Google Git
Sign in
fuchsia / third_party / Vulkan-ValidationLayers / HEAD / . / layers / sync / sync_validation.cpp
blob: 8731eec8e4b0973bbf7db5541fe0201fe34cb7e7 [file] [log] [blame]
/* Copyright (c) 2019-2023 The Khronos Group Inc.
* Copyright (c) 2019-2023 Valve Corporation
* Copyright (c) 2019-2023 LunarG, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include <algorithm>
#include <limits>
#include <memory>
#include <vector>
#include "sync/sync_validation.h"
#include "sync/sync_utils.h"
// Utilities to DRY up Get... calls
template <typename Map, typename Key = typename Map::key_type, typename RetVal = std::optional<typename Map::mapped_type>>
RetVal GetMappedOptional(const Map &map, const Key &key) {
RetVal ret_val;
auto it = map.find(key);
if (it != map.cend()) {
ret_val.emplace(it->second);
}
return ret_val;
}
template <typename Map, typename Fn>
typename Map::mapped_type GetMapped(const Map &map, const typename Map::key_type &key, Fn &&default_factory) {
auto value = GetMappedOptional(map, key);
return (value) ? *value : default_factory();
}
template <typename Map, typename Key = typename Map::key_type, typename Mapped = typename Map::mapped_type,
typename Value = typename Mapped::element_type>
Value *GetMappedPlainFromShared(const Map &map, const Key &key) {
auto value = GetMappedOptional<Map, Key>(map, key);
if (value) return value->get();
return nullptr;
}
static bool SimpleBinding(const BINDABLE &bindable) { return !bindable.sparse && bindable.Binding(); }
static const ResourceAccessRange kFullRange(std::numeric_limits<VkDeviceSize>::min(), std::numeric_limits<VkDeviceSize>::max());
static const char *string_SyncHazardVUID(SyncHazard hazard) {
switch (hazard) {
case SyncHazard::NONE:
return "SYNC-HAZARD-NONE";
break;
case SyncHazard::READ_AFTER_WRITE:
return "SYNC-HAZARD-READ-AFTER-WRITE";
break;
case SyncHazard::WRITE_AFTER_READ:
return "SYNC-HAZARD-WRITE-AFTER-READ";
break;
case SyncHazard::WRITE_AFTER_WRITE:
return "SYNC-HAZARD-WRITE-AFTER-WRITE";
break;
case SyncHazard::READ_RACING_WRITE:
return "SYNC-HAZARD-READ-RACING-WRITE";
break;
case SyncHazard::WRITE_RACING_WRITE:
return "SYNC-HAZARD-WRITE-RACING-WRITE";
break;
case SyncHazard::WRITE_RACING_READ:
return "SYNC-HAZARD-WRITE-RACING-READ";
break;
case SyncHazard::READ_AFTER_PRESENT:
return "SYNC-HAZARD-READ-AFTER-PRESENT";
break;
case SyncHazard::WRITE_AFTER_PRESENT:
return "SYNC-HAZARD-WRITE-AFTER-PRESENT";
break;
case SyncHazard::PRESENT_AFTER_WRITE:
return "SYNC-HAZARD-PRESENT-AFTER-WRITE";
break;
case SyncHazard::PRESENT_AFTER_READ:
return "SYNC-HAZARD-PRESENT-AFTER-READ";
break;
default:
assert(0);
}
return "SYNC-HAZARD-INVALID";
}
static bool IsHazardVsRead(SyncHazard hazard) {
bool vs_read = false;
switch (hazard) {
case SyncHazard::WRITE_AFTER_READ:
vs_read = true;
break;
case SyncHazard::WRITE_RACING_READ:
vs_read = true;
break;
case SyncHazard::PRESENT_AFTER_READ:
vs_read = true;
break;
default:
break;
}
return vs_read;
}
static const char *string_SyncHazard(SyncHazard hazard) {
switch (hazard) {
case SyncHazard::NONE:
return "NONE";
break;
case SyncHazard::READ_AFTER_WRITE:
return "READ_AFTER_WRITE";
break;
case SyncHazard::WRITE_AFTER_READ:
return "WRITE_AFTER_READ";
break;
case SyncHazard::WRITE_AFTER_WRITE:
return "WRITE_AFTER_WRITE";
break;
case SyncHazard::READ_RACING_WRITE:
return "READ_RACING_WRITE";
break;
case SyncHazard::WRITE_RACING_WRITE:
return "WRITE_RACING_WRITE";
break;
case SyncHazard::WRITE_RACING_READ:
return "WRITE_RACING_READ";
break;
case SyncHazard::READ_AFTER_PRESENT:
return "READ_AFTER_PRESENT";
break;
case SyncHazard::WRITE_AFTER_PRESENT:
return "WRITE_AFTER_PRESENT";
break;
case SyncHazard::PRESENT_AFTER_WRITE:
return "PRESENT_AFTER_WRITE";
break;
case SyncHazard::PRESENT_AFTER_READ:
return "PRESENT_AFTER_READ";
break;
default:
assert(0);
}
return "INVALID HAZARD";
}
static const SyncStageAccessInfoType *SyncStageAccessInfoFromMask(SyncStageAccessFlags flags) {
// Return the info for the first bit found
const SyncStageAccessInfoType *info = nullptr;
for (size_t i = 0; i < flags.size(); i++) {
if (flags.test(i)) {
info = &syncStageAccessInfoByStageAccessIndex()[i];
break;
}
}
return info;
}
static std::string string_SyncStageAccessFlags(const SyncStageAccessFlags &flags, const char *sep = "|") {
std::string out_str;
if (flags.none()) {
out_str = "0";
} else {
for (size_t i = 0; i < syncStageAccessInfoByStageAccessIndex().size(); i++) {
const auto &info = syncStageAccessInfoByStageAccessIndex()[i];
if ((flags & info.stage_access_bit).any()) {
if (!out_str.empty()) {
out_str.append(sep);
}
out_str.append(info.name);
}
}
if (out_str.length() == 0) {
out_str.append("Unhandled SyncStageAccess");
}
}
return out_str;
}
struct SyncNodeFormatter {
const debug_report_data *report_data;
const BASE_NODE *node;
const char *label;
SyncNodeFormatter(const SyncValidator &sync_state, const CMD_BUFFER_STATE *cb_state)
: report_data(sync_state.report_data), node(cb_state), label("command_buffer") {}
SyncNodeFormatter(const SyncValidator &sync_state, const IMAGE_STATE *image)
: report_data(sync_state.report_data), node(image), label("image") {}
SyncNodeFormatter(const SyncValidator &sync_state, const QUEUE_STATE *q_state)
: report_data(sync_state.report_data), node(q_state), label("queue") {}
SyncNodeFormatter(const SyncValidator &sync_state, const BASE_NODE *base_node, const char *label_ = nullptr)
: report_data(sync_state.report_data), node(base_node), label(label_) {}
};
std::ostream &operator<<(std::ostream &out, const SyncNodeFormatter &formatter) {
if (formatter.label) {
out << formatter.label << ": ";
}
if (formatter.node) {
out << formatter.report_data->FormatHandle(*formatter.node).c_str();
if (formatter.node->Destroyed()) {
out << " (destroyed)";
}
} else {
out << "null handle";
}
return out;
}
std::ostream &operator<<(std::ostream &out, const NamedHandle::FormatterState &formatter) {
const NamedHandle &handle = formatter.that;
bool labeled = false;
if (!handle.name.empty()) {
out << handle.name;
labeled = true;
}
if (handle.IsIndexed()) {
out << "[" << handle.index << "]";
labeled = true;
}
if (labeled) {
out << ": ";
}
out << formatter.state.FormatHandle(handle.handle);
return out;
}
std::ostream &operator<<(std::ostream &out, const ResourceUsageRecord::FormatterState &formatter) {
const ResourceUsageRecord &record = formatter.record;
if (record.alt_usage) {
out << record.alt_usage.Formatter(formatter.sync_state);
} else {
out << "command: " << vvl::String(record.command);
out << ", seq_no: " << record.seq_num;
if (record.sub_command != 0) {
out << ", subcmd: " << record.sub_command;
}
// Note: ex_cb_state set to null forces output of record.cb_state
if (!formatter.ex_cb_state || (formatter.ex_cb_state != record.cb_state)) {
out << ", " << SyncNodeFormatter(formatter.sync_state, record.cb_state);
}
for (const auto &named_handle : record.handles) {
out << "," << named_handle.Formatter(formatter.sync_state);
}
out << ", reset_no: " << std::to_string(record.reset_count);
}
return out;
}
std::ostream &operator<<(std::ostream &out, const HazardResult::HazardState &hazard) {
assert(hazard.usage_index < static_cast<SyncStageAccessIndex>(syncStageAccessInfoByStageAccessIndex().size()));
const auto &usage_info = syncStageAccessInfoByStageAccessIndex()[hazard.usage_index];
const auto *info = SyncStageAccessInfoFromMask(hazard.prior_access);
const char *stage_access_name = info ? info->name : "INVALID_STAGE_ACCESS";
out << "(";
if (!hazard.recorded_access.get()) {
// if we have a recorded usage the usage is reported from the recorded contexts point of view
out << "usage: " << usage_info.name << ", ";
}
out << "prior_usage: " << stage_access_name;
if (IsHazardVsRead(hazard.hazard)) {
const auto barriers = hazard.access_state->GetReadBarriers(hazard.prior_access);
out << ", read_barriers: " << string_VkPipelineStageFlags2(barriers);
} else {
SyncStageAccessFlags write_barrier = hazard.access_state->GetWriteBarriers();
out << ", write_barriers: " << string_SyncStageAccessFlags(write_barrier);
}
return out;
}
struct NoopBarrierAction {
explicit NoopBarrierAction() {}
void operator()(ResourceAccessState *access) const {}
const bool layout_transition = false;
};
static void InitSubpassContexts(VkQueueFlags queue_flags, const RENDER_PASS_STATE &rp_state, const AccessContext *external_context,
std::vector<AccessContext> &subpass_contexts) {
const auto &create_info = rp_state.createInfo;
// Add this for all subpasses here so that they exsist during next subpass validation
subpass_contexts.clear();
subpass_contexts.reserve(create_info.subpassCount);
for (uint32_t pass = 0; pass < create_info.subpassCount; pass++) {
subpass_contexts.emplace_back(pass, queue_flags, rp_state.subpass_dependencies, subpass_contexts, external_context);
}
}
// NOTE: Make sure the proxy doesn't outlive from, as the proxy is pointing directly to access contexts owned by from.
CommandBufferAccessContext::CommandBufferAccessContext(const CommandBufferAccessContext &from, AsProxyContext dummy)
: CommandBufferAccessContext(from.sync_state_) {
// Copy only the needed fields out of from for a temporary, proxy command buffer context
cb_state_ = from.cb_state_;
access_log_ = std::make_shared<AccessLog>(*from.access_log_); // potentially large, but no choice given tagging lookup.
command_number_ = from.command_number_;
subcommand_number_ = from.subcommand_number_;
reset_count_ = from.reset_count_;
const auto *from_context = from.GetCurrentAccessContext();
assert(from_context);
// Construct a fully resolved single access context out of from
const NoopBarrierAction noop_barrier;
from_context->ResolveAccessRange(kFullRange, noop_barrier, &cb_access_context_.GetAccessStateMap(), nullptr);
// The proxy has flatten the current render pass context (if any), but the async contexts are needed for hazard detection
cb_access_context_.ImportAsyncContexts(*from_context);
events_context_ = from.events_context_;
// We don't want to copy the full render_pass_context_ history just for the proxy.
}
std::string CommandBufferAccessContext::FormatUsage(const ResourceUsageTag tag) const {
if (tag >= access_log_->size()) return std::string();
std::stringstream out;
assert(tag < access_log_->size());
const auto &record = (*access_log_)[tag];
out << record.Formatter(*sync_state_, cb_state_);
return out.str();
}
std::string CommandBufferAccessContext::FormatUsage(const ResourceFirstAccess &access) const {
std::stringstream out;
assert(access.usage_info);
out << "(recorded_usage: " << access.usage_info->name;
out << ", " << FormatUsage(access.tag) << ")";
return out.str();
}
std::string CommandExecutionContext::FormatHazard(const HazardResult &hazard) const {
std::stringstream out;
assert(hazard.IsHazard());
out << hazard.State();
out << ", " << FormatUsage(hazard.Tag()) << ")";
return out.str();
}
bool CommandExecutionContext::ValidForSyncOps() const {
const bool valid = GetCurrentEventsContext() && GetCurrentAccessContext();
assert(valid);
return valid;
}
// NOTE: the attachement read flag is put *only* in the access scope and not in the exect scope, since the ordering
// rules apply only to this specific access for this stage, and not the stage as a whole. The ordering detection
// also reflects this special case for read hazard detection (using access instead of exec scope)
static constexpr VkPipelineStageFlags2KHR kColorAttachmentExecScope = VK_PIPELINE_STAGE_2_COLOR_ATTACHMENT_OUTPUT_BIT_KHR;
static const SyncStageAccessFlags kColorAttachmentAccessScope =
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_READ_BIT |
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_READ_NONCOHERENT_BIT_EXT |
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE_BIT |
SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ_BIT; // Note: this is intentionally not in the exec scope
static constexpr VkPipelineStageFlags2KHR kDepthStencilAttachmentExecScope =
VK_PIPELINE_STAGE_2_EARLY_FRAGMENT_TESTS_BIT_KHR | VK_PIPELINE_STAGE_2_LATE_FRAGMENT_TESTS_BIT_KHR;
static const SyncStageAccessFlags kDepthStencilAttachmentAccessScope =
SYNC_EARLY_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_READ_BIT | SYNC_EARLY_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE_BIT |
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_READ_BIT | SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE_BIT |
SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ_BIT; // Note: this is intentionally not in the exec scope
static constexpr VkPipelineStageFlags2KHR kRasterAttachmentExecScope = kDepthStencilAttachmentExecScope | kColorAttachmentExecScope;
static const SyncStageAccessFlags kRasterAttachmentAccessScope = kDepthStencilAttachmentAccessScope | kColorAttachmentAccessScope;
ResourceAccessState::OrderingBarriers ResourceAccessState::kOrderingRules = {
{{VK_PIPELINE_STAGE_2_NONE_KHR, SyncStageAccessFlags()},
{kColorAttachmentExecScope, kColorAttachmentAccessScope},
{kDepthStencilAttachmentExecScope, kDepthStencilAttachmentAccessScope},
{kRasterAttachmentExecScope, kRasterAttachmentAccessScope}}};
// Sometimes we have an internal access conflict, and we using the kInvalidTag to set and detect in temporary/proxy contexts
static const ResourceUsageTag kInvalidTag(ResourceUsageRecord::kMaxIndex);
static VkDeviceSize ResourceBaseAddress(const BUFFER_STATE &buffer) { return buffer.GetFakeBaseAddress(); }
template <typename T>
static ResourceAccessRange MakeRange(const T &has_offset_and_size) {
return ResourceAccessRange(has_offset_and_size.offset, (has_offset_and_size.offset + has_offset_and_size.size));
}
static ResourceAccessRange MakeRange(VkDeviceSize start, VkDeviceSize size) { return ResourceAccessRange(start, (start + size)); }
static ResourceAccessRange MakeRange(const BUFFER_STATE &buffer, VkDeviceSize offset, VkDeviceSize size) {
return MakeRange(offset, buffer.ComputeSize(offset, size));
}
static ResourceAccessRange MakeRange(const BUFFER_VIEW_STATE &buf_view_state) {
return MakeRange(*buf_view_state.buffer_state.get(), buf_view_state.create_info.offset, buf_view_state.create_info.range);
}
static ResourceAccessRange MakeRange(VkDeviceSize offset, uint32_t first_index, uint32_t count, uint32_t stride) {
const VkDeviceSize range_start = offset + (first_index * stride);
const VkDeviceSize range_size = count * stride;
return MakeRange(range_start, range_size);
}
static ResourceAccessRange MakeRange(const BufferBinding &binding, uint32_t first_index, const std::optional<uint32_t> &count,
uint32_t stride) {
if (count) {
return MakeRange(binding.offset, first_index, count.value(), stride);
}
return MakeRange(binding);
}
// Range generators for to allow event scope filtration to be limited to the top of the resource access traversal pipeline
//
// Note: there is no "begin/end" or reset facility. These are each written as "one time through" generators.
//
// Usage:
// Constructor() -- initializes the generator to point to the begin of the space declared.
// * -- the current range of the generator empty signfies end
// ++ -- advance to the next non-empty range (or end)
// A wrapper for a single range with the same semantics as the actual generators below
template <typename KeyType>
class SingleRangeGenerator {
public:
SingleRangeGenerator(const KeyType &range) : current_(range) {}
const KeyType &operator*() const { return current_; }
const KeyType *operator->() const { return &current_; }
SingleRangeGenerator &operator++() {
current_ = KeyType(); // just one real range
return *this;
}
bool operator==(const SingleRangeGenerator &other) const { return current_ == other.current_; }
private:
SingleRangeGenerator() = default;
const KeyType range_;
KeyType current_;
};
// Generate the ranges that are the intersection of range and the entries in the RangeMap
template <typename RangeMap, typename KeyType = typename RangeMap::key_type>
class MapRangesRangeGenerator {
public:
// Default constructed is safe to dereference for "empty" test, but for no other operation.
MapRangesRangeGenerator() : range_(), map_(nullptr), map_pos_(), current_() {
// Default construction for KeyType *must* be empty range
assert(current_.empty());
}
MapRangesRangeGenerator(const RangeMap &filter, const KeyType &range) : range_(range), map_(&filter), map_pos_(), current_() {
SeekBegin();
}
MapRangesRangeGenerator(const MapRangesRangeGenerator &from) = default;
const KeyType &operator*() const { return current_; }
const KeyType *operator->() const { return &current_; }
MapRangesRangeGenerator &operator++() {
++map_pos_;
UpdateCurrent();
return *this;
}
bool operator==(const MapRangesRangeGenerator &other) const { return current_ == other.current_; }
protected:
void UpdateCurrent() {
if (map_pos_ != map_->cend()) {
current_ = range_ & map_pos_->first;
} else {
current_ = KeyType();
}
}
void SeekBegin() {
map_pos_ = map_->lower_bound(range_);
UpdateCurrent();
}
// Adding this functionality here, to avoid gratuitous Base:: qualifiers in the derived class
// Note: Not exposed in this classes public interface to encourage using a consistent ++/empty generator semantic
template <typename Pred>
MapRangesRangeGenerator &PredicatedIncrement(Pred &pred) {
do {
++map_pos_;
} while (map_pos_ != map_->cend() && map_pos_->first.intersects(range_) && !pred(map_pos_));
UpdateCurrent();
return *this;
}
const KeyType range_;
const RangeMap *map_;
typename RangeMap::const_iterator map_pos_;
KeyType current_;
};
using EventSimpleRangeGenerator = MapRangesRangeGenerator<SyncEventState::ScopeMap>;
// Generate the ranges that are the intersection of the RangeGen ranges and the entries in the FilterMap
// Templated to allow for different Range generators or map sources...
template <typename RangeMap, typename RangeGen, typename KeyType = typename RangeMap::key_type>
class FilteredGeneratorGenerator {
public:
// Default constructed is safe to dereference for "empty" test, but for no other operation.
FilteredGeneratorGenerator() : filter_(nullptr), gen_(), filter_pos_(), current_() {
// Default construction for KeyType *must* be empty range
assert(current_.empty());
}
FilteredGeneratorGenerator(const RangeMap &filter, RangeGen &gen) : filter_(&filter), gen_(gen), filter_pos_(), current_() {
SeekBegin();
}
FilteredGeneratorGenerator(const FilteredGeneratorGenerator &from) = default;
const KeyType &operator*() const { return current_; }
const KeyType *operator->() const { return &current_; }
FilteredGeneratorGenerator &operator++() {
KeyType gen_range = GenRange();
KeyType filter_range = FilterRange();
current_ = KeyType();
while (gen_range.non_empty() && filter_range.non_empty() && current_.empty()) {
if (gen_range.end > filter_range.end) {
// if the generated range is beyond the filter_range, advance the filter range
filter_range = AdvanceFilter();
} else {
gen_range = AdvanceGen();
}
current_ = gen_range & filter_range;
}
return *this;
}
bool operator==(const FilteredGeneratorGenerator &other) const { return current_ == other.current_; }
private:
KeyType AdvanceFilter() {
++filter_pos_;
auto filter_range = FilterRange();
if (filter_range.valid()) {
FastForwardGen(filter_range);
}
return filter_range;
}
KeyType AdvanceGen() {
++gen_;
auto gen_range = GenRange();
if (gen_range.valid()) {
FastForwardFilter(gen_range);
}
return gen_range;
}
KeyType FilterRange() const { return (filter_pos_ != filter_->cend()) ? filter_pos_->first : KeyType(); }
KeyType GenRange() const { return *gen_; }
KeyType FastForwardFilter(const KeyType &range) {
auto filter_range = FilterRange();
int retry_count = 0;
const static int kRetryLimit = 2; // TODO -- determine whether this limit is optimal
while (!filter_range.empty() && (filter_range.end <= range.begin)) {
if (retry_count < kRetryLimit) {
++filter_pos_;
filter_range = FilterRange();
retry_count++;
} else {
// Okay we've tried walking, do a seek.
filter_pos_ = filter_->lower_bound(range);
break;
}
}
return FilterRange();
}
// TODO: Consider adding "seek" (or an absolute bound "get" to range generators to make this walk
// faster.
KeyType FastForwardGen(const KeyType &range) {
auto gen_range = GenRange();
while (!gen_range.empty() && (gen_range.end <= range.begin)) {
++gen_;
gen_range = GenRange();
}
return gen_range;
}
void SeekBegin() {
auto gen_range = GenRange();
if (gen_range.empty()) {
current_ = KeyType();
filter_pos_ = filter_->cend();
} else {
filter_pos_ = filter_->lower_bound(gen_range);
current_ = gen_range & FilterRange();
}
}
const RangeMap *filter_;
RangeGen gen_;
typename RangeMap::const_iterator filter_pos_;
KeyType current_;
};
using EventImageRangeGenerator = FilteredGeneratorGenerator<SyncEventState::ScopeMap, subresource_adapter::ImageRangeGenerator>;
SyncStageAccessIndex GetSyncStageAccessIndexsByDescriptorSet(VkDescriptorType descriptor_type,
const ResourceInterfaceVariable &variable,
VkShaderStageFlagBits stage_flag) {
if (descriptor_type == VK_DESCRIPTOR_TYPE_INPUT_ATTACHMENT) {
assert(stage_flag == VK_SHADER_STAGE_FRAGMENT_BIT);
return SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ;
}
const auto stage_accesses = sync_utils::GetShaderStageAccesses(stage_flag);
if (descriptor_type == VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER || descriptor_type == VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER_DYNAMIC) {
return stage_accesses.uniform_read;
}
// If the desriptorSet is writable, we don't need to care SHADER_READ. SHADER_WRITE is enough.
// Because if write hazard happens, read hazard might or might not happen.
// But if write hazard doesn't happen, read hazard is impossible to happen.
if (variable.is_written_to) {
return stage_accesses.storage_write;
} else if (descriptor_type == VK_DESCRIPTOR_TYPE_SAMPLED_IMAGE ||
descriptor_type == VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER ||
descriptor_type == VK_DESCRIPTOR_TYPE_UNIFORM_TEXEL_BUFFER) {
return stage_accesses.sampled_read;
} else {
return stage_accesses.storage_read;
}
}
bool IsImageLayoutDepthWritable(VkImageLayout image_layout) {
return (image_layout == VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL ||
image_layout == VK_IMAGE_LAYOUT_DEPTH_ATTACHMENT_STENCIL_READ_ONLY_OPTIMAL ||
image_layout == VK_IMAGE_LAYOUT_DEPTH_ATTACHMENT_OPTIMAL);
}
bool IsImageLayoutStencilWritable(VkImageLayout image_layout) {
return (image_layout == VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL ||
image_layout == VK_IMAGE_LAYOUT_DEPTH_READ_ONLY_STENCIL_ATTACHMENT_OPTIMAL ||
image_layout == VK_IMAGE_LAYOUT_STENCIL_ATTACHMENT_OPTIMAL);
}
// Tranverse the attachment resolves for this a specific subpass, and do action() to them.
// Used by both validation and record operations
//
// The signature for Action() reflect the needs of both uses.
template <typename Action>
void ResolveOperation(Action &action, const RENDER_PASS_STATE &rp_state, const AttachmentViewGenVector &attachment_views,
uint32_t subpass) {
const auto &rp_ci = rp_state.createInfo;
const auto *attachment_ci = rp_ci.pAttachments;
const auto &subpass_ci = rp_ci.pSubpasses[subpass];
// Color resolves -- require an inuse color attachment and a matching inuse resolve attachment
const auto *color_attachments = subpass_ci.pColorAttachments;
const auto *color_resolve = subpass_ci.pResolveAttachments;
if (color_resolve && color_attachments) {
for (uint32_t i = 0; i < subpass_ci.colorAttachmentCount; i++) {
const auto &color_attach = color_attachments[i].attachment;
const auto &resolve_attach = subpass_ci.pResolveAttachments[i].attachment;
if ((color_attach != VK_ATTACHMENT_UNUSED) && (resolve_attach != VK_ATTACHMENT_UNUSED)) {
action("color", "resolve read", color_attach, resolve_attach, attachment_views[color_attach],
AttachmentViewGen::Gen::kRenderArea, SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_READ,
SyncOrdering::kColorAttachment);
action("color", "resolve write", color_attach, resolve_attach, attachment_views[resolve_attach],
AttachmentViewGen::Gen::kRenderArea, SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE,
SyncOrdering::kColorAttachment);
}
}
}
// Depth stencil resolve only if the extension is present
const auto ds_resolve = vku::FindStructInPNextChain<VkSubpassDescriptionDepthStencilResolve>(subpass_ci.pNext);
if (ds_resolve && ds_resolve->pDepthStencilResolveAttachment &&
(ds_resolve->pDepthStencilResolveAttachment->attachment != VK_ATTACHMENT_UNUSED) && subpass_ci.pDepthStencilAttachment &&
(subpass_ci.pDepthStencilAttachment->attachment != VK_ATTACHMENT_UNUSED)) {
const auto src_at = subpass_ci.pDepthStencilAttachment->attachment;
const auto src_ci = attachment_ci[src_at];
// The formats are required to match so we can pick either
const bool resolve_depth = (ds_resolve->depthResolveMode != VK_RESOLVE_MODE_NONE) && vkuFormatHasDepth(src_ci.format);
const bool resolve_stencil = (ds_resolve->stencilResolveMode != VK_RESOLVE_MODE_NONE) && vkuFormatHasStencil(src_ci.format);
const auto dst_at = ds_resolve->pDepthStencilResolveAttachment->attachment;
// Figure out which aspects are actually touched during resolve operations
const char *aspect_string = nullptr;
AttachmentViewGen::Gen gen_type = AttachmentViewGen::Gen::kRenderArea;
if (resolve_depth && resolve_stencil) {
aspect_string = "depth/stencil";
} else if (resolve_depth) {
// Validate depth only
gen_type = AttachmentViewGen::Gen::kDepthOnlyRenderArea;
aspect_string = "depth";
} else if (resolve_stencil) {
// Validate all stencil only
gen_type = AttachmentViewGen::Gen::kStencilOnlyRenderArea;
aspect_string = "stencil";
}
if (aspect_string) {
action(aspect_string, "resolve read", src_at, dst_at, attachment_views[src_at], gen_type,
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_READ, SyncOrdering::kRaster);
action(aspect_string, "resolve write", src_at, dst_at, attachment_views[dst_at], gen_type,
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE, SyncOrdering::kRaster);
}
}
}
// Action for validating resolve operations
class ValidateResolveAction {
public:
ValidateResolveAction(VkRenderPass render_pass, uint32_t subpass, const AccessContext &context,
const CommandExecutionContext &exec_context, vvl::Func command)
: render_pass_(render_pass),
subpass_(subpass),
context_(context),
exec_context_(exec_context),
command_(command),
skip_(false) {}
void operator()(const char *aspect_name, const char *attachment_name, uint32_t src_at, uint32_t dst_at,
const AttachmentViewGen &view_gen, AttachmentViewGen::Gen gen_type, SyncStageAccessIndex current_usage,
SyncOrdering ordering_rule) {
HazardResult hazard;
hazard = context_.DetectHazard(view_gen, gen_type, current_usage, ordering_rule);
if (hazard.IsHazard()) {
skip_ |= exec_context_.GetSyncState().LogError(
render_pass_, string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s in subpass %" PRIu32 "during %s %s, from attachment %" PRIu32 " to resolve attachment %" PRIu32
". Access info %s.",
vvl::String(command_), string_SyncHazard(hazard.Hazard()), subpass_, aspect_name, attachment_name, src_at, dst_at,
exec_context_.FormatHazard(hazard).c_str());
}
}
// Providing a mechanism for the constructing caller to get the result of the validation
bool GetSkip() const { return skip_; }
private:
VkRenderPass render_pass_;
const uint32_t subpass_;
const AccessContext &context_;
const CommandExecutionContext &exec_context_;
vvl::Func command_;
bool skip_;
};
// Update action for resolve operations
class UpdateStateResolveAction {
public:
UpdateStateResolveAction(AccessContext &context, ResourceUsageTag tag) : context_(context), tag_(tag) {}
void operator()(const char *, const char *, uint32_t, uint32_t, const AttachmentViewGen &view_gen,
AttachmentViewGen::Gen gen_type, SyncStageAccessIndex current_usage, SyncOrdering ordering_rule) {
// Ignores validation only arguments...
context_.UpdateAccessState(view_gen, gen_type, current_usage, ordering_rule, tag_);
}
private:
AccessContext &context_;
const ResourceUsageTag tag_;
};
void HazardResult::Set(const ResourceAccessState *access_state_, const SyncStageAccessInfoType &usage_info_, SyncHazard hazard_,
const ResourceAccessWriteState &prior_write) {
state_.emplace(access_state_, usage_info_, hazard_, prior_write.Access().stage_access_bit, prior_write.Tag());
}
void HazardResult::Set(const ResourceAccessState *access_state_, const SyncStageAccessInfoType &usage_info_, SyncHazard hazard_,
const SyncStageAccessFlags &prior_, ResourceUsageTag tag_) {
state_.emplace(access_state_, usage_info_, hazard_, prior_, tag_);
}
void HazardResult::AddRecordedAccess(const ResourceFirstAccess &first_access) {
assert(state_.has_value());
state_->recorded_access = std::make_unique<const ResourceFirstAccess>(first_access);
}
bool HazardResult::IsWAWHazard() const {
assert(state_.has_value());
return (state_->hazard == WRITE_AFTER_WRITE) && (state_->prior_access[state_->usage_index]);
}
AccessContext::AccessContext(uint32_t subpass, VkQueueFlags queue_flags,
const std::vector<SubpassDependencyGraphNode> &dependencies,
const std::vector<AccessContext> &contexts, const AccessContext *external_context) {
Reset();
const auto &subpass_dep = dependencies[subpass];
const bool has_barrier_from_external = subpass_dep.barrier_from_external.size() > 0U;
prev_.reserve(subpass_dep.prev.size() + (has_barrier_from_external ? 1U : 0U));
prev_by_subpass_.resize(subpass, nullptr); // Can't be more prevs than the subpass we're on
for (const auto &prev_dep : subpass_dep.prev) {
const auto prev_pass = prev_dep.first->pass;
const auto &prev_barriers = prev_dep.second;
assert(prev_dep.second.size());
prev_.emplace_back(&contexts[prev_pass], queue_flags, prev_barriers);
prev_by_subpass_[prev_pass] = &prev_.back();
}
async_.reserve(subpass_dep.async.size());
for (const auto async_subpass : subpass_dep.async) {
// Start tags are not known at creation time (as it's done at BeginRenderpass)
async_.emplace_back(contexts[async_subpass], kInvalidTag);
}
if (has_barrier_from_external) {
// Store the barrier from external with the reat, but save pointer for "by subpass" lookups.
prev_.emplace_back(external_context, queue_flags, subpass_dep.barrier_from_external);
src_external_ = &prev_.back();
}
if (subpass_dep.barrier_to_external.size()) {
dst_external_ = TrackBack(this, queue_flags, subpass_dep.barrier_to_external);
}
}
void AccessContext::Trim() {
auto normalize = [](ResourceAccessRangeMap::value_type &access) { access.second.Normalize(); };
ForAll(normalize);
// Consolidate map after normalization, combines directly adjacent ranges with common values.
sparse_container::consolidate(access_state_map_);
}
void AccessContext::AddReferencedTags(ResourceUsageTagSet &used) const {
auto gather = [&used](const ResourceAccessRangeMap::value_type &access) { access.second.GatherReferencedTags(used); };
ConstForAll(gather);
}
template <typename Detector>
HazardResult AccessContext::DetectPreviousHazard(Detector &detector, const ResourceAccessRange &range) const {
ResourceAccessRangeMap descent_map;
ResolvePreviousAccess(range, &descent_map, nullptr);
HazardResult hazard;
for (auto prev = descent_map.begin(); prev != descent_map.end() && !hazard.IsHazard(); ++prev) {
hazard = detector.Detect(prev);
}
return hazard;
}
template <typename Action>
void AccessContext::ForAll(Action &&action) {
for (auto &access : access_state_map_) {
action(access);
}
}
template <typename Action>
void AccessContext::ConstForAll(Action &&action) const {
for (auto &access : access_state_map_) {
action(access);
}
}
template <typename Predicate>
void AccessContext::EraseIf(Predicate &&pred) {
// Note: Don't forward, we don't want r-values moved, since we're going to make multiple calls.
vvl::EraseIf(access_state_map_, pred);
}
template <typename Detector, typename RangeGen>
HazardResult AccessContext::DetectHazard(Detector &detector, RangeGen &range_gen, DetectOptions options) const {
for (; range_gen->non_empty(); ++range_gen) {
HazardResult hazard = DetectHazard(detector, *range_gen, options);
if (hazard.IsHazard()) return hazard;
}
return HazardResult();
}
// A recursive range walker for hazard detection, first for the current context and the (DetectHazardRecur) to walk
// the DAG of the contexts (for example subpasses)
template <typename Detector>
HazardResult AccessContext::DetectHazard(Detector &detector, const ResourceAccessRange &range, DetectOptions options) const {
HazardResult hazard;
if (static_cast<uint32_t>(options) & DetectOptions::kDetectAsync) {
// Async checks don't require recursive lookups, as the async lists are exhaustive for the top-level context
// so we'll check these first
for (const auto &async_ref : async_) {
hazard = async_ref.Context().DetectAsyncHazard(detector, range, async_ref.StartTag());
if (hazard.IsHazard()) return hazard;
}
}
const bool detect_prev = (static_cast<uint32_t>(options) & DetectOptions::kDetectPrevious) != 0;
const auto the_end = access_state_map_.cend(); // End is not invalidated
auto pos = access_state_map_.lower_bound(range);
ResourceAccessRange gap = {range.begin, range.begin};
while (pos != the_end && pos->first.begin < range.end) {
// Cover any leading gap, or gap between entries
if (detect_prev) {
// TODO: After profiling we may want to change the descent logic such that we don't recur per gap...
// Cover any leading gap, or gap between entries
gap.end = pos->first.begin; // We know this begin is < range.end
if (gap.non_empty()) {
// Recur on all gaps
hazard = DetectPreviousHazard(detector, gap);
if (hazard.IsHazard()) return hazard;
}
// Set up for the next gap. If pos..end is >= range.end, loop will exit, and trailing gap will be empty
gap.begin = pos->first.end;
}
hazard = detector.Detect(pos);
if (hazard.IsHazard()) return hazard;
++pos;
}
if (detect_prev) {
// Detect in the trailing empty as needed
gap.end = range.end;
if (gap.non_empty()) {
hazard = DetectPreviousHazard(detector, gap);
}
}
return hazard;
}
// A non recursive range walker for the asynchronous contexts (those we have no barriers with)
template <typename Detector>
HazardResult AccessContext::DetectAsyncHazard(const Detector &detector, const ResourceAccessRange &range,
ResourceUsageTag async_tag) const {
auto pos = access_state_map_.lower_bound(range);
const auto the_end = access_state_map_.end();
HazardResult hazard;
while (pos != the_end && pos->first.begin < range.end) {
hazard = detector.DetectAsync(pos, async_tag);
if (hazard.IsHazard()) break;
++pos;
}
return hazard;
}
struct ApplySubpassTransitionBarriersAction {
explicit ApplySubpassTransitionBarriersAction(const std::vector<SyncBarrier> &barriers_) : barriers(barriers_) {}
void operator()(ResourceAccessState *access) const {
assert(access);
access->ApplyBarriers(barriers, true);
}
const std::vector<SyncBarrier> &barriers;
};
struct QueueTagOffsetBarrierAction {
QueueTagOffsetBarrierAction(QueueId qid, ResourceUsageTag offset) : queue_id(qid), tag_offset(offset) {}
void operator()(ResourceAccessState *access) const {
access->OffsetTag(tag_offset);
access->SetQueueId(queue_id);
};
QueueId queue_id;
ResourceUsageTag tag_offset;
};
struct ApplyTrackbackStackAction {
explicit ApplyTrackbackStackAction(const std::vector<SyncBarrier> &barriers_,
const ResourceAccessStateFunction *previous_barrier_ = nullptr)
: barriers(barriers_), previous_barrier(previous_barrier_) {}
void operator()(ResourceAccessState *access) const {
assert(access);
assert(!access->HasPendingState());
access->ApplyBarriers(barriers, false);
// NOTE: We can use invalid tag, as these barriers do no include layout transitions (would assert in SetWrite)
access->ApplyPendingBarriers(kInvalidTag);
if (previous_barrier) {
assert(bool(*previous_barrier));
(*previous_barrier)(access);
}
}
const std::vector<SyncBarrier> &barriers;
const ResourceAccessStateFunction *previous_barrier;
};
static SyncBarrier MergeBarriers(const std::vector<SyncBarrier> &barriers) {
SyncBarrier merged = {};
for (const auto &barrier : barriers) {
merged.Merge(barrier);
}
return merged;
}
template <typename BarrierAction>
void AccessContext::ResolveAccessRange(const ResourceAccessRange &range, BarrierAction &barrier_action,
ResourceAccessRangeMap *resolve_map, const ResourceAccessState *infill_state,
bool recur_to_infill) const {
if (!range.non_empty()) return;
ResourceRangeMergeIterator current(*resolve_map, access_state_map_, range.begin);
while (current->range.non_empty() && range.includes(current->range.begin)) {
const auto current_range = current->range & range;
if (current->pos_B->valid) {
const auto &src_pos = current->pos_B->lower_bound;
ResourceAccessState access(src_pos->second); // intentional copy
barrier_action(&access);
if (current->pos_A->valid) {
const auto trimmed = sparse_container::split(current->pos_A->lower_bound, *resolve_map, current_range);
trimmed->second.Resolve(access);
current.invalidate_A(trimmed);
} else {
auto inserted = resolve_map->insert(current->pos_A->lower_bound, std::make_pair(current_range, access));
current.invalidate_A(inserted); // Update the parallel iterator to point at the insert segment
}
} else {
// we have to descend to fill this gap
if (recur_to_infill) {
ResourceAccessRange recurrence_range = current_range;
// The current context is empty for the current range, so recur to fill the gap.
// Since we will be recurring back up the DAG, expand the gap descent to cover the full range for which B
// is not valid, to minimize that recurrence
if (current->pos_B.at_end()) {
// Do the remainder here....
recurrence_range.end = range.end;
} else {
// Recur only over the range until B becomes valid (within the limits of range).
recurrence_range.end = std::min(range.end, current->pos_B->lower_bound->first.begin);
}
ResolvePreviousAccessStack(recurrence_range, resolve_map, infill_state, barrier_action);
// Given that there could be gaps we need to seek carefully to not repeatedly search the same gaps in the next
// iterator of the outer while.
// Set the parallel iterator to the end of this range s.t. ++ will move us to the next range whether or
// not the end of the range is a gap. For the seek to work, first we need to warn the parallel iterator
// we stepped on the dest map
const auto seek_to = recurrence_range.end - 1; // The subtraction is safe as range can't be empty (loop condition)
current.invalidate_A(); // Changes current->range
current.seek(seek_to);
} else if (!current->pos_A->valid && infill_state) {
// If we didn't find anything in the current range, and we aren't reccuring... we infill if required
auto inserted = resolve_map->insert(current->pos_A->lower_bound, std::make_pair(current->range, *infill_state));
current.invalidate_A(inserted); // Update the parallel iterator to point at the correct segment after insert
}
}
if (current->range.non_empty()) {
++current;
}
}
// Infill if range goes passed both the current and resolve map prior contents
if (recur_to_infill && (current->range.end < range.end)) {
ResourceAccessRange trailing_fill_range = {current->range.end, range.end};
ResolvePreviousAccessStack<BarrierAction>(trailing_fill_range, resolve_map, infill_state, barrier_action);
}
}
template <typename BarrierAction>
void AccessContext::ResolvePreviousAccessStack(const ResourceAccessRange &range, ResourceAccessRangeMap *descent_map,
const ResourceAccessState *infill_state,
const BarrierAction &previous_barrier) const {
ResourceAccessStateFunction stacked_barrier(std::ref(previous_barrier));
ResolvePreviousAccess(range, descent_map, infill_state, &stacked_barrier);
}
void AccessContext::ResolvePreviousAccess(const ResourceAccessRange &range, ResourceAccessRangeMap *descent_map,
const ResourceAccessState *infill_state,
const ResourceAccessStateFunction *previous_barrier) const {
if (prev_.size() == 0) {
if (range.non_empty() && infill_state) {
// Fill the empty poritions of descent_map with the default_state with the barrier function applied (iff present)
ResourceAccessState state_copy;
if (previous_barrier) {
assert(bool(*previous_barrier));
state_copy = *infill_state;
(*previous_barrier)(&state_copy);
infill_state = &state_copy;
}
sparse_container::update_range_value(*descent_map, range, *infill_state,
sparse_container::value_precedence::prefer_dest);
}
} else {
// Look for something to fill the gap further along.
for (const auto &prev_dep : prev_) {
const ApplyTrackbackStackAction barrier_action(prev_dep.barriers, previous_barrier);
prev_dep.source_subpass->ResolveAccessRange(range, barrier_action, descent_map, infill_state);
}
}
}
// Non-lazy import of all accesses, WaitEvents needs this.
void AccessContext::ResolvePreviousAccesses() {
ResourceAccessState default_state;
if (!prev_.size()) return; // If no previous contexts, nothing to do
ResolvePreviousAccess(kFullRange, &access_state_map_, &default_state);
}
static SyncStageAccessIndex ColorLoadUsage(VkAttachmentLoadOp load_op) {
const auto stage_access = (load_op == VK_ATTACHMENT_LOAD_OP_NONE_EXT)
? SYNC_ACCESS_INDEX_NONE
: ((load_op == VK_ATTACHMENT_LOAD_OP_LOAD) ? SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_READ
: SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE);
return stage_access;
}
static SyncStageAccessIndex DepthStencilLoadUsage(VkAttachmentLoadOp load_op) {
const auto stage_access =
(load_op == VK_ATTACHMENT_LOAD_OP_NONE_EXT)
? SYNC_ACCESS_INDEX_NONE
: ((load_op == VK_ATTACHMENT_LOAD_OP_LOAD) ? SYNC_EARLY_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_READ
: SYNC_EARLY_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE);
return stage_access;
}
// Caller must manage returned pointer
static AccessContext *CreateStoreResolveProxyContext(const AccessContext &context, const RENDER_PASS_STATE &rp_state,
uint32_t subpass, const AttachmentViewGenVector &attachment_views) {
auto *proxy = new AccessContext(context);
proxy->UpdateAttachmentResolveAccess(rp_state, attachment_views, subpass, kInvalidTag);
proxy->UpdateAttachmentStoreAccess(rp_state, attachment_views, subpass, kInvalidTag);
return proxy;
}
template <typename BarrierAction>
void AccessContext::ResolveAccessRange(const AttachmentViewGen &view_gen, AttachmentViewGen::Gen gen_type,
BarrierAction &barrier_action, ResourceAccessRangeMap *descent_map,
const ResourceAccessState *infill_state) const {
const std::optional<ImageRangeGen> &attachment_gen = view_gen.GetRangeGen(gen_type);
if (!attachment_gen) return;
subresource_adapter::ImageRangeGenerator range_gen(*attachment_gen);
for (; range_gen->non_empty(); ++range_gen) {
ResolveAccessRange(*range_gen, barrier_action, descent_map, infill_state);
}
}
template <typename ResolveOp>
void AccessContext::ResolveFromContext(ResolveOp &&resolve_op, const AccessContext &from_context,
const ResourceAccessState *infill_state, bool recur_to_infill) {
from_context.ResolveAccessRange(kFullRange, resolve_op, &access_state_map_, infill_state, recur_to_infill);
}
template <typename ResolveOp, typename RangeGenerator>
void AccessContext::ResolveFromContext(ResolveOp &&resolve_op, const AccessContext &from_context, RangeGenerator range_gen,
const ResourceAccessState *infill_state, bool recur_to_infill) {
for (; range_gen->non_empty(); ++range_gen) {
from_context.ResolveAccessRange(*range_gen, resolve_op, &access_state_map_, infill_state, recur_to_infill);
}
}
// Layout transitions are handled as if the were occuring in the beginning of the next subpass
bool AccessContext::ValidateLayoutTransitions(const CommandExecutionContext &exec_context, const RENDER_PASS_STATE &rp_state,
const VkRect2D &render_area, uint32_t subpass,
const AttachmentViewGenVector &attachment_views, vvl::Func command) const {
bool skip = false;
// As validation methods are const and precede the record/update phase, for any tranistions from the immediately
// previous subpass, we have to validate them against a copy of the AccessContext, with resolve operations applied, as
// those affects have not been recorded yet.
//
// Note: we could be more efficient by tracking whether or not we actually *have* any changes (e.g. attachment resolve)
// to apply and only copy then, if this proves a hot spot.
std::unique_ptr<AccessContext> proxy_for_prev;
TrackBack proxy_track_back;
const auto &transitions = rp_state.subpass_transitions[subpass];
for (const auto &transition : transitions) {
const bool prev_needs_proxy = transition.prev_pass != VK_SUBPASS_EXTERNAL && (transition.prev_pass + 1 == subpass);
const auto *track_back = GetTrackBackFromSubpass(transition.prev_pass);
assert(track_back);
if (prev_needs_proxy) {
if (!proxy_for_prev) {
proxy_for_prev.reset(
CreateStoreResolveProxyContext(*track_back->source_subpass, rp_state, transition.prev_pass, attachment_views));
proxy_track_back = *track_back;
proxy_track_back.source_subpass = proxy_for_prev.get();
}
track_back = &proxy_track_back;
}
auto hazard = DetectSubpassTransitionHazard(*track_back, attachment_views[transition.attachment]);
if (hazard.IsHazard()) {
if (hazard.Tag() == kInvalidTag) {
skip |= exec_context.GetSyncState().LogError(
rp_state.renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s in subpass %" PRIu32 " for attachment %" PRIu32
" image layout transition (old_layout: %s, new_layout: %s) after store/resolve operation in subpass %" PRIu32,
vvl::String(command), string_SyncHazard(hazard.Hazard()), subpass, transition.attachment,
string_VkImageLayout(transition.old_layout), string_VkImageLayout(transition.new_layout), transition.prev_pass);
} else {
skip |= exec_context.GetSyncState().LogError(
rp_state.renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s in subpass %" PRIu32 " for attachment %" PRIu32
" image layout transition (old_layout: %s, new_layout: %s). Access info %s.",
vvl::String(command), string_SyncHazard(hazard.Hazard()), subpass, transition.attachment,
string_VkImageLayout(transition.old_layout), string_VkImageLayout(transition.new_layout),
exec_context.FormatHazard(hazard).c_str());
}
}
}
return skip;
}
bool AccessContext::ValidateLoadOperation(const CommandExecutionContext &exec_context, const RENDER_PASS_STATE &rp_state,
const VkRect2D &render_area, uint32_t subpass,
const AttachmentViewGenVector &attachment_views, vvl::Func command) const {
bool skip = false;
const auto *attachment_ci = rp_state.createInfo.pAttachments;
for (uint32_t i = 0; i < rp_state.createInfo.attachmentCount; i++) {
if (subpass == rp_state.attachment_first_subpass[i]) {
const auto &view_gen = attachment_views[i];
if (!view_gen.IsValid()) continue;
const auto &ci = attachment_ci[i];
// Need check in the following way
// 1) if the usage bit isn't in the dest_access_scope, and there is layout traniition for initial use, report hazard
// vs. transition
// 2) if there isn't a layout transition, we need to look at the external context with a "detect hazard" operation
// for each aspect loaded.
const bool has_depth = vkuFormatHasDepth(ci.format);
const bool has_stencil = vkuFormatHasStencil(ci.format);
const bool is_color = !(has_depth || has_stencil);
const SyncStageAccessIndex load_index = has_depth ? DepthStencilLoadUsage(ci.loadOp) : ColorLoadUsage(ci.loadOp);
const SyncStageAccessIndex stencil_load_index = has_stencil ? DepthStencilLoadUsage(ci.stencilLoadOp) : load_index;
HazardResult hazard;
const char *aspect = nullptr;
bool checked_stencil = false;
if (is_color && (load_index != SYNC_ACCESS_INDEX_NONE)) {
hazard = DetectHazard(view_gen, AttachmentViewGen::Gen::kRenderArea, load_index, SyncOrdering::kColorAttachment);
aspect = "color";
} else {
if (has_depth && (load_index != SYNC_ACCESS_INDEX_NONE)) {
hazard = DetectHazard(view_gen, AttachmentViewGen::Gen::kDepthOnlyRenderArea, load_index,
SyncOrdering::kDepthStencilAttachment);
aspect = "depth";
}
if (!hazard.IsHazard() && has_stencil && (stencil_load_index != SYNC_ACCESS_INDEX_NONE)) {
hazard = DetectHazard(view_gen, AttachmentViewGen::Gen::kStencilOnlyRenderArea, stencil_load_index,
SyncOrdering::kDepthStencilAttachment);
aspect = "stencil";
checked_stencil = true;
}
}
if (hazard.IsHazard()) {
auto load_op_string = string_VkAttachmentLoadOp(checked_stencil ? ci.stencilLoadOp : ci.loadOp);
const auto &sync_state = exec_context.GetSyncState();
if (hazard.Tag() == kInvalidTag) {
// Hazard vs. ILT
skip |= sync_state.LogError(rp_state.renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s vs. layout transition in subpass %" PRIu32 " for attachment %" PRIu32
" aspect %s during load with loadOp %s.",
vvl::String(command), string_SyncHazard(hazard.Hazard()), subpass, i, aspect,
load_op_string);
} else {
skip |= sync_state.LogError(rp_state.renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s in subpass %" PRIu32 " for attachment %" PRIu32
" aspect %s during load with loadOp %s. Access info %s.",
vvl::String(command), string_SyncHazard(hazard.Hazard()), subpass, i, aspect,
load_op_string, exec_context.FormatHazard(hazard).c_str());
}
}
}
}
return skip;
}
// Store operation validation can ignore resolve (before it) and layout tranistions after it. The first is ignored
// because of the ordering guarantees w.r.t. sample access and that the resolve validation hasn't altered the state, because
// store is part of the same Next/End operation.
// The latter is handled in layout transistion validation directly
bool AccessContext::ValidateStoreOperation(const CommandExecutionContext &exec_context, const RENDER_PASS_STATE &rp_state,
const VkRect2D &render_area, uint32_t subpass,
const AttachmentViewGenVector &attachment_views, vvl::Func command) const {
bool skip = false;
const auto *attachment_ci = rp_state.createInfo.pAttachments;
for (uint32_t i = 0; i < rp_state.createInfo.attachmentCount; i++) {
if (subpass == rp_state.attachment_last_subpass[i]) {
const AttachmentViewGen &view_gen = attachment_views[i];
if (!view_gen.IsValid()) continue;
const auto &ci = attachment_ci[i];
// The spec states that "don't care" is an operation with VK_ACCESS_COLOR_ATTACHMENT_WRITE_BIT,
// so we assume that an implementation is *free* to write in that case, meaning that for correctness
// sake, we treat DONT_CARE as writing.
const bool has_depth = vkuFormatHasDepth(ci.format);
const bool has_stencil = vkuFormatHasStencil(ci.format);
const bool is_color = !(has_depth || has_stencil);
const bool store_op_stores = ci.storeOp != VK_ATTACHMENT_STORE_OP_NONE_EXT;
if (!has_stencil && !store_op_stores) continue;
HazardResult hazard;
const char *aspect = nullptr;
bool checked_stencil = false;
if (is_color) {
hazard = DetectHazard(view_gen, AttachmentViewGen::Gen::kRenderArea,
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE, SyncOrdering::kRaster);
aspect = "color";
} else {
const bool stencil_op_stores = ci.stencilStoreOp != VK_ATTACHMENT_STORE_OP_NONE_EXT;
if (has_depth && store_op_stores) {
hazard = DetectHazard(view_gen, AttachmentViewGen::Gen::kDepthOnlyRenderArea,
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE, SyncOrdering::kRaster);
aspect = "depth";
}
if (!hazard.IsHazard() && has_stencil && stencil_op_stores) {
hazard = DetectHazard(view_gen, AttachmentViewGen::Gen::kStencilOnlyRenderArea,
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE, SyncOrdering::kRaster);
aspect = "stencil";
checked_stencil = true;
}
}
if (hazard.IsHazard()) {
const char *const op_type_string = checked_stencil ? "stencilStoreOp" : "storeOp";
const char *const store_op_string = string_VkAttachmentStoreOp(checked_stencil ? ci.stencilStoreOp : ci.storeOp);
skip |= exec_context.GetSyncState().LogError(rp_state.renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s in subpass %" PRIu32 " for attachment %" PRIu32
" %s aspect during store with %s %s. Access info %s",
vvl::String(command), string_SyncHazard(hazard.Hazard()), subpass, i,
aspect, op_type_string, store_op_string,
exec_context.FormatHazard(hazard).c_str());
}
}
}
return skip;
}
bool AccessContext::ValidateResolveOperations(const CommandExecutionContext &exec_context, const RENDER_PASS_STATE &rp_state,
const VkRect2D &render_area, const AttachmentViewGenVector &attachment_views,
vvl::Func command, uint32_t subpass) const {
ValidateResolveAction validate_action(rp_state.renderPass(), subpass, *this, exec_context, command);
ResolveOperation(validate_action, rp_state, attachment_views, subpass);
return validate_action.GetSkip();
}
void AccessContext::AddAsyncContext(const AccessContext *context, ResourceUsageTag tag) {
if (context) {
async_.emplace_back(*context, tag);
}
}
class HazardDetector {
const SyncStageAccessInfoType &usage_info_;
public:
HazardResult Detect(const ResourceAccessRangeMap::const_iterator &pos) const { return pos->second.DetectHazard(usage_info_); }
HazardResult DetectAsync(const ResourceAccessRangeMap::const_iterator &pos, ResourceUsageTag start_tag) const {
return pos->second.DetectAsyncHazard(usage_info_, start_tag);
}
explicit HazardDetector(SyncStageAccessIndex usage_index) : usage_info_(SyncStageAccess::UsageInfo(usage_index)) {}
};
class HazardDetectorWithOrdering {
const SyncStageAccessInfoType &usage_info_;
const SyncOrdering ordering_rule_;
public:
HazardResult Detect(const ResourceAccessRangeMap::const_iterator &pos) const {
return pos->second.DetectHazard(usage_info_, ordering_rule_, QueueSyncState::kQueueIdInvalid);
}
HazardResult DetectAsync(const ResourceAccessRangeMap::const_iterator &pos, ResourceUsageTag start_tag) const {
return pos->second.DetectAsyncHazard(usage_info_, start_tag);
}
HazardDetectorWithOrdering(SyncStageAccessIndex usage_index, SyncOrdering ordering)
: usage_info_(SyncStageAccess::UsageInfo(usage_index)), ordering_rule_(ordering) {}
};
HazardResult AccessContext::DetectHazard(const BUFFER_STATE &buffer, SyncStageAccessIndex usage_index,
const ResourceAccessRange &range) const {
if (!SimpleBinding(buffer)) return HazardResult();
const auto base_address = ResourceBaseAddress(buffer);
HazardDetector detector(usage_index);
return DetectHazard(detector, (range + base_address), DetectOptions::kDetectAll);
}
template <typename Detector>
HazardResult AccessContext::DetectHazard(Detector &detector, const AttachmentViewGen &view_gen, AttachmentViewGen::Gen gen_type,
DetectOptions options) const {
const std::optional<ImageRangeGen> &attachment_gen = view_gen.GetRangeGen(gen_type);
if (!attachment_gen) return HazardResult();
subresource_adapter::ImageRangeGenerator range_gen(*attachment_gen);
return DetectHazard(detector, range_gen, options);
}
template <typename Detector>
HazardResult AccessContext::DetectHazard(Detector &detector, const ImageState &image,
const VkImageSubresourceRange &subresource_range, const VkOffset3D &offset,
const VkExtent3D &extent, bool is_depth_sliced, DetectOptions options) const {
// range_gen is non-temporary to avoid additional copy
ImageRangeGen range_gen = image.MakeImageRangeGen(subresource_range, offset, extent, is_depth_sliced);
return DetectHazard(detector, range_gen, options);
}
template <typename Detector>
HazardResult AccessContext::DetectHazard(Detector &detector, const ImageState &image,
const VkImageSubresourceRange &subresource_range, bool is_depth_sliced,
DetectOptions options) const {
// range_gen is non-temporary to avoid additional copy
ImageRangeGen range_gen = image.MakeImageRangeGen(subresource_range, is_depth_sliced);
return DetectHazard(detector, range_gen, options);
}
HazardResult AccessContext::DetectHazard(const ImageState &image, SyncStageAccessIndex current_usage,
const VkImageSubresourceLayers &subresource, const VkOffset3D &offset,
const VkExtent3D &extent, bool is_depth_sliced) const {
VkImageSubresourceRange subresource_range = {subresource.aspectMask, subresource.mipLevel, 1, subresource.baseArrayLayer,
subresource.layerCount};
HazardDetector detector(current_usage);
return DetectHazard(detector, image, subresource_range, offset, extent, is_depth_sliced, DetectOptions::kDetectAll);
}
HazardResult AccessContext::DetectHazard(const ImageState &image, SyncStageAccessIndex current_usage,
const VkImageSubresourceRange &subresource_range, bool is_depth_sliced) const {
HazardDetector detector(current_usage);
return DetectHazard(detector, image, subresource_range, is_depth_sliced, DetectOptions::kDetectAll);
}
HazardResult AccessContext::DetectHazard(const ImageViewState &image_view, SyncStageAccessIndex current_usage) const {
// Get is const, but callee will copy
HazardDetector detector(current_usage);
return DetectHazard(detector, image_view.GetFullViewImageRangeGen(), DetectOptions::kDetectAll);
}
HazardResult AccessContext::DetectHazard(const ImageViewState &image_view, SyncStageAccessIndex current_usage,
SyncOrdering ordering_rule, const VkOffset3D &offset, const VkExtent3D &extent) const {
// range_gen is non-temporary to avoid an additional copy
ImageRangeGen range_gen(image_view.MakeImageRangeGen(offset, extent));
HazardDetectorWithOrdering detector(current_usage, ordering_rule);
return DetectHazard(detector, range_gen, DetectOptions::kDetectAll);
}
HazardResult AccessContext::DetectHazard(const AttachmentViewGen &view_gen, AttachmentViewGen::Gen gen_type,
SyncStageAccessIndex current_usage, SyncOrdering ordering_rule) const {
HazardDetectorWithOrdering detector(current_usage, ordering_rule);
return DetectHazard(detector, view_gen, gen_type, DetectOptions::kDetectAll);
}
HazardResult AccessContext::DetectHazard(const ImageState &image, SyncStageAccessIndex current_usage,
const VkImageSubresourceRange &subresource_range, SyncOrdering ordering_rule,
const VkOffset3D &offset, const VkExtent3D &extent, bool is_depth_sliced) const {
HazardDetectorWithOrdering detector(current_usage, ordering_rule);
return DetectHazard(detector, image, subresource_range, offset, extent, is_depth_sliced, DetectOptions::kDetectAll);
}
class BarrierHazardDetector {
public:
BarrierHazardDetector(SyncStageAccessIndex usage_index, VkPipelineStageFlags2KHR src_exec_scope,
SyncStageAccessFlags src_access_scope)
: usage_info_(SyncStageAccess::UsageInfo(usage_index)),
src_exec_scope_(src_exec_scope),
src_access_scope_(src_access_scope) {}
HazardResult Detect(const ResourceAccessRangeMap::const_iterator &pos) const {
return pos->second.DetectBarrierHazard(usage_info_, QueueSyncState::kQueueIdInvalid, src_exec_scope_, src_access_scope_);
}
HazardResult DetectAsync(const ResourceAccessRangeMap::const_iterator &pos, ResourceUsageTag start_tag) const {
// Async barrier hazard detection can use the same path as the usage index is not IsRead, but is IsWrite
return pos->second.DetectAsyncHazard(usage_info_, start_tag);
}
private:
const SyncStageAccessInfoType &usage_info_;
VkPipelineStageFlags2KHR src_exec_scope_;
SyncStageAccessFlags src_access_scope_;
};
class EventBarrierHazardDetector {
public:
EventBarrierHazardDetector(SyncStageAccessIndex usage_index, VkPipelineStageFlags2KHR src_exec_scope,
SyncStageAccessFlags src_access_scope, const SyncEventState::ScopeMap &event_scope, QueueId queue_id,
ResourceUsageTag scope_tag)
: usage_info_(SyncStageAccess::UsageInfo(usage_index)),
src_exec_scope_(src_exec_scope),
src_access_scope_(src_access_scope),
event_scope_(event_scope),
scope_queue_id_(queue_id),
scope_tag_(scope_tag),
scope_pos_(event_scope.cbegin()),
scope_end_(event_scope.cend()) {}
HazardResult Detect(const ResourceAccessRangeMap::const_iterator &pos) {
// Need to piece together coverage of pos->first range:
// Copy the range as we'll be chopping it up as needed
ResourceAccessRange range = pos->first;
const ResourceAccessState &access = pos->second;
HazardResult hazard;
bool in_scope = AdvanceScope(range);
bool unscoped_tested = false;
while (in_scope && !hazard.IsHazard()) {
if (range.begin < ScopeBegin()) {
if (!unscoped_tested) {
unscoped_tested = true;
hazard = access.DetectHazard(usage_info_);
}
// Note: don't need to check for in_scope as AdvanceScope true means range and ScopeRange intersect.
// Thus a [ ScopeBegin, range.end ) will be non-empty.
range.begin = ScopeBegin();
} else { // in_scope implied that ScopeRange and range intersect
hazard = access.DetectBarrierHazard(usage_info_, ScopeState(), src_exec_scope_, src_access_scope_, scope_queue_id_,
scope_tag_);
if (!hazard.IsHazard()) {
range.begin = ScopeEnd();
in_scope = AdvanceScope(range); // contains a non_empty check
}
}
}
if (range.non_empty() && !hazard.IsHazard() && !unscoped_tested) {
hazard = access.DetectHazard(usage_info_);
}
return hazard;
}
HazardResult DetectAsync(const ResourceAccessRangeMap::const_iterator &pos, ResourceUsageTag start_tag) const {
// Async barrier hazard detection can use the same path as the usage index is not IsRead, but is IsWrite
return pos->second.DetectAsyncHazard(usage_info_, start_tag);
}
private:
bool ScopeInvalid() const { return scope_pos_ == scope_end_; }
bool ScopeValid() const { return !ScopeInvalid(); }
void ScopeSeek(const ResourceAccessRange &range) { scope_pos_ = event_scope_.lower_bound(range); }
// Hiding away the std::pair grunge...
ResourceAddress ScopeBegin() const { return scope_pos_->first.begin; }
ResourceAddress ScopeEnd() const { return scope_pos_->first.end; }
const ResourceAccessRange &ScopeRange() const { return scope_pos_->first; }
const ResourceAccessState &ScopeState() const { return scope_pos_->second; }
bool AdvanceScope(const ResourceAccessRange &range) {
// Note: non_empty is (valid && !empty), so don't change !non_empty to empty...
if (!range.non_empty()) return false;
if (ScopeInvalid()) return false;
if (ScopeRange().strictly_less(range)) {
ScopeSeek(range);
}
return ScopeValid() && ScopeRange().intersects(range);
}
const SyncStageAccessInfoType usage_info_;
VkPipelineStageFlags2KHR src_exec_scope_;
SyncStageAccessFlags src_access_scope_;
const SyncEventState::ScopeMap &event_scope_;
QueueId scope_queue_id_;
const ResourceUsageTag scope_tag_;
SyncEventState::ScopeMap::const_iterator scope_pos_;
SyncEventState::ScopeMap::const_iterator scope_end_;
};
HazardResult AccessContext::DetectImageBarrierHazard(const ImageState &image, const VkImageSubresourceRange &subresource_range,
VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope, QueueId queue_id,
const SyncEventState &sync_event, AccessContext::DetectOptions options) const {
// It's not particularly DRY to get the address type in this function as well as lower down, but we have to select the
// first access scope map to use, and there's no easy way to plumb it in below.
const auto &event_scope = sync_event.FirstScope();
EventBarrierHazardDetector detector(SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION, src_exec_scope, src_access_scope,
event_scope, queue_id, sync_event.first_scope_tag);
return DetectHazard(detector, image, subresource_range, false, options);
}
HazardResult AccessContext::DetectImageBarrierHazard(const AttachmentViewGen &view_gen, const SyncBarrier &barrier,
DetectOptions options) const {
BarrierHazardDetector detector(SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION, barrier.src_exec_scope.exec_scope,
barrier.src_access_scope);
return DetectHazard(detector, view_gen, AttachmentViewGen::Gen::kViewSubresource, options);
}
HazardResult AccessContext::DetectImageBarrierHazard(const ImageState &image, VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope,
const VkImageSubresourceRange &subresource_range,
const DetectOptions options) const {
BarrierHazardDetector detector(SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION, src_exec_scope, src_access_scope);
return DetectHazard(detector, image, subresource_range, false, options);
}
HazardResult AccessContext::DetectImageBarrierHazard(const SyncImageMemoryBarrier &image_barrier) const {
return DetectImageBarrierHazard(*image_barrier.image.get(), image_barrier.barrier.src_exec_scope.exec_scope,
image_barrier.barrier.src_access_scope, image_barrier.range, kDetectAll);
}
template <typename Flags, typename Map>
SyncStageAccessFlags AccessScopeImpl(Flags flag_mask, const Map &map) {
SyncStageAccessFlags scope;
for (const auto &bit_scope : map) {
if (flag_mask < bit_scope.first) break;
if (flag_mask & bit_scope.first) {
scope |= bit_scope.second;
}
}
return scope;
}
SyncStageAccessFlags SyncStageAccess::AccessScopeByStage(VkPipelineStageFlags2KHR stages) {
return AccessScopeImpl(stages, syncStageAccessMaskByStageBit());
}
SyncStageAccessFlags SyncStageAccess::AccessScopeByAccess(VkAccessFlags2KHR accesses) {
return AccessScopeImpl(sync_utils::ExpandAccessFlags(accesses), syncStageAccessMaskByAccessBit());
}
// Getting from stage mask and access mask to stage/access masks is something we need to be good at...
SyncStageAccessFlags SyncStageAccess::AccessScope(VkPipelineStageFlags2KHR stages, VkAccessFlags2KHR accesses) {
// The access scope is the intersection of all stage/access types possible for the enabled stages and the enables
// accesses (after doing a couple factoring of common terms the union of stage/access intersections is the intersections
// of the union of all stage/access types for all the stages and the same unions for the access mask...
return AccessScopeByStage(stages) & AccessScopeByAccess(accesses);
}
// The semantics of the InfillUpdateOps of infill_update_range are slightly different than for the UpdateMemoryAccessState Action
// operations, as this simplifies the generic traversal. So we wrap them in a semantics Adapter to get the same effect.
template <typename Action>
struct ActionToOpsAdapter {
using Map = ResourceAccessRangeMap;
using Range = typename Map::key_type;
using Iterator = typename Map::iterator;
using IndexType = typename Map::index_type;
void infill(Map &accesses, const Iterator &pos, const Range &infill_range) const {
// Combine Infill and update operations to make the generic implementation simpler
Iterator infill = action.Infill(&accesses, pos, infill_range);
if (infill == accesses.end()) return; // Allow action to 'pass' on filling in the blanks
// Need to apply the action to the Infill. 'infill_update_range' expect ops.infill to be completely done with
// the infill_range, where as Action::Infill assumes the caller will apply the action() logic to the infill_range
for (; infill != pos; ++infill) {
assert(infill != accesses.end());
action(infill);
}
}
void update(const Iterator &pos) const { action(pos); }
const Action &action;
};
template <typename Action, typename RangeGen>
void UpdateMemoryAccessState(ResourceAccessRangeMap &accesses, const Action &action, RangeGen &range_gen) {
ActionToOpsAdapter<Action> ops{action};
for (; range_gen->non_empty(); ++range_gen) {
infill_update_range(accesses, *range_gen, ops);
}
}
template <typename Action>
void UpdateMemoryAccessRangeState(ResourceAccessRangeMap &accesses, Action &action, const ResourceAccessRange &range) {
ActionToOpsAdapter<Action> ops{action};
infill_update_range(accesses, range, ops);
}
struct UpdateMemoryAccessStateFunctor {
using Iterator = ResourceAccessRangeMap::iterator;
Iterator Infill(ResourceAccessRangeMap *accesses, const Iterator &pos, const ResourceAccessRange &range) const {
// this is only called on gaps, and never returns a gap.
ResourceAccessState default_state;
context.ResolvePreviousAccess(range, accesses, &default_state);
return accesses->lower_bound(range);
}
void operator()(const Iterator &pos) const {
auto &access_state = pos->second;
access_state.Update(usage_info, ordering_rule, tag);
}
UpdateMemoryAccessStateFunctor(const AccessContext &context_, SyncStageAccessIndex usage_, SyncOrdering ordering_rule_,
ResourceUsageTag tag_)
: context(context_), usage_info(SyncStageAccess::UsageInfo(usage_)), ordering_rule(ordering_rule_), tag(tag_) {}
const AccessContext &context;
const SyncStageAccessInfoType &usage_info;
const SyncOrdering ordering_rule;
const ResourceUsageTag tag;
};
// The barrier operation for pipeline and subpass dependencies`
struct PipelineBarrierOp {
SyncBarrier barrier;
bool layout_transition;
ResourceAccessState::QueueScopeOps scope;
PipelineBarrierOp(QueueId queue_id, const SyncBarrier &barrier_, bool layout_transition_)
: barrier(barrier_), layout_transition(layout_transition_), scope(queue_id) {
if (queue_id != QueueSyncState::kQueueIdInvalid) {
// This is a submit time application... supress layout transitions to not taint the QueueBatchContext write state
layout_transition = false;
}
}
PipelineBarrierOp(const PipelineBarrierOp &rhs)
: barrier(rhs.barrier), layout_transition(rhs.layout_transition), scope(rhs.scope) {}
void operator()(ResourceAccessState *access_state) const { access_state->ApplyBarrier(scope, barrier, layout_transition); }
};
// Batch barrier ops don't modify in place, and thus don't need to hold pending state, and also are *never* layout transitions.
struct BatchBarrierOp : public PipelineBarrierOp {
void operator()(ResourceAccessState *access_state) const {
access_state->ApplyBarrier(scope, barrier, layout_transition);
access_state->ApplyPendingBarriers(kInvalidTag); // There can't be any need for this tag
}
BatchBarrierOp(QueueId queue_id, const SyncBarrier &barrier_) : PipelineBarrierOp(queue_id, barrier_, false) {}
};
// The barrier operation for wait events
struct WaitEventBarrierOp {
ResourceAccessState::EventScopeOps scope_ops;
SyncBarrier barrier;
bool layout_transition;
WaitEventBarrierOp(const QueueId scope_queue_, const ResourceUsageTag scope_tag_, const SyncBarrier &barrier_,
bool layout_transition_)
: scope_ops(scope_queue_, scope_tag_), barrier(barrier_), layout_transition(layout_transition_) {
if (scope_queue_ != QueueSyncState::kQueueIdInvalid) {
// This is a submit time application... supress layout transitions to not taint the QueueBatchContext write state
layout_transition = false;
}
}
void operator()(ResourceAccessState *access_state) const { access_state->ApplyBarrier(scope_ops, barrier, layout_transition); }
};
// This functor applies a collection of barriers, updating the "pending state" in each touched memory range, and optionally
// resolves the pending state. Suitable for processing Global memory barriers, or Subpass Barriers when the "final" barrier
// of a collection is known/present.
template <typename BarrierOp, typename OpVector = std::vector<BarrierOp>>
class ApplyBarrierOpsFunctor {
public:
using Iterator = ResourceAccessRangeMap::iterator;
// Only called with a gap, and pos at the lower_bound(range)
inline Iterator Infill(ResourceAccessRangeMap *accesses, const Iterator &pos, const ResourceAccessRange &range) const {
if (!infill_default_) {
return pos;
}
ResourceAccessState default_state;
auto inserted = accesses->insert(pos, std::make_pair(range, default_state));
return inserted;
}
void operator()(const Iterator &pos) const {
auto &access_state = pos->second;
for (const auto &op : barrier_ops_) {
op(&access_state);
}
if (resolve_) {
// If this is the last (or only) batch, we can do the pending resolve as the last step in this operation to avoid
// another walk
access_state.ApplyPendingBarriers(tag_);
}
}
// A valid tag is required IFF layout_transition is true, as transitions are write ops
ApplyBarrierOpsFunctor(bool resolve, typename OpVector::size_type size_hint, ResourceUsageTag tag)
: resolve_(resolve), infill_default_(false), barrier_ops_(), tag_(tag) {
barrier_ops_.reserve(size_hint);
}
void EmplaceBack(const BarrierOp &op) {
barrier_ops_.emplace_back(op);
infill_default_ |= op.layout_transition;
}
private:
bool resolve_;
bool infill_default_;
OpVector barrier_ops_;
const ResourceUsageTag tag_;
};
// This functor applies a single barrier, updating the "pending state" in each touched memory range, but does not
// resolve the pendinging state. Suitable for processing Image and Buffer barriers from PipelineBarriers or Events
template <typename BarrierOp>
class ApplyBarrierFunctor : public ApplyBarrierOpsFunctor<BarrierOp, small_vector<BarrierOp, 1>> {
using Base = ApplyBarrierOpsFunctor<BarrierOp, small_vector<BarrierOp, 1>>;
public:
ApplyBarrierFunctor(const BarrierOp &barrier_op) : Base(false, 1, kInvalidTag) { Base::EmplaceBack(barrier_op); }
};
// This functor resolves the pendinging state.
class ResolvePendingBarrierFunctor : public ApplyBarrierOpsFunctor<NoopBarrierAction, small_vector<NoopBarrierAction, 1>> {
using Base = ApplyBarrierOpsFunctor<NoopBarrierAction, small_vector<NoopBarrierAction, 1>>;
public:
ResolvePendingBarrierFunctor(ResourceUsageTag tag) : Base(true, 0, tag) {}
};
void AccessContext::UpdateAccessState(const BUFFER_STATE &buffer, SyncStageAccessIndex current_usage, SyncOrdering ordering_rule,
const ResourceAccessRange &range, const ResourceUsageTag tag) {
if (!SimpleBinding(buffer)) return;
const auto base_address = ResourceBaseAddress(buffer);
UpdateMemoryAccessStateFunctor action(*this, current_usage, ordering_rule, tag);
UpdateMemoryAccessRangeState(access_state_map_, action, range + base_address);
}
void AccessContext::UpdateAccessState(const ImageState &image, SyncStageAccessIndex current_usage, SyncOrdering ordering_rule,
const VkImageSubresourceRange &subresource_range, const ResourceUsageTag &tag) {
// range_gen is non-temporary to avoid an additional copy
ImageRangeGen range_gen = image.MakeImageRangeGen(subresource_range, false);
UpdateAccessState(range_gen, current_usage, ordering_rule, tag);
}
void AccessContext::UpdateAccessState(const ImageState &image, SyncStageAccessIndex current_usage, SyncOrdering ordering_rule,
const VkImageSubresourceRange &subresource_range, const VkOffset3D &offset,
const VkExtent3D &extent, const ResourceUsageTag tag) {
// range_gen is non-temporary to avoid an additional copy
ImageRangeGen range_gen = image.MakeImageRangeGen(subresource_range, offset, extent, false);
UpdateAccessState(range_gen, current_usage, ordering_rule, tag);
}
void AccessContext::UpdateAccessState(const ImageViewState &image_view, SyncStageAccessIndex current_usage,
SyncOrdering ordering_rule, const VkOffset3D &offset, const VkExtent3D &extent,
const ResourceUsageTag tag) {
// range_gen is non-temporary to avoid an additional copy
ImageRangeGen range_gen(image_view.MakeImageRangeGen(offset, extent));
UpdateAccessState(range_gen, current_usage, ordering_rule, tag);
}
void AccessContext::UpdateAccessState(const ImageViewState &image_view, SyncStageAccessIndex current_usage,
SyncOrdering ordering_rule, ResourceUsageTag tag) {
// Get is const, and will be copied in callee
UpdateAccessState(image_view.GetFullViewImageRangeGen(), current_usage, ordering_rule, tag);
}
void AccessContext::UpdateAccessState(const AttachmentViewGen &view_gen, AttachmentViewGen::Gen gen_type,
SyncStageAccessIndex current_usage, SyncOrdering ordering_rule, const ResourceUsageTag tag) {
const std::optional<ImageRangeGen> &attachment_gen = view_gen.GetRangeGen(gen_type);
if (attachment_gen) {
// Value of const optional is const, and will be copied in callee
UpdateAccessState(*attachment_gen, current_usage, ordering_rule, tag);
}
}
void AccessContext::UpdateAccessState(const ImageState &image, SyncStageAccessIndex current_usage, SyncOrdering ordering_rule,
const VkImageSubresourceLayers &subresource, const VkOffset3D &offset,
const VkExtent3D &extent, const ResourceUsageTag tag) {
VkImageSubresourceRange subresource_range = {subresource.aspectMask, subresource.mipLevel, 1, subresource.baseArrayLayer,
subresource.layerCount};
UpdateAccessState(image, current_usage, ordering_rule, subresource_range, offset, extent, tag);
}
void AccessContext::UpdateAccessState(ImageRangeGen &range_gen, SyncStageAccessIndex current_usage, SyncOrdering ordering_rule,
ResourceUsageTag tag) {
UpdateMemoryAccessStateFunctor action(*this, current_usage, ordering_rule, tag);
UpdateMemoryAccessState(access_state_map_, action, range_gen);
}
void AccessContext::UpdateAccessState(const ImageRangeGen &range_gen, SyncStageAccessIndex current_usage,
SyncOrdering ordering_rule, ResourceUsageTag tag) {
// range_gen is non-temporary to avoid infinite call recursion
ImageRangeGen mutable_range_gen(range_gen);
UpdateAccessState(mutable_range_gen, current_usage, ordering_rule, tag);
}
template <typename Action>
void AccessContext::ApplyUpdateAction(const AttachmentViewGen &view_gen, AttachmentViewGen::Gen gen_type, const Action &action) {
const std::optional<ImageRangeGen> &ref_range_gen = view_gen.GetRangeGen(gen_type);
if (ref_range_gen) {
ImageRangeGen range_gen(*ref_range_gen);
UpdateMemoryAccessState(access_state_map_, action, range_gen);
}
}
void AccessContext::UpdateAttachmentResolveAccess(const RENDER_PASS_STATE &rp_state,
const AttachmentViewGenVector &attachment_views, uint32_t subpass,
const ResourceUsageTag tag) {
UpdateStateResolveAction update(*this, tag);
ResolveOperation(update, rp_state, attachment_views, subpass);
}
void AccessContext::UpdateAttachmentStoreAccess(const RENDER_PASS_STATE &rp_state, const AttachmentViewGenVector &attachment_views,
uint32_t subpass, const ResourceUsageTag tag) {
const auto *attachment_ci = rp_state.createInfo.pAttachments;
for (uint32_t i = 0; i < rp_state.createInfo.attachmentCount; i++) {
if (rp_state.attachment_last_subpass[i] == subpass) {
const auto &view_gen = attachment_views[i];
if (!view_gen.IsValid()) continue; // UNUSED
const auto &ci = attachment_ci[i];
const bool has_depth = vkuFormatHasDepth(ci.format);
const bool has_stencil = vkuFormatHasStencil(ci.format);
const bool is_color = !(has_depth || has_stencil);
const bool store_op_stores = ci.storeOp != VK_ATTACHMENT_STORE_OP_NONE_EXT;
if (is_color && store_op_stores) {
UpdateAccessState(view_gen, AttachmentViewGen::Gen::kRenderArea,
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE, SyncOrdering::kRaster, tag);
} else {
if (has_depth && store_op_stores) {
UpdateAccessState(view_gen, AttachmentViewGen::Gen::kDepthOnlyRenderArea,
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE, SyncOrdering::kRaster, tag);
}
const bool stencil_op_stores = ci.stencilStoreOp != VK_ATTACHMENT_STORE_OP_NONE_EXT;
if (has_stencil && stencil_op_stores) {
UpdateAccessState(view_gen, AttachmentViewGen::Gen::kStencilOnlyRenderArea,
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE, SyncOrdering::kRaster, tag);
}
}
}
}
}
template <typename Action>
void AccessContext::ApplyToContext(const Action &barrier_action) {
// Note: Barriers do *not* cross context boundaries, applying to accessess within.... (at least for renderpass subpasses)
UpdateMemoryAccessRangeState(access_state_map_, barrier_action, kFullRange);
}
void AccessContext::ResolveChildContexts(const std::vector<AccessContext> &contexts) {
for (uint32_t subpass_index = 0; subpass_index < contexts.size(); subpass_index++) {
auto &context = contexts[subpass_index];
ApplyTrackbackStackAction barrier_action(context.GetDstExternalTrackBack().barriers);
context.ResolveAccessRange(kFullRange, barrier_action, &access_state_map_, nullptr, false);
}
}
// Caller must ensure that lifespan of this is less than the lifespan of from
void AccessContext::ImportAsyncContexts(const AccessContext &from) {
async_.insert(async_.end(), from.async_.begin(), from.async_.end());
}
// Suitable only for *subpass* access contexts
HazardResult AccessContext::DetectSubpassTransitionHazard(const TrackBack &track_back, const AttachmentViewGen &attach_view) const {
if (!attach_view.IsValid()) return HazardResult();
// We should never ask for a transition from a context we don't have
assert(track_back.source_subpass);
// Do the detection against the specific prior context independent of other contexts. (Synchronous only)
// Hazard detection for the transition can be against the merged of the barriers (it only uses src_...)
const auto merged_barrier = MergeBarriers(track_back.barriers);
HazardResult hazard = track_back.source_subpass->DetectImageBarrierHazard(attach_view, merged_barrier, kDetectPrevious);
if (!hazard.IsHazard()) {
// The Async hazard check is against the current context's async set.
SyncBarrier null_barrier = {};
hazard = DetectImageBarrierHazard(attach_view, null_barrier, kDetectAsync);
}
return hazard;
}
void AccessContext::RecordLayoutTransitions(const RENDER_PASS_STATE &rp_state, uint32_t subpass,
const AttachmentViewGenVector &attachment_views, const ResourceUsageTag tag) {
const auto &transitions = rp_state.subpass_transitions[subpass];
const ResourceAccessState empty_infill;
for (const auto &transition : transitions) {
const auto prev_pass = transition.prev_pass;
const auto &view_gen = attachment_views[transition.attachment];
if (!view_gen.IsValid()) continue;
const auto *trackback = GetTrackBackFromSubpass(prev_pass);
assert(trackback);
// Import the attachments into the current context
const auto *prev_context = trackback->source_subpass;
assert(prev_context);
ApplySubpassTransitionBarriersAction barrier_action(trackback->barriers);
prev_context->ResolveAccessRange(view_gen, AttachmentViewGen::Gen::kViewSubresource, barrier_action, &access_state_map_,
&empty_infill);
}
// If there were no transitions skip this global map walk
if (transitions.size()) {
ResolvePendingBarrierFunctor apply_pending_action(tag);
ApplyToContext(apply_pending_action);
}
}
bool CommandBufferAccessContext::ValidateDispatchDrawDescriptorSet(VkPipelineBindPoint pipelineBindPoint,
const Location &loc) const {
bool skip = false;
const PIPELINE_STATE *pipe = nullptr;
const std::vector<LAST_BOUND_STATE::PER_SET> *per_sets = nullptr;
cb_state_->GetCurrentPipelineAndDesriptorSets(pipelineBindPoint, &pipe, &per_sets);
if (!pipe || !per_sets) {
return skip;
}
using DescriptorClass = cvdescriptorset::DescriptorClass;
using BufferDescriptor = cvdescriptorset::BufferDescriptor;
using ImageDescriptor = cvdescriptorset::ImageDescriptor;
using TexelDescriptor = cvdescriptorset::TexelDescriptor;
for (const auto &stage_state : pipe->stage_states) {
const auto raster_state = pipe->RasterizationState();
if (stage_state.GetStage() == VK_SHADER_STAGE_FRAGMENT_BIT && raster_state && raster_state->rasterizerDiscardEnable) {
continue;
} else if (!stage_state.entrypoint) {
continue;
}
for (const auto &variable : stage_state.entrypoint->resource_interface_variables) {
if (variable.decorations.set >= per_sets->size()) {
// This should be caught by Core validation, but if core checks are disabled SyncVal should not crash.
continue;
}
const auto *descriptor_set = (*per_sets)[variable.decorations.set].bound_descriptor_set.get();
if (!descriptor_set) continue;
auto binding = descriptor_set->GetBinding(variable.decorations.binding);
const auto descriptor_type = binding->type;
SyncStageAccessIndex sync_index =
GetSyncStageAccessIndexsByDescriptorSet(descriptor_type, variable, stage_state.GetStage());
for (uint32_t index = 0; index < binding->count; index++) {
const auto *descriptor = binding->GetDescriptor(index);
switch (descriptor->GetClass()) {
case DescriptorClass::ImageSampler:
case DescriptorClass::Image: {
if (descriptor->Invalid()) {
continue;
}
// NOTE: ImageSamplerDescriptor inherits from ImageDescriptor, so this cast works for both types.
const auto *image_descriptor = static_cast<const ImageDescriptor *>(descriptor);
const auto *img_view_state =
static_cast<const syncval_state::ImageViewState *>(image_descriptor->GetImageViewState());
VkImageLayout image_layout = image_descriptor->GetImageLayout();
if (img_view_state->IsDepthSliced()) {
// NOTE: 2D ImageViews of VK_IMAGE_CREATE_2D_ARRAY_COMPATIBLE_BIT Images are not allowed in
// Descriptors, unless VK_EXT_image_2d_view_of_3d is supported, which it isn't at the moment.
// See: VUID 00343
continue;
}
HazardResult hazard;
if (sync_index == SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ) {
const VkExtent3D extent = CastTo3D(cb_state_->active_render_pass_begin_info.renderArea.extent);
const VkOffset3D offset = CastTo3D(cb_state_->active_render_pass_begin_info.renderArea.offset);
// Input attachments are subject to raster ordering rules
hazard =
current_context_->DetectHazard(*img_view_state, sync_index, SyncOrdering::kRaster, offset, extent);
} else {
hazard = current_context_->DetectHazard(*img_view_state, sync_index);
}
if (hazard.IsHazard() && !sync_state_->SupressedBoundDescriptorWAW(hazard)) {
skip |= sync_state_->LogError(
string_SyncHazardVUID(hazard.Hazard()), img_view_state->image_view(), loc,
"Hazard %s for %s, in %s, and %s, %s, type: %s, imageLayout: %s, binding #%" PRIu32
", index %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), sync_state_->FormatHandle(img_view_state->image_view()).c_str(),
sync_state_->FormatHandle(cb_state_->commandBuffer()).c_str(),
sync_state_->FormatHandle(pipe->pipeline()).c_str(),
sync_state_->FormatHandle(descriptor_set->GetSet()).c_str(),
string_VkDescriptorType(descriptor_type), string_VkImageLayout(image_layout),
variable.decorations.binding, index, FormatHazard(hazard).c_str());
}
break;
}
case DescriptorClass::TexelBuffer: {
const auto *texel_descriptor = static_cast<const TexelDescriptor *>(descriptor);
if (texel_descriptor->Invalid()) {
continue;
}
const auto *buf_view_state = texel_descriptor->GetBufferViewState();
const auto *buf_state = buf_view_state->buffer_state.get();
const ResourceAccessRange range = MakeRange(*buf_view_state);
auto hazard = current_context_->DetectHazard(*buf_state, sync_index, range);
if (hazard.IsHazard() && !sync_state_->SupressedBoundDescriptorWAW(hazard)) {
skip |= sync_state_->LogError(
string_SyncHazardVUID(hazard.Hazard()), buf_view_state->buffer_view(), loc,
"Hazard %s for %s in %s, %s, and %s, type: %s, binding #%d index %d. Access info %s.",
string_SyncHazard(hazard.Hazard()),
sync_state_->FormatHandle(buf_view_state->buffer_view()).c_str(),
sync_state_->FormatHandle(cb_state_->commandBuffer()).c_str(),
sync_state_->FormatHandle(pipe->pipeline()).c_str(),
sync_state_->FormatHandle(descriptor_set->GetSet()).c_str(),
string_VkDescriptorType(descriptor_type), variable.decorations.binding, index,
FormatHazard(hazard).c_str());
}
break;
}
case DescriptorClass::GeneralBuffer: {
const auto *buffer_descriptor = static_cast<const BufferDescriptor *>(descriptor);
if (buffer_descriptor->Invalid()) {
continue;
}
const auto *buf_state = buffer_descriptor->GetBufferState();
const ResourceAccessRange range =
MakeRange(*buf_state, buffer_descriptor->GetOffset(), buffer_descriptor->GetRange());
auto hazard = current_context_->DetectHazard(*buf_state, sync_index, range);
if (hazard.IsHazard() && !sync_state_->SupressedBoundDescriptorWAW(hazard)) {
skip |= sync_state_->LogError(
string_SyncHazardVUID(hazard.Hazard()), buf_state->buffer(), loc,
"Hazard %s for %s in %s, %s, and %s, type: %s, binding #%d index %d. Access info %s.",
string_SyncHazard(hazard.Hazard()), sync_state_->FormatHandle(buf_state->buffer()).c_str(),
sync_state_->FormatHandle(cb_state_->commandBuffer()).c_str(),
sync_state_->FormatHandle(pipe->pipeline()).c_str(),
sync_state_->FormatHandle(descriptor_set->GetSet()).c_str(),
string_VkDescriptorType(descriptor_type), variable.decorations.binding, index,
FormatHazard(hazard).c_str());
}
break;
}
// TODO: INLINE_UNIFORM_BLOCK_EXT, ACCELERATION_STRUCTURE_KHR
default:
break;
}
}
}
}
return skip;
}
void CommandBufferAccessContext::RecordDispatchDrawDescriptorSet(VkPipelineBindPoint pipelineBindPoint,
const ResourceUsageTag tag) {
const PIPELINE_STATE *pipe = nullptr;
const std::vector<LAST_BOUND_STATE::PER_SET> *per_sets = nullptr;
cb_state_->GetCurrentPipelineAndDesriptorSets(pipelineBindPoint, &pipe, &per_sets);
if (!pipe || !per_sets) {
return;
}
using DescriptorClass = cvdescriptorset::DescriptorClass;
using BufferDescriptor = cvdescriptorset::BufferDescriptor;
using ImageDescriptor = cvdescriptorset::ImageDescriptor;
using TexelDescriptor = cvdescriptorset::TexelDescriptor;
for (const auto &stage_state : pipe->stage_states) {
const auto raster_state = pipe->RasterizationState();
if (stage_state.GetStage() == VK_SHADER_STAGE_FRAGMENT_BIT && raster_state && raster_state->rasterizerDiscardEnable) {
continue;
} else if (!stage_state.entrypoint) {
continue;
}
for (const auto &variable : stage_state.entrypoint->resource_interface_variables) {
if (variable.decorations.set >= per_sets->size()) {
// This should be caught by Core validation, but if core checks are disabled SyncVal should not crash.
continue;
}
const auto *descriptor_set = (*per_sets)[variable.decorations.set].bound_descriptor_set.get();
if (!descriptor_set) continue;
auto binding = descriptor_set->GetBinding(variable.decorations.binding);
const auto descriptor_type = binding->type;
SyncStageAccessIndex sync_index =
GetSyncStageAccessIndexsByDescriptorSet(descriptor_type, variable, stage_state.GetStage());
for (uint32_t i = 0; i < binding->count; i++) {
const auto *descriptor = binding->GetDescriptor(i);
switch (descriptor->GetClass()) {
case DescriptorClass::ImageSampler:
case DescriptorClass::Image: {
// NOTE: ImageSamplerDescriptor inherits from ImageDescriptor, so this cast works for both types.
const auto *image_descriptor = static_cast<const ImageDescriptor *>(descriptor);
if (image_descriptor->Invalid()) {
continue;
}
const auto *img_view_state =
static_cast<const syncval_state::ImageViewState *>(image_descriptor->GetImageViewState());
if (img_view_state->IsDepthSliced()) {
// NOTE: 2D ImageViews of VK_IMAGE_CREATE_2D_ARRAY_COMPATIBLE_BIT Images are not allowed in
// Descriptors, unless VK_EXT_image_2d_view_of_3d is supported, which it isn't at the moment.
// See: VUID 00343
continue;
}
if (sync_index == SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ) {
const VkExtent3D extent = CastTo3D(cb_state_->active_render_pass_begin_info.renderArea.extent);
const VkOffset3D offset = CastTo3D(cb_state_->active_render_pass_begin_info.renderArea.offset);
current_context_->UpdateAccessState(*img_view_state, sync_index, SyncOrdering::kRaster, offset, extent,
tag);
} else {
current_context_->UpdateAccessState(*img_view_state, sync_index, SyncOrdering::kNonAttachment, tag);
}
break;
}
case DescriptorClass::TexelBuffer: {
const auto *texel_descriptor = static_cast<const TexelDescriptor *>(descriptor);
if (texel_descriptor->Invalid()) {
continue;
}
const auto *buf_view_state = texel_descriptor->GetBufferViewState();
const auto *buf_state = buf_view_state->buffer_state.get();
const ResourceAccessRange range = MakeRange(*buf_view_state);
current_context_->UpdateAccessState(*buf_state, sync_index, SyncOrdering::kNonAttachment, range, tag);
break;
}
case DescriptorClass::GeneralBuffer: {
const auto *buffer_descriptor = static_cast<const BufferDescriptor *>(descriptor);
if (buffer_descriptor->Invalid()) {
continue;
}
const auto *buf_state = buffer_descriptor->GetBufferState();
const ResourceAccessRange range =
MakeRange(*buf_state, buffer_descriptor->GetOffset(), buffer_descriptor->GetRange());
current_context_->UpdateAccessState(*buf_state, sync_index, SyncOrdering::kNonAttachment, range, tag);
break;
}
// TODO: INLINE_UNIFORM_BLOCK_EXT, ACCELERATION_STRUCTURE_KHR
default:
break;
}
}
}
}
}
bool CommandBufferAccessContext::ValidateDrawVertex(const std::optional<uint32_t> &vertexCount, uint32_t firstVertex,
const Location &loc) const {
bool skip = false;
const auto *pipe = cb_state_->GetCurrentPipeline(VK_PIPELINE_BIND_POINT_GRAPHICS);
if (!pipe) {
return skip;
}
const auto &binding_buffers = cb_state_->current_vertex_buffer_binding_info.vertex_buffer_bindings;
const auto &binding_buffers_size = binding_buffers.size();
const auto &binding_descriptions_size = pipe->vertex_input_state->binding_descriptions.size();
for (size_t i = 0; i < binding_descriptions_size; ++i) {
const auto &binding_description = pipe->vertex_input_state->binding_descriptions[i];
if (binding_description.binding < binding_buffers_size) {
const auto &binding_buffer = binding_buffers[binding_description.binding];
if (!binding_buffer.bound()) continue;
auto *buf_state = binding_buffer.buffer_state.get();
const ResourceAccessRange range = MakeRange(binding_buffer, firstVertex, vertexCount, binding_description.stride);
auto hazard = current_context_->DetectHazard(*buf_state, SYNC_VERTEX_ATTRIBUTE_INPUT_VERTEX_ATTRIBUTE_READ, range);
if (hazard.IsHazard()) {
skip |= sync_state_->LogError(string_SyncHazardVUID(hazard.Hazard()), buf_state->buffer(), loc,
"Hazard %s for vertex %s in %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
sync_state_->FormatHandle(buf_state->buffer()).c_str(),
sync_state_->FormatHandle(cb_state_->commandBuffer()).c_str(),
FormatHazard(hazard).c_str());
}
}
}
return skip;
}
void CommandBufferAccessContext::RecordDrawVertex(const std::optional<uint32_t> &vertexCount, uint32_t firstVertex,
const ResourceUsageTag tag) {
const auto *pipe = cb_state_->GetCurrentPipeline(VK_PIPELINE_BIND_POINT_GRAPHICS);
if (!pipe) {
return;
}
const auto &binding_buffers = cb_state_->current_vertex_buffer_binding_info.vertex_buffer_bindings;
const auto &binding_buffers_size = binding_buffers.size();
const auto &binding_descriptions_size = pipe->vertex_input_state->binding_descriptions.size();
for (size_t i = 0; i < binding_descriptions_size; ++i) {
const auto &binding_description = pipe->vertex_input_state->binding_descriptions[i];
if (binding_description.binding < binding_buffers_size) {
const auto &binding_buffer = binding_buffers[binding_description.binding];
if (!binding_buffer.bound()) continue;
auto *buf_state = binding_buffer.buffer_state.get();
const ResourceAccessRange range = MakeRange(binding_buffer, firstVertex, vertexCount, binding_description.stride);
current_context_->UpdateAccessState(*buf_state, SYNC_VERTEX_ATTRIBUTE_INPUT_VERTEX_ATTRIBUTE_READ,
SyncOrdering::kNonAttachment, range, tag);
}
}
}
bool CommandBufferAccessContext::ValidateDrawVertexIndex(const std::optional<uint32_t> &index_count, uint32_t firstIndex,
const Location &loc) const {
bool skip = false;
if (!cb_state_->index_buffer_binding.bound()) {
return skip;
}
const auto &index_binding = cb_state_->index_buffer_binding;
auto *index_buf_state = index_binding.buffer_state.get();
const auto index_size = GetIndexAlignment(index_binding.index_type);
const ResourceAccessRange range = MakeRange(index_binding, firstIndex, index_count, index_size);
auto hazard = current_context_->DetectHazard(*index_buf_state, SYNC_INDEX_INPUT_INDEX_READ, range);
if (hazard.IsHazard()) {
skip |= sync_state_->LogError(string_SyncHazardVUID(hazard.Hazard()), index_buf_state->buffer(), loc,
"Hazard %s for index %s in %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
sync_state_->FormatHandle(index_buf_state->buffer()).c_str(),
sync_state_->FormatHandle(cb_state_->commandBuffer()).c_str(), FormatHazard(hazard).c_str());
}
// TODO: For now, we detect the whole vertex buffer. Index buffer could be changed until SubmitQueue.
// We will detect more accurate range in the future.
skip |= ValidateDrawVertex(std::optional<uint32_t>(), 0, loc);
return skip;
}
void CommandBufferAccessContext::RecordDrawVertexIndex(const std::optional<uint32_t> &indexCount, uint32_t firstIndex,
const ResourceUsageTag tag) {
if (!cb_state_->index_buffer_binding.bound()) return;
const auto &index_binding = cb_state_->index_buffer_binding;
auto *index_buf_state = index_binding.buffer_state.get();
const auto index_size = GetIndexAlignment(index_binding.index_type);
const ResourceAccessRange range = MakeRange(index_binding, firstIndex, indexCount, index_size);
current_context_->UpdateAccessState(*index_buf_state, SYNC_INDEX_INPUT_INDEX_READ, SyncOrdering::kNonAttachment, range, tag);
// TODO: For now, we detect the whole vertex buffer. Index buffer could be changed until SubmitQueue.
// We will detect more accurate range in the future.
RecordDrawVertex(std::optional<uint32_t>(), 0, tag);
}
bool CommandBufferAccessContext::ValidateDrawSubpassAttachment(const Location &loc) const {
bool skip = false;
if (!current_renderpass_context_) return skip;
skip |= current_renderpass_context_->ValidateDrawSubpassAttachment(GetExecutionContext(), *cb_state_, loc.function);
return skip;
}
void CommandBufferAccessContext::RecordDrawSubpassAttachment(const ResourceUsageTag tag) {
if (current_renderpass_context_) {
current_renderpass_context_->RecordDrawSubpassAttachment(*cb_state_, tag);
}
}
QueueId CommandBufferAccessContext::GetQueueId() const { return QueueSyncState::kQueueIdInvalid; }
ResourceUsageTag CommandBufferAccessContext::RecordBeginRenderPass(
vvl::Func command, const RENDER_PASS_STATE &rp_state, const VkRect2D &render_area,
const std::vector<const syncval_state::ImageViewState *> &attachment_views) {
// Create an access context the current renderpass.
const auto barrier_tag = NextCommandTag(command, NamedHandle("renderpass", rp_state.Handle()),
ResourceUsageRecord::SubcommandType::kSubpassTransition);
const auto load_tag = NextSubcommandTag(command, ResourceUsageRecord::SubcommandType::kLoadOp);
render_pass_contexts_.emplace_back(
std::make_unique<RenderPassAccessContext>(rp_state, render_area, GetQueueFlags(), attachment_views, &cb_access_context_));
current_renderpass_context_ = render_pass_contexts_.back().get();
current_renderpass_context_->RecordBeginRenderPass(barrier_tag, load_tag);
current_context_ = &current_renderpass_context_->CurrentContext();
return barrier_tag;
}
ResourceUsageTag CommandBufferAccessContext::RecordNextSubpass(vvl::Func command) {
assert(current_renderpass_context_);
if (!current_renderpass_context_) return NextCommandTag(command);
auto store_tag = NextCommandTag(command, NamedHandle("renderpass", current_renderpass_context_->GetRenderPassState()->Handle()),
ResourceUsageRecord::SubcommandType::kStoreOp);
auto barrier_tag = NextSubcommandTag(command, ResourceUsageRecord::SubcommandType::kSubpassTransition);
auto load_tag = NextSubcommandTag(command, ResourceUsageRecord::SubcommandType::kLoadOp);
current_renderpass_context_->RecordNextSubpass(store_tag, barrier_tag, load_tag);
current_context_ = &current_renderpass_context_->CurrentContext();
return barrier_tag;
}
ResourceUsageTag CommandBufferAccessContext::RecordEndRenderPass(vvl::Func command) {
assert(current_renderpass_context_);
if (!current_renderpass_context_) return NextCommandTag(command);
auto store_tag = NextCommandTag(command, NamedHandle("renderpass", current_renderpass_context_->GetRenderPassState()->Handle()),
ResourceUsageRecord::SubcommandType::kStoreOp);
auto barrier_tag = NextSubcommandTag(command, ResourceUsageRecord::SubcommandType::kSubpassTransition);
current_renderpass_context_->RecordEndRenderPass(&cb_access_context_, store_tag, barrier_tag);
current_context_ = &cb_access_context_;
current_renderpass_context_ = nullptr;
return barrier_tag;
}
void CommandBufferAccessContext::RecordDestroyEvent(EVENT_STATE *event_state) { GetCurrentEventsContext()->Destroy(event_state); }
bool ReplayState::DetectFirstUseHazard(const ResourceUsageRange &first_use_range) const {
bool skip = false;
if (first_use_range.non_empty()) {
HazardResult hazard;
// We're allowing for the Replay(Validate|Record) to modify the exec_context (e.g. for Renderpass operations), so
// we need to fetch the current access context each time
hazard = GetRecordedAccessContext()->DetectFirstUseHazard(exec_context_.GetQueueId(), first_use_range,
*exec_context_.GetCurrentAccessContext());
if (hazard.IsHazard()) {
const SyncValidator &sync_state = exec_context_.GetSyncState();
const auto handle = exec_context_.Handle();
const auto recorded_handle = recorded_context_.GetCBState().commandBuffer();
skip = sync_state.LogError(string_SyncHazardVUID(hazard.Hazard()), handle, error_obj_.location,
"Hazard %s for entry %" PRIu32 ", %s, Recorded access info %s. Access info %s.",
string_SyncHazard(hazard.Hazard()), index_, sync_state.FormatHandle(recorded_handle).c_str(),
recorded_context_.FormatUsage(*hazard.RecordedAccess()).c_str(),
recorded_context_.FormatHazard(hazard).c_str());
}
}
return skip;
}
bool ReplayState::ValidateFirstUse() {
if (!exec_context_.ValidForSyncOps()) return false;
bool skip = false;
ResourceUsageRange first_use_range = {0, 0};
for (const auto &sync_op : recorded_context_.GetSyncOps()) {
// Set the range to cover all accesses until the next sync_op, and validate
first_use_range.end = sync_op.tag;
skip |= DetectFirstUseHazard(first_use_range);
// Call to replay validate support for syncop with non-trivial replay
skip |= sync_op.sync_op->ReplayValidate(*this, sync_op.tag);
// Record the barrier into the proxy context.
sync_op.sync_op->ReplayRecord(exec_context_, base_tag_ + sync_op.tag);
first_use_range.begin = sync_op.tag + 1;
}
// and anything after the last syncop
first_use_range.end = ResourceUsageRecord::kMaxIndex;
skip |= DetectFirstUseHazard(first_use_range);
return skip;
}
void CommandBufferAccessContext::RecordExecutedCommandBuffer(const CommandBufferAccessContext &recorded_cb_context) {
const AccessContext *recorded_context = recorded_cb_context.GetCurrentAccessContext();
assert(recorded_context);
// Just run through the barriers ignoring the usage from the recorded context, as Resolve will overwrite outdated state
const ResourceUsageTag base_tag = GetTagLimit();
for (const auto &sync_op : recorded_cb_context.GetSyncOps()) {
// we update the range to any include layout transition first use writes,
// as they are stored along with the source scope (as effective barrier) when recorded
sync_op.sync_op->ReplayRecord(*this, base_tag + sync_op.tag);
}
ResourceUsageRange tag_range = ImportRecordedAccessLog(recorded_cb_context);
assert(base_tag == tag_range.begin); // to ensure the to offset calculation agree
ResolveExecutedCommandBuffer(*recorded_context, tag_range.begin);
}
void CommandBufferAccessContext::ResolveExecutedCommandBuffer(const AccessContext &recorded_context, ResourceUsageTag offset) {
auto tag_offset = [offset](ResourceAccessState *access) { access->OffsetTag(offset); };
GetCurrentAccessContext()->ResolveFromContext(tag_offset, recorded_context);
}
ResourceUsageRange CommandExecutionContext::ImportRecordedAccessLog(const CommandBufferAccessContext &recorded_context) {
// The execution references ensure lifespan for the referenced child CB's...
ResourceUsageRange tag_range(GetTagLimit(), 0);
InsertRecordedAccessLogEntries(recorded_context);
tag_range.end = GetTagLimit();
return tag_range;
}
void CommandBufferAccessContext::InsertRecordedAccessLogEntries(const CommandBufferAccessContext &recorded_context) {
cbs_referenced_->emplace(recorded_context.GetCBStateShared());
access_log_->insert(access_log_->end(), recorded_context.access_log_->cbegin(), recorded_context.access_log_->cend());
}
ResourceUsageTag CommandBufferAccessContext::NextSubcommandTag(vvl::Func command, ResourceUsageRecord::SubcommandType subcommand) {
return NextSubcommandTag(command, NamedHandle(), subcommand);
}
ResourceUsageTag CommandBufferAccessContext::NextSubcommandTag(vvl::Func command, NamedHandle &&handle,
ResourceUsageRecord::SubcommandType subcommand) {
ResourceUsageTag next = access_log_->size();
access_log_->emplace_back(command, command_number_, subcommand, ++subcommand_number_, cb_state_, reset_count_);
if (command_handles_.size()) {
// This is a duplication, but it keeps tags->log information flat (i.e not depending on some "command tag" entry
access_log_->back().handles = command_handles_;
}
if (handle) {
access_log_->back().AddHandle(std::move(handle));
}
return next;
}
ResourceUsageTag CommandBufferAccessContext::NextCommandTag(vvl::Func command, ResourceUsageRecord::SubcommandType subcommand) {
return NextCommandTag(command, NamedHandle(), subcommand);
}
ResourceUsageTag CommandBufferAccessContext::NextCommandTag(vvl::Func command, NamedHandle &&handle,
ResourceUsageRecord::SubcommandType subcommand) {
command_number_++;
command_handles_.clear();
subcommand_number_ = 0;
ResourceUsageTag next = access_log_->size();
access_log_->emplace_back(command, command_number_, subcommand, subcommand_number_, cb_state_, reset_count_);
if (handle) {
access_log_->back().AddHandle(handle);
command_handles_.emplace_back(std::move(handle));
}
return next;
}
ResourceUsageTag CommandBufferAccessContext::NextIndexedCommandTag(vvl::Func command, uint32_t index) {
if (index == 0) {
return NextCommandTag(command, ResourceUsageRecord::SubcommandType::kIndex);
}
return NextSubcommandTag(command, ResourceUsageRecord::SubcommandType::kIndex);
}
void CommandBufferAccessContext::RecordSyncOp(SyncOpPointer &&sync_op) {
auto tag = sync_op->Record(this);
// As renderpass operations can have side effects on the command buffer access context,
// update the sync operation to record these if any.
sync_ops_.emplace_back(tag, std::move(sync_op));
}
class HazardDetectFirstUse {
public:
HazardDetectFirstUse(const ResourceAccessState &recorded_use, QueueId queue_id, const ResourceUsageRange &tag_range)
: recorded_use_(recorded_use), queue_id_(queue_id), tag_range_(tag_range) {}
HazardResult Detect(const ResourceAccessRangeMap::const_iterator &pos) const {
return pos->second.DetectHazard(recorded_use_, queue_id_, tag_range_);
}
HazardResult DetectAsync(const ResourceAccessRangeMap::const_iterator &pos, ResourceUsageTag start_tag) const {
return pos->second.DetectAsyncHazard(recorded_use_, tag_range_, start_tag);
}
private:
const ResourceAccessState &recorded_use_;
const QueueId queue_id_;
const ResourceUsageRange &tag_range_;
};
// This is called with the *recorded* command buffers access context, with the *active* access context pass in, againsts which
// hazards will be detected
HazardResult AccessContext::DetectFirstUseHazard(QueueId queue_id, const ResourceUsageRange &tag_range,
const AccessContext &access_context) const {
HazardResult hazard;
for (const auto &recorded_access : access_state_map_) {
// Cull any entries not in the current tag range
if (!recorded_access.second.FirstAccessInTagRange(tag_range)) continue;
HazardDetectFirstUse detector(recorded_access.second, queue_id, tag_range);
hazard = access_context.DetectHazard(detector, recorded_access.first, DetectOptions::kDetectAll);
if (hazard.IsHazard()) break;
}
return hazard;
}
bool RenderPassAccessContext::ValidateDrawSubpassAttachment(const CommandExecutionContext &exec_context,
const CMD_BUFFER_STATE &cmd_buffer, vvl::Func command) const {
bool skip = false;
const auto &sync_state = exec_context.GetSyncState();
const auto lv_bind_point = ConvertToLvlBindPoint(VK_PIPELINE_BIND_POINT_GRAPHICS);
const auto last_bound_state = cmd_buffer.lastBound[lv_bind_point];
const auto *pipe = last_bound_state.pipeline_state;
if (!pipe) {
return skip;
}
const auto raster_state = pipe->RasterizationState();
if (raster_state && raster_state->rasterizerDiscardEnable) {
return skip;
}
const char *caller_name = vvl::String(command);
const auto &list = pipe->fragmentShader_writable_output_location_list;
const auto &subpass = rp_state_->createInfo.pSubpasses[current_subpass_];
const auto &current_context = CurrentContext();
// Subpass's inputAttachment has been done in ValidateDispatchDrawDescriptorSet
if (subpass.pColorAttachments && subpass.colorAttachmentCount && !list.empty()) {
for (const auto location : list) {
if (location >= subpass.colorAttachmentCount ||
subpass.pColorAttachments[location].attachment == VK_ATTACHMENT_UNUSED) {
continue;
}
const AttachmentViewGen &view_gen = attachment_views_[subpass.pColorAttachments[location].attachment];
if (!view_gen.IsValid()) continue;
HazardResult hazard =
current_context.DetectHazard(view_gen, AttachmentViewGen::Gen::kRenderArea,
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE, SyncOrdering::kColorAttachment);
if (hazard.IsHazard()) {
const VkImageView view_handle = view_gen.GetViewState()->image_view();
skip |=
sync_state.LogError(view_handle, string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s for %s in %s, Subpass #%d, and pColorAttachments #%d. Access info %s.",
caller_name, string_SyncHazard(hazard.Hazard()),
sync_state.FormatHandle(view_handle).c_str(), sync_state.FormatHandle(cmd_buffer).c_str(),
cmd_buffer.GetActiveSubpass(), location, exec_context.FormatHazard(hazard).c_str());
}
}
}
// PHASE1 TODO: Add layout based read/vs. write selection.
// PHASE1 TODO: Read operations for both depth and stencil are possible in the future.
const auto ds_state = pipe->DepthStencilState();
const uint32_t depth_stencil_attachment = GetSubpassDepthStencilAttachmentIndex(ds_state, subpass.pDepthStencilAttachment);
if ((depth_stencil_attachment != VK_ATTACHMENT_UNUSED) && attachment_views_[depth_stencil_attachment].IsValid()) {
const AttachmentViewGen &view_gen = attachment_views_[depth_stencil_attachment];
const IMAGE_VIEW_STATE &view_state = *view_gen.GetViewState();
bool depth_write = false, stencil_write = false;
const bool depth_write_enable = last_bound_state.IsDepthWriteEnable(); // implicitly means DepthTestEnable is set
const bool stencil_test_enable = last_bound_state.IsStencilTestEnable();
// PHASE1 TODO: These validation should be in core_checks.
if (!vkuFormatIsStencilOnly(view_state.create_info.format) && depth_write_enable &&
IsImageLayoutDepthWritable(subpass.pDepthStencilAttachment->layout)) {
depth_write = true;
}
// PHASE1 TODO: It needs to check if stencil is writable.
// If failOp, passOp, or depthFailOp are not KEEP, and writeMask isn't 0, it's writable.
// If depth test is disable, it's considered depth test passes, and then depthFailOp doesn't run.
// PHASE1 TODO: These validation should be in core_checks.
if (!vkuFormatIsDepthOnly(view_state.create_info.format) && stencil_test_enable &&
IsImageLayoutStencilWritable(subpass.pDepthStencilAttachment->layout)) {
stencil_write = true;
}
// PHASE1 TODO: Add EARLY stage detection based on ExecutionMode.
if (depth_write) {
HazardResult hazard = current_context.DetectHazard(view_gen, AttachmentViewGen::Gen::kDepthOnlyRenderArea,
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE,
SyncOrdering::kDepthStencilAttachment);
if (hazard.IsHazard()) {
skip |= sync_state.LogError(
view_state.image_view(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s for %s in %s, Subpass #%d, and depth part of pDepthStencilAttachment. Access info %s.",
caller_name, string_SyncHazard(hazard.Hazard()), sync_state.FormatHandle(view_state).c_str(),
sync_state.FormatHandle(cmd_buffer).c_str(), cmd_buffer.GetActiveSubpass(),
exec_context.FormatHazard(hazard).c_str());
}
}
if (stencil_write) {
HazardResult hazard = current_context.DetectHazard(view_gen, AttachmentViewGen::Gen::kStencilOnlyRenderArea,
SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE,
SyncOrdering::kDepthStencilAttachment);
if (hazard.IsHazard()) {
skip |= sync_state.LogError(
view_state.image_view(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s for %s in %s, Subpass #%d, and stencil part of pDepthStencilAttachment. Access info %s.",
caller_name, string_SyncHazard(hazard.Hazard()), sync_state.FormatHandle(view_state).c_str(),
sync_state.FormatHandle(cmd_buffer).c_str(), cmd_buffer.GetActiveSubpass(),
exec_context.FormatHazard(hazard).c_str());
}
}
}
return skip;
}
void RenderPassAccessContext::RecordDrawSubpassAttachment(const CMD_BUFFER_STATE &cmd_buffer, const ResourceUsageTag tag) {
const auto lv_bind_point = ConvertToLvlBindPoint(VK_PIPELINE_BIND_POINT_GRAPHICS);
const auto last_bound_state = cmd_buffer.lastBound[lv_bind_point];
const auto *pipe = last_bound_state.pipeline_state;
if (!pipe) {
return;
}
const auto *raster_state = pipe->RasterizationState();
if (raster_state && raster_state->rasterizerDiscardEnable) {
return;
}
const auto &list = pipe->fragmentShader_writable_output_location_list;
const auto &subpass = rp_state_->createInfo.pSubpasses[current_subpass_];
auto &current_context = CurrentContext();
// Subpass's inputAttachment has been done in RecordDispatchDrawDescriptorSet
if (subpass.pColorAttachments && subpass.colorAttachmentCount && !list.empty()) {
for (const auto location : list) {
if (location >= subpass.colorAttachmentCount ||
subpass.pColorAttachments[location].attachment == VK_ATTACHMENT_UNUSED) {
continue;
}
const AttachmentViewGen &view_gen = attachment_views_[subpass.pColorAttachments[location].attachment];
current_context.UpdateAccessState(view_gen, AttachmentViewGen::Gen::kRenderArea,
SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE, SyncOrdering::kColorAttachment,
tag);
}
}
// PHASE1 TODO: Add layout based read/vs. write selection.
// PHASE1 TODO: Read operations for both depth and stencil are possible in the future.
const auto *ds_state = pipe->DepthStencilState();
const uint32_t depth_stencil_attachment = GetSubpassDepthStencilAttachmentIndex(ds_state, subpass.pDepthStencilAttachment);
if ((depth_stencil_attachment != VK_ATTACHMENT_UNUSED) && attachment_views_[depth_stencil_attachment].IsValid()) {
const AttachmentViewGen &view_gen = attachment_views_[depth_stencil_attachment];
const IMAGE_VIEW_STATE &view_state = *view_gen.GetViewState();
bool depth_write = false, stencil_write = false;
const bool has_depth = 0 != (view_state.normalized_subresource_range.aspectMask & VK_IMAGE_ASPECT_DEPTH_BIT);
const bool has_stencil = 0 != (view_state.normalized_subresource_range.aspectMask & VK_IMAGE_ASPECT_STENCIL_BIT);
const bool depth_write_enable = last_bound_state.IsDepthWriteEnable(); // implicitly means DepthTestEnable is set
const bool stencil_test_enable = last_bound_state.IsStencilTestEnable();
// PHASE1 TODO: These validation should be in core_checks.
if (has_depth && !vkuFormatIsStencilOnly(view_state.create_info.format) && depth_write_enable &&
IsImageLayoutDepthWritable(subpass.pDepthStencilAttachment->layout)) {
depth_write = true;
}
// PHASE1 TODO: It needs to check if stencil is writable.
// If failOp, passOp, or depthFailOp are not KEEP, and writeMask isn't 0, it's writable.
// If depth test is disable, it's considered depth test passes, and then depthFailOp doesn't run.
// PHASE1 TODO: These validation should be in core_checks.
if (has_stencil && !vkuFormatIsDepthOnly(view_state.create_info.format) && stencil_test_enable &&
IsImageLayoutStencilWritable(subpass.pDepthStencilAttachment->layout)) {
stencil_write = true;
}
if (depth_write || stencil_write) {
const auto ds_gentype = view_gen.GetDepthStencilRenderAreaGenType(depth_write, stencil_write);
// PHASE1 TODO: Add EARLY stage detection based on ExecutionMode.
current_context.UpdateAccessState(view_gen, ds_gentype, SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE,
SyncOrdering::kDepthStencilAttachment, tag);
}
}
}
static constexpr VkImageAspectFlags kColorAspects =
VK_IMAGE_ASPECT_COLOR_BIT | VK_IMAGE_ASPECT_PLANE_0_BIT | VK_IMAGE_ASPECT_PLANE_1_BIT | VK_IMAGE_ASPECT_PLANE_2_BIT;
static constexpr VkImageAspectFlags kDepthStencilAspects = VK_IMAGE_ASPECT_DEPTH_BIT | VK_IMAGE_ASPECT_STENCIL_BIT;
static std::optional<uint32_t> GetAttachmentIndex(const RenderPassAccessContext &rp_context,
const VkClearAttachment &clear_attachment) {
const auto &rpci = rp_context.GetRenderPassState()->createInfo;
const auto &subpass = rpci.pSubpasses[rp_context.GetCurrentSubpass()];
uint32_t attachment_index = VK_ATTACHMENT_UNUSED;
if (clear_attachment.aspectMask & VK_IMAGE_ASPECT_COLOR_BIT) {
if (clear_attachment.colorAttachment < subpass.colorAttachmentCount) {
attachment_index = subpass.pColorAttachments[clear_attachment.colorAttachment].attachment;
}
} else if (clear_attachment.aspectMask & kDepthStencilAspects) {
if (subpass.pDepthStencilAttachment) {
attachment_index = subpass.pDepthStencilAttachment->attachment;
}
}
const bool invalid_index = (attachment_index == VK_ATTACHMENT_UNUSED || attachment_index >= rpci.attachmentCount);
return invalid_index ? std::optional<uint32_t>() : attachment_index;
}
static VkImageAspectFlags GetAspectsToClear(VkImageAspectFlags clear_aspect_mask, VkImageAspectFlags view_aspect_mask) {
// Check if clear request is valid.
const bool clear_color = (clear_aspect_mask & VK_IMAGE_ASPECT_COLOR_BIT) != 0;
const bool clear_depth = (clear_aspect_mask & VK_IMAGE_ASPECT_DEPTH_BIT) != 0;
const bool clear_stencil = (clear_aspect_mask & VK_IMAGE_ASPECT_STENCIL_BIT) != 0;
if (!clear_color && !clear_depth && !clear_stencil) {
return 0; // nothing to clear
}
if (clear_color && (clear_depth || clear_stencil)) {
return 0; // according to spec it's not allowed
}
// Collect aspects that should be cleared.
VkImageAspectFlags aspects_to_clear = VK_IMAGE_ASPECT_NONE;
if (clear_color && (view_aspect_mask & kColorAspects) != 0) {
assert(GetBitSetCount(view_aspect_mask) == 1);
aspects_to_clear |= view_aspect_mask;
}
if (clear_depth && (view_aspect_mask & VK_IMAGE_ASPECT_DEPTH_BIT) != 0) {
aspects_to_clear |= VK_IMAGE_ASPECT_DEPTH_BIT;
}
if (clear_stencil && (view_aspect_mask & VK_IMAGE_ASPECT_STENCIL_BIT) != 0) {
aspects_to_clear |= VK_IMAGE_ASPECT_STENCIL_BIT;
}
return aspects_to_clear;
}
static std::optional<VkImageSubresourceRange> RestrictSubresourceRange(const VkImageSubresourceRange &normalized_subresource_range,
const VkClearRect &clear_rect) {
assert(normalized_subresource_range.layerCount != VK_REMAINING_ARRAY_LAYERS); // contract of this function
assert(clear_rect.layerCount != VK_REMAINING_ARRAY_LAYERS); // according to spec
const uint32_t first = std::max(normalized_subresource_range.baseArrayLayer, clear_rect.baseArrayLayer);
const uint32_t last_range = normalized_subresource_range.baseArrayLayer + normalized_subresource_range.layerCount;
const uint32_t last_clear = clear_rect.baseArrayLayer + clear_rect.layerCount;
const uint32_t last = std::min(last_range, last_clear);
std::optional<VkImageSubresourceRange> result;
if (first < last) {
result = normalized_subresource_range;
result->baseArrayLayer = first;
result->layerCount = last - first;
}
return result;
}
std::optional<RenderPassAccessContext::ClearAttachmentInfo> RenderPassAccessContext::GetClearAttachmentInfo(
const VkClearAttachment &clear_attachment, const VkClearRect &rect) const {
const auto attachment_index = GetAttachmentIndex(*this, clear_attachment);
if (!attachment_index) {
return {};
}
const auto &view_subresource_range = attachment_views_[attachment_index.value()].GetViewState()->normalized_subresource_range;
const auto aspects = GetAspectsToClear(clear_attachment.aspectMask, view_subresource_range.aspectMask);
if (!aspects) {
return {};
}
const auto subresource_range = RestrictSubresourceRange(view_subresource_range, rect);
if (!subresource_range) {
return {};
}
return ClearAttachmentInfo{attachment_index.value(), aspects, subresource_range.value()};
}
bool RenderPassAccessContext::ValidateClearAttachment(const CommandExecutionContext &exec_context,
const CMD_BUFFER_STATE &cmd_buffer, const Location &loc,
const VkClearAttachment &clear_attachment, const VkClearRect &rect,
uint32_t rect_index) const {
const auto info = GetClearAttachmentInfo(clear_attachment, rect);
if (!info) {
return false;
}
const auto &view_state = *attachment_views_[info->attachment_index].GetViewState();
const VkOffset3D offset = CastTo3D(rect.rect.offset);
const VkExtent3D extent = CastTo3D(rect.rect.extent);
auto subresource_range = info->subresource_range;
bool skip = false;
if (info->aspects_to_clear & kColorAspects) {
assert(GetBitSetCount(info->aspects_to_clear) == 1);
subresource_range.aspectMask = info->aspects_to_clear;
HazardResult hazard = CurrentContext().DetectHazard(
*view_state.GetImageState(), SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE, subresource_range,
SyncOrdering::kColorAttachment, offset, extent, view_state.IsDepthSliced());
if (hazard.IsHazard()) {
const LogObjectList objlist(cmd_buffer.commandBuffer(), view_state.image_view());
skip |= exec_context.GetSyncState().LogError(string_SyncHazardVUID(hazard.Hazard()), objlist, loc,
"Hazard %s when clearing pRects[%" PRIu32
"] region of color attachment %" PRIu32 " in subpass %" PRIu32
". Access info %s.",
string_SyncHazard(hazard.Hazard()), rect_index, info->attachment_index,
cmd_buffer.GetActiveSubpass(), exec_context.FormatHazard(hazard).c_str());
}
}
constexpr VkImageAspectFlagBits depth_stencil_aspects[2] = {VK_IMAGE_ASPECT_DEPTH_BIT, VK_IMAGE_ASPECT_STENCIL_BIT};
for (const auto aspect : depth_stencil_aspects) {
if (info->aspects_to_clear & aspect) {
// Original aspect mask can contain both stencil and depth but here we track each aspect separately
subresource_range.aspectMask = aspect;
// vkCmdClearAttachments depth/stencil writes are executed by the EARLY_FRAGMENT_TESTS_BIT and LATE_FRAGMENT_TESTS_BIT
// stages. The implementation tracks the most recent access, which happens in the LATE_FRAGMENT_TESTS_BIT stage.
HazardResult hazard = CurrentContext().DetectHazard(
*view_state.GetImageState(), SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE, subresource_range,
SyncOrdering::kDepthStencilAttachment, offset, extent, view_state.IsDepthSliced());
if (hazard.IsHazard()) {
const LogObjectList objlist(cmd_buffer.commandBuffer(), view_state.image_view());
skip |= exec_context.GetSyncState().LogError(
string_SyncHazardVUID(hazard.Hazard()), objlist, loc,
"Hazard %s when clearing pRects[%" PRIu32 "] region of %s aspect of depth-stencil attachment %" PRIu32
" in subpass %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), rect_index, string_VkImageAspectFlagBits(aspect), info->attachment_index,
cmd_buffer.GetActiveSubpass(), exec_context.FormatHazard(hazard).c_str());
}
}
}
return skip;
}
void RenderPassAccessContext::RecordClearAttachment(const CMD_BUFFER_STATE &cmd_buffer, ResourceUsageTag tag,
const VkClearAttachment &clear_attachment, const VkClearRect &rect) {
const auto info = GetClearAttachmentInfo(clear_attachment, rect);
if (!info) {
return;
}
const auto &view_state = *attachment_views_[info->attachment_index].GetViewState();
const VkOffset3D offset = CastTo3D(rect.rect.offset);
const VkExtent3D extent = CastTo3D(rect.rect.extent);
auto subresource_range = info->subresource_range;
// Original subresource range can include aspects that are not cleared, they should not be tracked
subresource_range.aspectMask = info->aspects_to_clear;
if (info->aspects_to_clear & kColorAspects) {
assert((info->aspects_to_clear & kDepthStencilAspects) == 0);
CurrentContext().UpdateAccessState(*view_state.GetImageState(), SYNC_COLOR_ATTACHMENT_OUTPUT_COLOR_ATTACHMENT_WRITE,
SyncOrdering::kColorAttachment, subresource_range, offset, extent, tag);
} else {
assert((info->aspects_to_clear & kColorAspects) == 0);
CurrentContext().UpdateAccessState(*view_state.GetImageState(), SYNC_LATE_FRAGMENT_TESTS_DEPTH_STENCIL_ATTACHMENT_WRITE,
SyncOrdering::kDepthStencilAttachment, subresource_range, offset, extent, tag);
}
}
bool RenderPassAccessContext::ValidateNextSubpass(const CommandExecutionContext &exec_context, vvl::Func command) const {
// PHASE1 TODO: Add Validate Preserve attachments
bool skip = false;
skip |= CurrentContext().ValidateResolveOperations(exec_context, *rp_state_, render_area_, attachment_views_, command,
current_subpass_);
skip |= CurrentContext().ValidateStoreOperation(exec_context, *rp_state_, render_area_, current_subpass_, attachment_views_,
command);
const auto next_subpass = current_subpass_ + 1;
if (next_subpass >= subpass_contexts_.size()) {
return skip;
}
const auto &next_context = subpass_contexts_[next_subpass];
skip |=
next_context.ValidateLayoutTransitions(exec_context, *rp_state_, render_area_, next_subpass, attachment_views_, command);
if (!skip) {
// To avoid complex (and buggy) duplication of the affect of layout transitions on load operations, we'll record them
// on a copy of the (empty) next context.
// Note: The resource access map should be empty so hopefully this copy isn't too horrible from a perf POV.
AccessContext temp_context(next_context);
temp_context.RecordLayoutTransitions(*rp_state_, next_subpass, attachment_views_, kInvalidTag);
skip |=
temp_context.ValidateLoadOperation(exec_context, *rp_state_, render_area_, next_subpass, attachment_views_, command);
}
return skip;
}
bool RenderPassAccessContext::ValidateEndRenderPass(const CommandExecutionContext &exec_context, vvl::Func command) const {
// PHASE1 TODO: Validate Preserve
bool skip = false;
skip |= CurrentContext().ValidateResolveOperations(exec_context, *rp_state_, render_area_, attachment_views_, command,
current_subpass_);
skip |= CurrentContext().ValidateStoreOperation(exec_context, *rp_state_, render_area_, current_subpass_, attachment_views_,
command);
skip |= ValidateFinalSubpassLayoutTransitions(exec_context, command);
return skip;
}
AccessContext *RenderPassAccessContext::CreateStoreResolveProxy() const {
return CreateStoreResolveProxyContext(CurrentContext(), *rp_state_, current_subpass_, attachment_views_);
}
bool RenderPassAccessContext::ValidateFinalSubpassLayoutTransitions(const CommandExecutionContext &exec_context,
vvl::Func command) const {
bool skip = false;
// As validation methods are const and precede the record/update phase, for any tranistions from the current (last)
// subpass, we have to validate them against a copy of the current AccessContext, with resolve operations applied.
// Note: we could be more efficient by tracking whether or not we actually *have* any changes (e.g. attachment resolve)
// to apply and only copy then, if this proves a hot spot.
std::unique_ptr<AccessContext> proxy_for_current;
// Validate the "finalLayout" transitions to external
// Get them from where there we're hidding in the extra entry.
const auto &final_transitions = rp_state_->subpass_transitions.back();
for (const auto &transition : final_transitions) {
const auto &view_gen = attachment_views_[transition.attachment];
const auto &trackback = subpass_contexts_[transition.prev_pass].GetDstExternalTrackBack();
assert(trackback.source_subpass); // Transitions are given implicit transitions if the StateTracker is working correctly
auto *context = trackback.source_subpass;
if (transition.prev_pass == current_subpass_) {
if (!proxy_for_current) {
// We haven't recorded resolve ofor the current_subpass, so we need to copy current and update it *as if*
proxy_for_current.reset(CreateStoreResolveProxy());
}
context = proxy_for_current.get();
}
// Use the merged barrier for the hazard check (safe since it just considers the src (first) scope.
const auto merged_barrier = MergeBarriers(trackback.barriers);
auto hazard = context->DetectImageBarrierHazard(view_gen, merged_barrier, AccessContext::DetectOptions::kDetectPrevious);
if (hazard.IsHazard()) {
if (hazard.Tag() == kInvalidTag) {
// Hazard vs. ILT
skip |= exec_context.GetSyncState().LogError(
rp_state_->renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s vs. store/resolve operations in subpass %" PRIu32 " for attachment %" PRIu32
" final image layout transition (old_layout: %s, new_layout: %s).",
vvl::String(command), string_SyncHazard(hazard.Hazard()), transition.prev_pass, transition.attachment,
string_VkImageLayout(transition.old_layout), string_VkImageLayout(transition.new_layout));
} else {
skip |= exec_context.GetSyncState().LogError(
rp_state_->renderPass(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s with last use subpass %" PRIu32 " for attachment %" PRIu32
" final image layout transition (old_layout: %s, new_layout: %s). Access info %s.",
vvl::String(command), string_SyncHazard(hazard.Hazard()), transition.prev_pass, transition.attachment,
string_VkImageLayout(transition.old_layout), string_VkImageLayout(transition.new_layout),
exec_context.FormatHazard(hazard).c_str());
}
}
}
return skip;
}
void RenderPassAccessContext::RecordLayoutTransitions(const ResourceUsageTag tag) {
// Add layout transitions...
subpass_contexts_[current_subpass_].RecordLayoutTransitions(*rp_state_, current_subpass_, attachment_views_, tag);
}
void RenderPassAccessContext::RecordLoadOperations(const ResourceUsageTag tag) {
const auto *attachment_ci = rp_state_->createInfo.pAttachments;
auto &subpass_context = subpass_contexts_[current_subpass_];
for (uint32_t i = 0; i < rp_state_->createInfo.attachmentCount; i++) {
if (rp_state_->attachment_first_subpass[i] == current_subpass_) {
const AttachmentViewGen &view_gen = attachment_views_[i];
if (!view_gen.IsValid()) continue; // UNUSED
const auto &ci = attachment_ci[i];
const bool has_depth = vkuFormatHasDepth(ci.format);
const bool has_stencil = vkuFormatHasStencil(ci.format);
const bool is_color = !(has_depth || has_stencil);
if (is_color) {
const SyncStageAccessIndex load_op = ColorLoadUsage(ci.loadOp);
if (load_op != SYNC_ACCESS_INDEX_NONE) {
subpass_context.UpdateAccessState(view_gen, AttachmentViewGen::Gen::kRenderArea, load_op,
SyncOrdering::kColorAttachment, tag);
}
} else {
if (has_depth) {
const SyncStageAccessIndex load_op = DepthStencilLoadUsage(ci.loadOp);
if (load_op != SYNC_ACCESS_INDEX_NONE) {
subpass_context.UpdateAccessState(view_gen, AttachmentViewGen::Gen::kDepthOnlyRenderArea, load_op,
SyncOrdering::kDepthStencilAttachment, tag);
}
}
if (has_stencil) {
const SyncStageAccessIndex load_op = DepthStencilLoadUsage(ci.stencilLoadOp);
if (load_op != SYNC_ACCESS_INDEX_NONE) {
subpass_context.UpdateAccessState(view_gen, AttachmentViewGen::Gen::kStencilOnlyRenderArea, load_op,
SyncOrdering::kDepthStencilAttachment, tag);
}
}
}
}
}
}
AttachmentViewGenVector RenderPassAccessContext::CreateAttachmentViewGen(
const VkRect2D &render_area, const std::vector<const syncval_state::ImageViewState *> &attachment_views) {
AttachmentViewGenVector view_gens;
VkExtent3D extent = CastTo3D(render_area.extent);
VkOffset3D offset = CastTo3D(render_area.offset);
view_gens.reserve(attachment_views.size());
for (const auto *view : attachment_views) {
view_gens.emplace_back(view, offset, extent);
}
return view_gens;
}
RenderPassAccessContext::RenderPassAccessContext(const RENDER_PASS_STATE &rp_state, const VkRect2D &render_area,
VkQueueFlags queue_flags,
const std::vector<const syncval_state::ImageViewState *> &attachment_views,
const AccessContext *external_context)
: rp_state_(&rp_state), render_area_(render_area), current_subpass_(0U), attachment_views_() {
// Add this for all subpasses here so that they exist during next subpass validation
InitSubpassContexts(queue_flags, rp_state, external_context, subpass_contexts_);
attachment_views_ = CreateAttachmentViewGen(render_area, attachment_views);
}
void RenderPassAccessContext::RecordBeginRenderPass(const ResourceUsageTag barrier_tag, const ResourceUsageTag load_tag) {
assert(0 == current_subpass_);
AccessContext &current_context = subpass_contexts_[current_subpass_];
current_context.SetStartTag(barrier_tag);
RecordLayoutTransitions(barrier_tag);
RecordLoadOperations(load_tag);
}
void RenderPassAccessContext::RecordNextSubpass(const ResourceUsageTag store_tag, const ResourceUsageTag barrier_tag,
const ResourceUsageTag load_tag) {
// Resolves are against *prior* subpass context and thus *before* the subpass increment
CurrentContext().UpdateAttachmentResolveAccess(*rp_state_, attachment_views_, current_subpass_, store_tag);
CurrentContext().UpdateAttachmentStoreAccess(*rp_state_, attachment_views_, current_subpass_, store_tag);
if (current_subpass_ + 1 >= subpass_contexts_.size()) {
return;
}
// Move to the next sub-command for the new subpass. The resolve and store are logically part of the previous
// subpass, so their tag needs to be different from the layout and load operations below.
current_subpass_++;
AccessContext &current_context = subpass_contexts_[current_subpass_];
current_context.SetStartTag(barrier_tag);
RecordLayoutTransitions(barrier_tag);
RecordLoadOperations(load_tag);
}
void RenderPassAccessContext::RecordEndRenderPass(AccessContext *external_context, const ResourceUsageTag store_tag,
const ResourceUsageTag barrier_tag) {
// Add the resolve and store accesses
CurrentContext().UpdateAttachmentResolveAccess(*rp_state_, attachment_views_, current_subpass_, store_tag);
CurrentContext().UpdateAttachmentStoreAccess(*rp_state_, attachment_views_, current_subpass_, store_tag);
// Export the accesses from the renderpass...
external_context->ResolveChildContexts(subpass_contexts_);
// Add the "finalLayout" transitions to external
// Get them from where there we're hidding in the extra entry.
// Not that since *final* always comes from *one* subpass per view, we don't have to accumulate the barriers
// TODO Aliasing we may need to reconsider barrier accumulation... though I don't know that it would be valid for aliasing
// that had mulitple final layout transistions from mulitple final subpasses.
const auto &final_transitions = rp_state_->subpass_transitions.back();
for (const auto &transition : final_transitions) {
const AttachmentViewGen &view_gen = attachment_views_[transition.attachment];
const auto &last_trackback = subpass_contexts_[transition.prev_pass].GetDstExternalTrackBack();
assert(&subpass_contexts_[transition.prev_pass] == last_trackback.source_subpass);
ApplyBarrierOpsFunctor<PipelineBarrierOp> barrier_action(true /* resolve */, last_trackback.barriers.size(), barrier_tag);
for (const auto &barrier : last_trackback.barriers) {
barrier_action.EmplaceBack(PipelineBarrierOp(QueueSyncState::kQueueIdInvalid, barrier, true));
}
external_context->ApplyUpdateAction(view_gen, AttachmentViewGen::Gen::kViewSubresource, barrier_action);
}
}
SyncExecScope SyncExecScope::MakeSrc(VkQueueFlags queue_flags, VkPipelineStageFlags2KHR mask_param,
const VkPipelineStageFlags2KHR disabled_feature_mask) {
SyncExecScope result;
result.mask_param = mask_param;
result.expanded_mask = sync_utils::ExpandPipelineStages(mask_param, queue_flags, disabled_feature_mask);
result.exec_scope = sync_utils::WithEarlierPipelineStages(result.expanded_mask);
result.valid_accesses = SyncStageAccess::AccessScopeByStage(result.expanded_mask);
return result;
}
SyncExecScope SyncExecScope::MakeDst(VkQueueFlags queue_flags, VkPipelineStageFlags2KHR mask_param) {
SyncExecScope result;
result.mask_param = mask_param;
result.expanded_mask = sync_utils::ExpandPipelineStages(mask_param, queue_flags);
result.exec_scope = sync_utils::WithLaterPipelineStages(result.expanded_mask);
result.valid_accesses = SyncStageAccess::AccessScopeByStage(result.expanded_mask);
return result;
}
SyncBarrier::SyncBarrier(const SyncExecScope &src, const SyncExecScope &dst)
: src_exec_scope(src), src_access_scope(0), dst_exec_scope(dst), dst_access_scope(0) {}
SyncBarrier::SyncBarrier(const SyncExecScope &src, const SyncExecScope &dst, const SyncBarrier::AllAccess &)
: src_exec_scope(src), src_access_scope(src.valid_accesses), dst_exec_scope(dst), dst_access_scope(dst.valid_accesses) {}
template <typename Barrier>
SyncBarrier::SyncBarrier(const Barrier &barrier, const SyncExecScope &src, const SyncExecScope &dst)
: src_exec_scope(src),
src_access_scope(SyncStageAccess::AccessScope(src.valid_accesses, barrier.srcAccessMask)),
dst_exec_scope(dst),
dst_access_scope(SyncStageAccess::AccessScope(dst.valid_accesses, barrier.dstAccessMask)) {}
SyncBarrier::SyncBarrier(VkQueueFlags queue_flags, const VkSubpassDependency2 &subpass) {
const auto barrier = vku::FindStructInPNextChain<VkMemoryBarrier2KHR>(subpass.pNext);
if (barrier) {
auto src = SyncExecScope::MakeSrc(queue_flags, barrier->srcStageMask);
src_exec_scope = src;
src_access_scope = SyncStageAccess::AccessScope(src.valid_accesses, barrier->srcAccessMask);
auto dst = SyncExecScope::MakeDst(queue_flags, barrier->dstStageMask);
dst_exec_scope = dst;
dst_access_scope = SyncStageAccess::AccessScope(dst.valid_accesses, barrier->dstAccessMask);
} else {
auto src = SyncExecScope::MakeSrc(queue_flags, subpass.srcStageMask);
src_exec_scope = src;
src_access_scope = SyncStageAccess::AccessScope(src.valid_accesses, subpass.srcAccessMask);
auto dst = SyncExecScope::MakeDst(queue_flags, subpass.dstStageMask);
dst_exec_scope = dst;
dst_access_scope = SyncStageAccess::AccessScope(dst.valid_accesses, subpass.dstAccessMask);
}
}
template <typename Barrier>
SyncBarrier::SyncBarrier(VkQueueFlags queue_flags, const Barrier &barrier) {
auto src = SyncExecScope::MakeSrc(queue_flags, barrier.srcStageMask);
src_exec_scope = src.exec_scope;
src_access_scope = SyncStageAccess::AccessScope(src.valid_accesses, barrier.srcAccessMask);
auto dst = SyncExecScope::MakeDst(queue_flags, barrier.dstStageMask);
dst_exec_scope = dst.exec_scope;
dst_access_scope = SyncStageAccess::AccessScope(dst.valid_accesses, barrier.dstAccessMask);
}
// Apply a list of barriers, without resolving pending state, useful for subpass layout transitions
void ResourceAccessState::ApplyBarriers(const std::vector<SyncBarrier> &barriers, bool layout_transition) {
const UntaggedScopeOps scope;
for (const auto &barrier : barriers) {
ApplyBarrier(scope, barrier, layout_transition);
}
}
// ApplyBarriers is design for *fully* inclusive barrier lists without layout tranistions. Designed use was for
// inter-subpass barriers for lazy-evaluation of parent context memory ranges. Subpass layout transistions are *not* done
// lazily, s.t. no previous access reports should need layout transitions.
void ResourceAccessState::ApplyBarriersImmediate(const std::vector<SyncBarrier> &barriers) {
assert(!HasPendingState()); // This should never be call in the middle of another barrier application
const UntaggedScopeOps scope;
for (const auto &barrier : barriers) {
ApplyBarrier(scope, barrier, false);
}
ApplyPendingBarriers(kInvalidTag); // There can't be any need for this tag
}
HazardResult ResourceAccessState::DetectHazard(const SyncStageAccessInfoType &usage_info) const {
HazardResult hazard;
const auto &usage_stage = usage_info.stage_mask;
if (IsRead(usage_info)) {
if (IsRAWHazard(usage_info)) {
hazard.Set(this, usage_info, READ_AFTER_WRITE, *last_write);
}
} else {
// Write operation:
// Check for read operations more recent than last_write (as setting last_write clears reads, that would be *any*
// If reads exists -- test only against them because either:
// * the reads were hazards, and we've reported the hazard, so just test the current write vs. the read operations
// * the read weren't hazards, and thus if the write is safe w.r.t. the reads, no hazard vs. last_write is possible if
// the current write happens after the reads, so just test the write against the reades
// Otherwise test against last_write
//
// Look for casus belli for WAR
if (last_reads.size()) {
for (const auto &read_access : last_reads) {
if (IsReadHazard(usage_stage, read_access)) {
hazard.Set(this, usage_info, WRITE_AFTER_READ, read_access.access, read_access.tag);
break;
}
}
} else if (last_write.has_value() && last_write->IsWriteHazard(usage_info)) {
// Write-After-Write check -- if we have a previous write to test against
hazard.Set(this, usage_info, WRITE_AFTER_WRITE, *last_write);
}
}
return hazard;
}
HazardResult ResourceAccessState::DetectHazard(const SyncStageAccessInfoType &usage_info, const SyncOrdering ordering_rule,
QueueId queue_id) const {
const auto &ordering = GetOrderingRules(ordering_rule);
return DetectHazard(usage_info, ordering, queue_id);
}
HazardResult ResourceAccessState::DetectHazard(const SyncStageAccessInfoType &usage_info, const OrderingBarrier &ordering,
QueueId queue_id) const {
// The ordering guarantees act as barriers to the last accesses, independent of synchronization operations
HazardResult hazard;
const auto &usage_bit = usage_info.stage_access_bit;
const auto &usage_stage = usage_info.stage_mask;
const auto &usage_index = usage_info.stage_access_index;
const bool input_attachment_ordering = ordering.access_scope[SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ];
if (IsRead(usage_info)) {
// Exclude RAW if no write, or write not most "most recent" operation w.r.t. usage;
bool is_raw_hazard = IsRAWHazard(usage_info);
if (is_raw_hazard) {
// NOTE: we know last_write is non-zero
// See if the ordering rules save us from the simple RAW check above
// First check to see if the current usage is covered by the ordering rules
const bool usage_is_input_attachment = (usage_index == SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ);
const bool usage_is_ordered =
(input_attachment_ordering && usage_is_input_attachment) || (0 != (usage_stage & ordering.exec_scope));
if (usage_is_ordered) {
// Now see of the most recent write (or a subsequent read) are ordered
const bool most_recent_is_ordered =
last_write->IsOrdered(ordering, queue_id) || (0 != GetOrderedStages(queue_id, ordering));
is_raw_hazard = !most_recent_is_ordered;
}
}
if (is_raw_hazard) {
hazard.Set(this, usage_info, READ_AFTER_WRITE, *last_write);
}
} else if (usage_index == SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION) {
// For Image layout transitions, the barrier represents the first synchronization/access scope of the layout transition
return DetectBarrierHazard(usage_info, queue_id, ordering.exec_scope, ordering.access_scope);
} else {
// Only check for WAW if there are no reads since last_write
const bool usage_write_is_ordered = (usage_bit & ordering.access_scope).any();
if (last_reads.size()) {
// Look for any WAR hazards outside the ordered set of stages
VkPipelineStageFlags2KHR ordered_stages = VK_PIPELINE_STAGE_2_NONE;
if (usage_write_is_ordered) {
// If the usage is ordered, we can ignore all ordered read stages w.r.t. WAR)
ordered_stages = GetOrderedStages(queue_id, ordering);
}
// If we're tracking any reads that aren't ordered against the current write, got to check 'em all.
if ((ordered_stages & last_read_stages) != last_read_stages) {
for (const auto &read_access : last_reads) {
if (read_access.stage & ordered_stages) continue; // but we can skip the ordered ones
if (IsReadHazard(usage_stage, read_access)) {
hazard.Set(this, usage_info, WRITE_AFTER_READ, read_access.access, read_access.tag);
break;
}
}
}
} else if (last_write.has_value() && !(last_write->IsOrdered(ordering, queue_id) && usage_write_is_ordered)) {
bool ilt_ilt_hazard = false;
if ((usage_index == SYNC_IMAGE_LAYOUT_TRANSITION) && (last_write->IsIndex(SYNC_IMAGE_LAYOUT_TRANSITION))) {
// ILT after ILT is a special case where we check the 2nd access scope of the first ILT against the first access
// scope of the second ILT, which has been passed (smuggled?) in the ordering barrier
ilt_ilt_hazard = !(last_write->Barriers() & ordering.access_scope).any();
}
if (ilt_ilt_hazard || last_write->IsWriteHazard(usage_info)) {
hazard.Set(this, usage_info, WRITE_AFTER_WRITE, *last_write);
}
}
}
return hazard;
}
HazardResult ResourceAccessState::DetectHazard(const ResourceAccessState &recorded_use, QueueId queue_id,
const ResourceUsageRange &tag_range) const {
HazardResult hazard;
using Size = FirstAccesses::size_type;
const auto &recorded_accesses = recorded_use.first_accesses_;
Size count = recorded_accesses.size();
if (count) {
// First access is only closed if the last is a write
bool do_write_last = recorded_use.first_access_closed_;
if (do_write_last) {
// Note: We know count > 0 so this is alway safe.
--count;
}
for (Size i = 0; i < count; ++count) {
const auto &first = recorded_accesses[i];
// Skip and quit logic
if (first.tag < tag_range.begin) continue;
if (first.tag >= tag_range.end) {
do_write_last = false; // ignore last since we know it can't be in tag_range
break;
}
hazard = DetectHazard(*first.usage_info, first.ordering_rule, queue_id);
if (hazard.IsHazard()) {
hazard.AddRecordedAccess(first);
break;
}
}
if (do_write_last) {
// Writes are a bit special... both for the "most recent" access logic, and layout transition specific logic
const auto &last_access = recorded_accesses.back();
if (tag_range.includes(last_access.tag)) {
OrderingBarrier barrier = GetOrderingRules(last_access.ordering_rule);
if (last_access.usage_info->stage_access_index == SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION) {
// Or in the layout first access scope as a barrier... IFF the usage is an ILT
// this was saved off in the "apply barriers" logic to simplify ILT access checks as they straddle
// the barrier that applies them
barrier |= recorded_use.first_write_layout_ordering_;
}
// Any read stages present in the recorded context (this) are most recent to the write, and thus mask those stages
// in the active context
if (recorded_use.first_read_stages_) {
// we need to ignore the first use read stage in the active context (so we add them to the ordering rule),
// reads in the active context are not "most recent" as all recorded context operations are *after* them
// This supresses only RAW checks for stages present in the recorded context, but not those only present in the
// active context.
barrier.exec_scope |= recorded_use.first_read_stages_;
// if there are any first use reads, we suppress WAW by injecting the active context write in the ordering rule
barrier.access_scope |= last_access.usage_info->stage_access_bit;
}
hazard = DetectHazard(*last_access.usage_info, barrier, queue_id);
if (hazard.IsHazard()) {
hazard.AddRecordedAccess(last_access);
}
}
}
}
return hazard;
}
// Asynchronous Hazards occur between subpasses with no connection through the DAG
HazardResult ResourceAccessState::DetectAsyncHazard(const SyncStageAccessInfoType &usage_info,
const ResourceUsageTag start_tag) const {
HazardResult hazard;
// Async checks need to not go back further than the start of the subpass, as we only want to find hazards between the async
// subpasses. Anything older than that should have been checked at the start of each subpass, taking into account all of
// the raster ordering rules.
if (IsRead(usage_info)) {
if (last_write.has_value() && (last_write->tag_ >= start_tag)) {
hazard.Set(this, usage_info, READ_RACING_WRITE, *last_write);
}
} else {
if (last_write.has_value() && (last_write->tag_ >= start_tag)) {
hazard.Set(this, usage_info, WRITE_RACING_WRITE, *last_write);
} else if (last_reads.size() > 0) {
// Any reads during the other subpass will conflict with this write, so we need to check them all.
for (const auto &read_access : last_reads) {
if (read_access.tag >= start_tag) {
hazard.Set(this, usage_info, WRITE_RACING_READ, read_access.access, read_access.tag);
break;
}
}
}
}
return hazard;
}
HazardResult ResourceAccessState::DetectAsyncHazard(const ResourceAccessState &recorded_use, const ResourceUsageRange &tag_range,
ResourceUsageTag start_tag) const {
HazardResult hazard;
for (const auto &first : recorded_use.first_accesses_) {
// Skip and quit logic
if (first.tag < tag_range.begin) continue;
if (first.tag >= tag_range.end) break;
hazard = DetectAsyncHazard(*first.usage_info, start_tag);
if (hazard.IsHazard()) {
hazard.AddRecordedAccess(first);
break;
}
}
return hazard;
}
HazardResult ResourceAccessState::DetectBarrierHazard(const SyncStageAccessInfoType &usage_info, QueueId queue_id,
VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
// Only supporting image layout transitions for now
assert(usage_info.stage_access_index == SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION);
HazardResult hazard;
// only test for WAW if there no intervening read operations.
// See DetectHazard(SyncStagetAccessIndex) above for more details.
if (last_reads.size()) {
// Look at the reads if any
for (const auto &read_access : last_reads) {
if (read_access.IsReadBarrierHazard(queue_id, src_exec_scope)) {
hazard.Set(this, usage_info, WRITE_AFTER_READ, read_access.access, read_access.tag);
break;
}
}
} else if (last_write.has_value() && IsWriteBarrierHazard(queue_id, src_exec_scope, src_access_scope)) {
hazard.Set(this, usage_info, WRITE_AFTER_WRITE, *last_write);
}
return hazard;
}
HazardResult ResourceAccessState::DetectBarrierHazard(const SyncStageAccessInfoType &usage_info,
const ResourceAccessState &scope_state,
VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope, QueueId event_queue,
ResourceUsageTag event_tag) const {
// Only supporting image layout transitions for now
assert(usage_info.stage_access_index == SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION);
HazardResult hazard;
if (last_write.has_value() && (last_write->tag_ >= event_tag)) {
// Any write after the event precludes the possibility of being in the first access scope for the layout transition
hazard.Set(this, usage_info, WRITE_AFTER_WRITE, *last_write);
} else {
// only test for WAW if there no intervening read operations.
// See DetectHazard(SyncStagetAccessIndex) above for more details.
if (last_reads.size()) {
// Look at the reads if any... if reads exist, they are either the reason the access is in the event
// first scope, or they are a hazard.
const ReadStates &scope_reads = scope_state.last_reads;
const ReadStates::size_type scope_read_count = scope_reads.size();
// Since the hasn't been a write:
// * The current read state is a superset of the scoped one
// * The stage order is the same.
assert(last_reads.size() >= scope_read_count);
for (ReadStates::size_type read_idx = 0; read_idx < scope_read_count; ++read_idx) {
const ReadState &scope_read = scope_reads[read_idx];
const ReadState &current_read = last_reads[read_idx];
assert(scope_read.stage == current_read.stage);
if (current_read.tag > event_tag) {
// The read is more recent than the set event scope, thus no barrier from the wait/ILT.
hazard.Set(this, usage_info, WRITE_AFTER_READ, current_read.access, current_read.tag);
} else {
// The read is in the events first synchronization scope, so we use a barrier hazard check
// If the read stage is not in the src sync scope
// *AND* not execution chained with an existing sync barrier (that's the or)
// then the barrier access is unsafe (R/W after R)
if (scope_read.IsReadBarrierHazard(event_queue, src_exec_scope)) {
hazard.Set(this, usage_info, WRITE_AFTER_READ, scope_read.access, scope_read.tag);
break;
}
}
}
if (!hazard.IsHazard() && (last_reads.size() > scope_read_count)) {
const ReadState &current_read = last_reads[scope_read_count];
hazard.Set(this, usage_info, WRITE_AFTER_READ, current_read.access, current_read.tag);
}
} else if (last_write.has_value()) {
// if there are no reads, the write is either the reason the access is in the event scope... they are a hazard
// The write is in the first sync scope of the event (sync their aren't any reads to be the reason)
// So do a normal barrier hazard check
if (scope_state.IsWriteBarrierHazard(event_queue, src_exec_scope, src_access_scope)) {
hazard.Set(&scope_state, usage_info, WRITE_AFTER_WRITE, *scope_state.last_write);
}
}
}
return hazard;
}
void ResourceAccessState::MergePending(const ResourceAccessState &other) {
pending_layout_transition |= other.pending_layout_transition;
}
void ResourceAccessState::MergeReads(const ResourceAccessState &other) {
// Merge the read states
const auto pre_merge_count = last_reads.size();
const auto pre_merge_stages = last_read_stages;
for (uint32_t other_read_index = 0; other_read_index < other.last_reads.size(); other_read_index++) {
auto &other_read = other.last_reads[other_read_index];
if (pre_merge_stages & other_read.stage) {
// Merge in the barriers for read stages that exist in *both* this and other
// TODO: This is N^2 with stages... perhaps the ReadStates should be sorted by stage index.
// but we should wait on profiling data for that.
for (uint32_t my_read_index = 0; my_read_index < pre_merge_count; my_read_index++) {
auto &my_read = last_reads[my_read_index];
if (other_read.stage == my_read.stage) {
if (my_read.tag < other_read.tag) {
// Other is more recent, copy in the state
my_read.access = other_read.access;
my_read.tag = other_read.tag;
my_read.queue = other_read.queue;
my_read.pending_dep_chain = other_read.pending_dep_chain;
// TODO: Phase 2 -- review the state merge logic to avoid false positive from overwriting the barriers
// May require tracking more than one access per stage.
my_read.barriers = other_read.barriers;
my_read.sync_stages = other_read.sync_stages;
if (my_read.stage == VK_PIPELINE_STAGE_2_FRAGMENT_SHADER_BIT_KHR) {
// Since I'm overwriting the fragement stage read, also update the input attachment info
// as this is the only stage that affects it.
input_attachment_read = other.input_attachment_read;
}
} else if (other_read.tag == my_read.tag) {
// The read tags match so merge the barriers
my_read.barriers |= other_read.barriers;
my_read.sync_stages |= other_read.sync_stages;
my_read.pending_dep_chain |= other_read.pending_dep_chain;
}
break;
}
}
} else {
// The other read stage doesn't exist in this, so add it.
last_reads.emplace_back(other_read);
last_read_stages |= other_read.stage;
if (other_read.stage == VK_PIPELINE_STAGE_2_FRAGMENT_SHADER_BIT_KHR) {
input_attachment_read = other.input_attachment_read;
}
}
}
read_execution_barriers |= other.read_execution_barriers;
}
// The logic behind resolves is the same as update, we assume that earlier hazards have be reported, and that no
// tranistive hazard can exists with a hazard between the earlier operations. Yes, an early hazard can mask that another
// exists, but if you fix *that* hazard it either fixes or unmasks the subsequent ones.
void ResourceAccessState::Resolve(const ResourceAccessState &other) {
bool skip_first = false;
if (last_write.has_value()) {
if (other.last_write.has_value()) {
if (last_write->Tag() < other.last_write->Tag()) {
// NOTE: Both last and other have writes, and thus first access is "closed". We are selecting other's
// first_access state, but it and this can only differ if there are async hazards
// error state.
//
// If this is a later write, we've reported any exsiting hazard, and we can just overwrite as the more recent
// operation
*this = other;
skip_first = true;
} else if (last_write->Tag() == other.last_write->Tag()) {
// In the *equals* case for write operations, we merged the write barriers and the read state (but without the
// dependency chaining logic or any stage expansion)
last_write->MergeBarriers(*other.last_write);
MergePending(other);
MergeReads(other);
} else {
// other write is before this write... in which case we keep this instead of other
// and can skip the "first_access" merge, since first_access has been closed since other write tag or before
skip_first = true;
}
} else {
// this has a write and other doesn't -- at best async read in other, which have been reported, and will be dropped
// Since this has a write first access is closed and shouldn't be updated by other
skip_first = true;
}
} else if (other.last_write.has_value()) { // && not this->last_write
// Other has write and this doesn't, thus keep it, See first access NOTE above
*this = other;
skip_first = true;
} else { // not this->last_write OR other.last_write
// Neither state has a write, just merge the reads
MergePending(other);
MergeReads(other);
}
// Merge first access information by making a copy of this first_access and reconstructing with a shuffle
// of the copy and other into this using the update first logic.
// NOTE: All sorts of additional cleverness could be put into short circuts. (for example back is write and is before front
// of the other first_accesses... )
if (!skip_first && !(first_accesses_ == other.first_accesses_) && !other.first_accesses_.empty()) {
FirstAccesses firsts(std::move(first_accesses_));
ClearFirstUse();
auto a = firsts.begin();
auto a_end = firsts.end();
for (auto &b : other.first_accesses_) {
// TODO: Determine whether some tag offset will be needed for PHASE II
while ((a != a_end) && (a->tag < b.tag)) {
UpdateFirst(a->tag, *a->usage_info, a->ordering_rule);
++a;
}
UpdateFirst(b.tag, *b.usage_info, b.ordering_rule);
}
for (; a != a_end; ++a) {
UpdateFirst(a->tag, *a->usage_info, a->ordering_rule);
}
}
}
void ResourceAccessState::Update(const SyncStageAccessInfoType &usage_info, SyncOrdering ordering_rule,
const ResourceUsageTag tag) {
// Move this logic in the ResourceStateTracker as methods, thereof (or we'll repeat it for every flavor of resource...
const auto &usage_bit = usage_info.stage_access_bit;
const auto &usage_stage = usage_info.stage_mask;
if (IsRead(usage_info)) {
// Mulitple outstanding reads may be of interest and do dependency chains independently
// However, for purposes of barrier tracking, only one read per pipeline stage matters
if (usage_stage & last_read_stages) {
const auto not_usage_stage = ~usage_stage;
for (auto &read_access : last_reads) {
if (read_access.stage == usage_stage) {
read_access.Set(usage_stage, usage_bit, 0, tag);
} else if (read_access.barriers & usage_stage) {
// If the current access is barriered to this stage, mark it as "known to happen after"
read_access.sync_stages |= usage_stage;
} else {
// If the current access is *NOT* barriered to this stage it needs to be cleared.
// Note: this is possible because semaphores can *clear* effective barriers, so the assumption
// that sync_stages is a subset of barriers may not apply.
read_access.sync_stages &= not_usage_stage;
}
}
} else {
for (auto &read_access : last_reads) {
if (read_access.barriers & usage_stage) {
read_access.sync_stages |= usage_stage;
}
}
last_reads.emplace_back(usage_stage, usage_bit, 0, tag);
last_read_stages |= usage_stage;
}
// Fragment shader reads come in two flavors, and we need to track if the one we're tracking is the special one.
if (usage_stage == VK_PIPELINE_STAGE_2_FRAGMENT_SHADER_BIT_KHR) {
// TODO Revisit re: multiple reads for a given stage
input_attachment_read = (usage_bit == SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ_BIT);
}
} else {
// Assume write
// TODO determine what to do with READ-WRITE operations if any
SetWrite(usage_info, tag);
}
UpdateFirst(tag, usage_info, ordering_rule);
}
// Clobber last read and all barriers... because all we have is DANGER, DANGER, WILL ROBINSON!!!
// if the last_reads/last_write were unsafe, we've reported them, in either case the prior access is irrelevant.
// We can overwrite them as *this* write is now after them.
//
// Note: intentionally ignore pending barriers and chains (i.e. don't apply or clear them), let ApplyPendingBarriers handle them.
void ResourceAccessState::SetWrite(const SyncStageAccessInfoType &usage_info, const ResourceUsageTag tag) {
ClearRead();
if (last_write.has_value()) {
last_write->Set(usage_info, tag);
} else {
last_write.emplace(usage_info, tag);
}
}
void ResourceAccessState::ClearWrite() { last_write.reset(); }
void ResourceAccessState::ClearRead() {
last_reads.clear();
last_read_stages = VK_PIPELINE_STAGE_2_NONE;
read_execution_barriers = VK_PIPELINE_STAGE_2_NONE;
input_attachment_read = false; // Denotes no outstanding input attachment read after the last write.
}
void ResourceAccessState::ClearPending() {
pending_layout_transition = false;
if (last_write.has_value()) last_write->ClearPending();
}
void ResourceAccessState::ClearFirstUse() {
first_accesses_.clear();
first_read_stages_ = VK_PIPELINE_STAGE_2_NONE;
first_write_layout_ordering_ = OrderingBarrier();
first_access_closed_ = false;
}
// Apply the memory barrier without updating the existing barriers. The execution barrier
// changes the "chaining" state, but to keep barriers independent, we defer this until all barriers
// of the batch have been processed. Also, depending on whether layout transition happens, we'll either
// replace the current write barriers or add to them, so accumulate to pending as well.
template <typename ScopeOps>
void ResourceAccessState::ApplyBarrier(ScopeOps &&scope, const SyncBarrier &barrier, bool layout_transition) {
// For independent barriers we need to track what the new barriers and dependency chain *will* be when we're done
// applying the memory barriers
// NOTE: We update the write barrier if the write is in the first access scope or if there is a layout
// transistion, under the theory of "most recent access". If the resource acces *isn't* safe
// vs. this layout transition DetectBarrierHazard should report it. We treat the layout
// transistion *as* a write and in scope with the barrier (it's before visibility).
if (layout_transition) {
if (!last_write.has_value()) {
last_write.emplace(UsageInfo(SYNC_ACCESS_INDEX_NONE), 0U);
}
last_write->UpdatePendingBarriers(barrier);
last_write->UpdatePendingLayoutOrdering(barrier);
pending_layout_transition = true;
} else {
if (scope.WriteInScope(barrier, *this)) {
last_write->UpdatePendingBarriers(barrier);
}
if (!pending_layout_transition) {
// Once we're dealing with a layout transition (which is modelled as a *write*) then the last reads/chains
// don't need to be tracked as we're just going to clear them.
VkPipelineStageFlags2 stages_in_scope = VK_PIPELINE_STAGE_2_NONE;
for (auto &read_access : last_reads) {
// The | implements the "dependency chain" logic for this access, as the barriers field stores the second sync
// scope
if (scope.ReadInScope(barrier, read_access)) {
// We'll apply the barrier in the next loop, because it's DRY'r to do it one place.
stages_in_scope |= read_access.stage;
}
}
for (auto &read_access : last_reads) {
if (0 != ((read_access.stage | read_access.sync_stages) & stages_in_scope)) {
// If this stage, or any stage known to be synchronized after it are in scope, apply the barrier to this
// read NOTE: Forwarding barriers to known prior stages changes the sync_stages from shallow to deep,
// because the
// barriers used to determine sync_stages have been propagated to all known earlier stages
read_access.ApplyReadBarrier(barrier.dst_exec_scope.exec_scope);
}
}
}
}
}
void ResourceAccessState::ApplyPendingBarriers(const ResourceUsageTag tag) {
if (pending_layout_transition) {
// SetWrite clobbers the last_reads array, and thus we don't have to clear the read_state out.
const SyncStageAccessInfoType &layout_usage_info = UsageInfo(SYNC_IMAGE_LAYOUT_TRANSITION);
SetWrite(layout_usage_info, tag); // Side effect notes below
UpdateFirst(tag, layout_usage_info, SyncOrdering::kNonAttachment);
TouchupFirstForLayoutTransition(tag, last_write->GetPendingLayoutOrdering());
last_write->ApplyPendingBarriers();
last_write->ClearPending();
pending_layout_transition = false;
} else {
// Apply the accumulate execution barriers (and thus update chaining information)
// for layout transition, last_reads is reset by SetWrite, so this will be skipped.
for (auto &read_access : last_reads) {
read_execution_barriers |= read_access.ApplyPendingBarriers();
}
// We OR in the accumulated write chain and barriers even in the case of a layout transition as SetWrite zeros them.
if (last_write.has_value()) {
last_write->ApplyPendingBarriers();
last_write->ClearPending();
}
}
}
// Assumes signal queue != wait queue
void ResourceAccessState::ApplySemaphore(const SemaphoreScope &signal, const SemaphoreScope wait) {
// Semaphores only guarantee the first scope of the signal is before the second scope of the wait.
// If any access isn't in the first scope, there are no guarantees, thus those barriers are cleared
assert(signal.queue != wait.queue);
for (auto &read_access : last_reads) {
if (read_access.ReadInQueueScopeOrChain(signal.queue, signal.exec_scope)) {
// Deflects WAR on wait queue
read_access.barriers = wait.exec_scope;
} else {
// Leave sync stages alone. Update method will clear unsynchronized stages on subsequent reads as needed.
read_access.barriers = VK_PIPELINE_STAGE_2_NONE;
}
}
if (WriteInQueueSourceScopeOrChain(signal.queue, signal.exec_scope, signal.valid_accesses)) {
assert(last_write.has_value());
// Will deflect RAW wait queue, WAW needs a chained barrier on wait queue
read_execution_barriers = wait.exec_scope;
last_write->barriers_ = wait.valid_accesses;
} else {
read_execution_barriers = VK_PIPELINE_STAGE_2_NONE;
if (last_write.has_value()) last_write->barriers_.reset();
}
if (last_write.has_value()) last_write->dependency_chain_ = read_execution_barriers;
}
// Read access predicate for queue wait
bool ResourceAccessState::WaitQueueTagPredicate::operator()(const ResourceAccessState::ReadState &read_access) const {
return (read_access.queue == queue) && (read_access.tag <= tag) &&
(read_access.stage != VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL);
}
bool ResourceAccessState::WaitQueueTagPredicate::operator()(const ResourceAccessState &access) const {
if (!access.last_write.has_value()) return false;
const auto &write_state = *access.last_write;
return write_state.IsQueue(queue) && (write_state.Tag() <= tag) &&
!write_state.IsIndex(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL);
}
// Read access predicate for queue wait
bool ResourceAccessState::WaitTagPredicate::operator()(const ResourceAccessState::ReadState &read_access) const {
return (read_access.tag <= tag) && (read_access.stage != VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL);
}
bool ResourceAccessState::WaitTagPredicate::operator()(const ResourceAccessState &access) const {
if (!access.last_write.has_value()) return false;
const auto &write_state = *access.last_write;
return (write_state.Tag() <= tag) && !write_state.IsIndex(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL);
}
// Present operations only matching only the *exactly* tagged present and acquire operations
bool ResourceAccessState::WaitAcquirePredicate::operator()(const ResourceAccessState::ReadState &read_access) const {
return (read_access.tag == acquire_tag) && (read_access.stage == VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL);
}
bool ResourceAccessState::WaitAcquirePredicate::operator()(const ResourceAccessState &access) const {
if (!access.last_write.has_value()) return false;
const auto &write_state = *access.last_write;
return (write_state.Tag() == present_tag) && write_state.IsIndex(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL);
}
// Return if the resulting state is "empty"
template <typename Predicate>
bool ResourceAccessState::ApplyPredicatedWait(Predicate &predicate) {
VkPipelineStageFlags2KHR sync_reads = VK_PIPELINE_STAGE_2_NONE;
// Use the predicate to build a mask of the read stages we are synchronizing
// Use the sync_stages to also detect reads known to be before any synchronized reads (first pass)
for (auto &read_access : last_reads) {
if (predicate(read_access)) {
// If we know this stage is before any stage we syncing, or if the predicate tells us that we are waited for..
sync_reads |= read_access.stage;
}
}
// Now that we know the reads directly in scopejust need to go over the list again to pick up the "known earlier" stages.
// NOTE: sync_stages is "deep" catching all stages synchronized after it because we forward barriers
uint32_t unsync_count = 0;
for (auto &read_access : last_reads) {
if (0 != ((read_access.stage | read_access.sync_stages) & sync_reads)) {
// This is redundant in the "stage" case, but avoids a second branch to get an accurate count
sync_reads |= read_access.stage;
} else {
++unsync_count;
}
}
if (unsync_count) {
if (sync_reads) {
// When have some remaining unsynchronized reads, we have to rewrite the last_reads array.
ReadStates unsync_reads;
unsync_reads.reserve(unsync_count);
VkPipelineStageFlags2KHR unsync_read_stages = VK_PIPELINE_STAGE_2_NONE;
for (auto &read_access : last_reads) {
if (0 == (read_access.stage & sync_reads)) {
unsync_reads.emplace_back(read_access);
unsync_read_stages |= read_access.stage;
}
}
last_read_stages = unsync_read_stages;
last_reads = std::move(unsync_reads);
}
} else {
// Nothing remains (or it was empty to begin with)
ClearRead();
}
bool all_clear = last_reads.size() == 0;
if (last_write.has_value()) {
if (predicate(*this) || sync_reads) {
// Clear any predicated write, or any the write from any any access with synchronized reads.
// This could drop RAW detection, but only if the synchronized reads were RAW hazards, and given
// MRR approach to reporting, this is consistent with other drops, especially since fixing the
// RAW wit the sync_reads stages would preclude a subsequent RAW.
ClearWrite();
} else {
all_clear = false;
}
}
return all_clear;
}
bool ResourceAccessState::FirstAccessInTagRange(const ResourceUsageRange &tag_range) const {
if (!first_accesses_.size()) return false;
const ResourceUsageRange first_access_range = {first_accesses_.front().tag, first_accesses_.back().tag + 1};
return tag_range.intersects(first_access_range);
}
void ResourceAccessState::OffsetTag(ResourceUsageTag offset) {
if (last_write.has_value()) last_write->OffsetTag(offset);
for (auto &read_access : last_reads) {
read_access.tag += offset;
}
for (auto &first : first_accesses_) {
first.tag += offset;
}
}
static const SyncStageAccessFlags kAllSyncStageAccessBits = ~SyncStageAccessFlags(0);
ResourceAccessState::ResourceAccessState()
: last_write(),
last_read_stages(0),
read_execution_barriers(VK_PIPELINE_STAGE_2_NONE),
last_reads(),
input_attachment_read(false),
pending_layout_transition(false),
first_accesses_(),
first_read_stages_(VK_PIPELINE_STAGE_2_NONE),
first_write_layout_ordering_(),
first_access_closed_(false) {}
// This should be just Bits or Index, but we don't have an invalid state for Index
VkPipelineStageFlags2KHR ResourceAccessState::GetReadBarriers(const SyncStageAccessFlags &usage_bit) const {
VkPipelineStageFlags2KHR barriers = VK_PIPELINE_STAGE_2_NONE;
for (const auto &read_access : last_reads) {
if ((read_access.access & usage_bit).any()) {
barriers = read_access.barriers;
break;
}
}
return barriers;
}
void ResourceAccessState::SetQueueId(QueueId id) {
for (auto &read_access : last_reads) {
if (read_access.queue == QueueSyncState::kQueueIdInvalid) {
read_access.queue = id;
}
}
if (last_write.has_value()) last_write->SetQueueId(id);
}
bool ResourceAccessState::IsWriteBarrierHazard(QueueId queue_id, VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
return last_write.has_value() && last_write->IsWriteBarrierHazard(queue_id, src_exec_scope, src_access_scope);
}
bool ResourceAccessState::WriteInSourceScopeOrChain(VkPipelineStageFlags2KHR src_exec_scope,
SyncStageAccessFlags src_access_scope) const {
return last_write.has_value() && last_write->WriteInSourceScopeOrChain(src_exec_scope, src_access_scope);
}
bool ResourceAccessState::WriteInQueueSourceScopeOrChain(QueueId queue, VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
return last_write.has_value() && last_write->WriteInQueueSourceScopeOrChain(queue, src_exec_scope, src_access_scope);
}
bool ResourceAccessState::WriteInEventScope(VkPipelineStageFlags2KHR src_exec_scope, const SyncStageAccessFlags &src_access_scope,
QueueId scope_queue, ResourceUsageTag scope_tag) const {
return last_write.has_value() && last_write->WriteInEventScope(src_exec_scope, src_access_scope, scope_queue, scope_tag);
}
// As ReadStates must be unique by stage, this is as good a sort as needed
bool operator<(const ResourceAccessState::ReadState &lhs, const ResourceAccessState::ReadState &rhs) {
return lhs.stage < rhs.stage;
}
void ResourceAccessState::Normalize() {
if (!last_reads.size()) {
ClearRead();
} else {
// Sort the reads in stage order for consistent comparisons
std::sort(last_reads.begin(), last_reads.end());
for (auto &read_access : last_reads) {
read_access.Normalize();
}
}
ClearPending();
ClearFirstUse();
}
void ResourceAccessState::GatherReferencedTags(ResourceUsageTagSet &used) const {
if (last_write.has_value()) {
used.CachedInsert(last_write->Tag());
}
for (const auto &read_access : last_reads) {
used.CachedInsert(read_access.tag);
}
}
bool ResourceAccessState::IsRAWHazard(const SyncStageAccessInfoType &usage_info) const {
assert(IsRead(usage_info));
// Only RAW vs. last_write if it doesn't happen-after any other read because either:
// * the previous reads are not hazards, and thus last_write must be visible and available to
// any reads that happen after.
// * the previous reads *are* hazards to last_write, have been reported, and if that hazard is fixed
// the current read will be also not be a hazard, thus reporting a hazard here adds no needed information.
return last_write.has_value() && (0 == (read_execution_barriers & usage_info.stage_mask)) &&
last_write->IsWriteHazard(usage_info);
}
VkPipelineStageFlags2 ResourceAccessState::GetOrderedStages(QueueId queue_id, const OrderingBarrier &ordering) const {
// At apply queue submission order limits on the effect of ordering
VkPipelineStageFlags2 non_qso_stages = VK_PIPELINE_STAGE_2_NONE;
if (queue_id != QueueSyncState::kQueueIdInvalid) {
for (const auto &read_access : last_reads) {
if (read_access.queue != queue_id) {
non_qso_stages |= read_access.stage;
}
}
}
// Whether the stage are in the ordering scope only matters if the current write is ordered
const VkPipelineStageFlags2 read_stages_in_qso = last_read_stages & ~non_qso_stages;
VkPipelineStageFlags2 ordered_stages = read_stages_in_qso & ordering.exec_scope;
// Special input attachment handling as always (not encoded in exec_scop)
const bool input_attachment_ordering = ordering.access_scope[SYNC_FRAGMENT_SHADER_INPUT_ATTACHMENT_READ];
if (input_attachment_ordering && input_attachment_read) {
// If we have an input attachment in last_reads and input attachments are ordered we all that stage
ordered_stages |= VK_PIPELINE_STAGE_2_FRAGMENT_SHADER_BIT_KHR;
}
return ordered_stages;
}
void ResourceAccessState::UpdateFirst(const ResourceUsageTag tag, const SyncStageAccessInfoType &usage_info,
SyncOrdering ordering_rule) {
// Only record until we record a write.
if (!first_access_closed_) {
const bool is_read = IsRead(usage_info);
const VkPipelineStageFlags2KHR usage_stage = is_read ? usage_info.stage_mask : 0U;
if (0 == (usage_stage & first_read_stages_)) {
// If this is a read we haven't seen or a write, record.
// We always need to know what stages were found prior to write
first_read_stages_ |= usage_stage;
if (0 == (read_execution_barriers & usage_stage)) {
// If this stage isn't masked then we add it (since writes map to usage_stage 0, this also records writes)
first_accesses_.emplace_back(tag, usage_info, ordering_rule);
first_access_closed_ = !is_read;
}
}
}
}
void ResourceAccessState::TouchupFirstForLayoutTransition(ResourceUsageTag tag, const OrderingBarrier &layout_ordering) {
// Only call this after recording an image layout transition
assert(first_accesses_.size());
if (first_accesses_.back().tag == tag) {
// If this layout transition is the the first write, add the additional ordering rules that guard the ILT
assert(first_accesses_.back().usage_info->stage_access_index == SyncStageAccessIndex::SYNC_IMAGE_LAYOUT_TRANSITION);
first_write_layout_ordering_ = layout_ordering;
}
}
ResourceAccessState::ReadState::ReadState(VkPipelineStageFlags2KHR stage_, SyncStageAccessFlags access_,
VkPipelineStageFlags2KHR barriers_, ResourceUsageTag tag_)
: stage(stage_),
access(access_),
barriers(barriers_),
sync_stages(VK_PIPELINE_STAGE_2_NONE),
tag(tag_),
queue(QueueSyncState::kQueueIdInvalid),
pending_dep_chain(VK_PIPELINE_STAGE_2_NONE) {}
void ResourceAccessState::ReadState::Set(VkPipelineStageFlags2KHR stage_, const SyncStageAccessFlags &access_,
VkPipelineStageFlags2KHR barriers_, ResourceUsageTag tag_) {
stage = stage_;
access = access_;
barriers = barriers_;
sync_stages = VK_PIPELINE_STAGE_2_NONE;
tag = tag_;
pending_dep_chain = VK_PIPELINE_STAGE_2_NONE; // If this is a new read, we aren't applying a barrier set.
}
// Scope test including "queue submission order" effects. Specifically, accesses from a different queue are not
// considered to be in "queue submission order" with barriers, events, or semaphore signalling, but any barriers
// that have bee applied (via semaphore) to those accesses can be chained off of.
bool ResourceAccessState::ReadState::ReadInQueueScopeOrChain(QueueId scope_queue, VkPipelineStageFlags2 exec_scope) const {
VkPipelineStageFlags2 effective_stages = barriers | ((scope_queue == queue) ? stage : VK_PIPELINE_STAGE_2_NONE);
return (exec_scope & effective_stages) != 0;
}
VkPipelineStageFlags2 ResourceAccessState::ReadState::ApplyPendingBarriers() {
barriers |= pending_dep_chain;
pending_dep_chain = VK_PIPELINE_STAGE_2_NONE;
return barriers;
}
ResourceUsageRange SyncValidator::ReserveGlobalTagRange(size_t tag_count) const {
ResourceUsageRange reserve;
reserve.begin = tag_limit_.fetch_add(tag_count);
reserve.end = reserve.begin + tag_count;
return reserve;
}
void SyncValidator::ApplyTaggedWait(QueueId queue_id, ResourceUsageTag tag) {
auto tagged_wait_op = [queue_id, tag](const std::shared_ptr<QueueBatchContext> &batch) {
batch->ApplyTaggedWait(queue_id, tag);
batch->Trim();
};
ForAllQueueBatchContexts(tagged_wait_op);
}
void SyncValidator::ApplyAcquireWait(const AcquiredImage &acquired) {
auto acq_wait_op = [&acquired](const std::shared_ptr<QueueBatchContext> &batch) {
batch->ApplyAcquireWait(acquired);
batch->Trim();
};
ForAllQueueBatchContexts(acq_wait_op);
}
template <typename BatchOp>
void SyncValidator::ForAllQueueBatchContexts(BatchOp &&op) {
// Often we need to go through every queue batch context and apply synchronization operations
// As usual -- two groups, the "last batch" and the signaled semaphores
QueueBatchContext::BatchSet queue_batch_contexts = GetQueueBatchSnapshot();
// Note: The const is to force the reference to const be on all platforms.
//
// It's not obivious (nor cross platform consitent), that the batch reference should be const
// but since it's pointing to the actual *key* for the set it must be. This doesn't make the
// object the shared pointer is referencing constant however.
for (const auto &batch : queue_batch_contexts) {
op(batch);
}
}
void SyncValidator::UpdateFenceWaitInfo(VkFence fence, QueueId queue_id, ResourceUsageTag tag) {
std::shared_ptr<const FENCE_STATE> fence_state = Get<FENCE_STATE>(fence);
UpdateFenceWaitInfo(fence_state, FenceSyncState(fence_state, queue_id, tag));
}
void SyncValidator::UpdateFenceWaitInfo(VkFence fence, const PresentedImage &image, ResourceUsageTag tag) {
std::shared_ptr<const FENCE_STATE> fence_state = Get<FENCE_STATE>(fence);
UpdateFenceWaitInfo(fence_state, FenceSyncState(fence_state, image, tag));
}
void SyncValidator::UpdateFenceWaitInfo(std::shared_ptr<const FENCE_STATE> &fence_state, FenceSyncState &&wait_info) {
if (BASE_NODE::Invalid(fence_state)) return;
waitable_fences_[fence_state->fence()] = std::move(wait_info);
}
void SyncValidator::WaitForFence(VkFence fence) {
auto fence_it = waitable_fences_.find(fence);
if (fence_it != waitable_fences_.end()) {
// The fence may no longer be waitable for several valid reasons.
FenceSyncState &wait_for = fence_it->second;
if (wait_for.acquired.Invalid()) {
// This is just a normal fence wait
ApplyTaggedWait(wait_for.queue_id, wait_for.tag);
} else {
// This a fence wait for a present operation
ApplyAcquireWait(wait_for.acquired);
}
waitable_fences_.erase(fence_it);
}
}
void SyncValidator::UpdateSyncImageMemoryBindState(uint32_t count, const VkBindImageMemoryInfo *infos) {
for (const auto &info : vvl::make_span(infos, count)) {
if (VK_NULL_HANDLE == info.image) continue;
auto image_state = Get<ImageState>(info.image);
if (image_state->IsTiled()) {
image_state->SetOpaqueBaseAddress(*this);
}
}
}
const QueueSyncState *SyncValidator::GetQueueSyncState(VkQueue queue) const {
return GetMappedPlainFromShared(queue_sync_states_, queue);
}
QueueSyncState *SyncValidator::GetQueueSyncState(VkQueue queue) { return GetMappedPlainFromShared(queue_sync_states_, queue); }
std::shared_ptr<const QueueSyncState> SyncValidator::GetQueueSyncStateShared(VkQueue queue) const {
return GetMapped(queue_sync_states_, queue, []() { return std::shared_ptr<QueueSyncState>(); });
}
std::shared_ptr<QueueSyncState> SyncValidator::GetQueueSyncStateShared(VkQueue queue) {
return GetMapped(queue_sync_states_, queue, []() { return std::shared_ptr<QueueSyncState>(); });
}
template <typename T>
struct GetBatchTraits {};
template <>
struct GetBatchTraits<std::shared_ptr<QueueSyncState>> {
using Batch = std::shared_ptr<QueueBatchContext>;
static Batch Get(const std::shared_ptr<QueueSyncState> &qss) { return qss ? qss->LastBatch() : Batch(); }
};
template <>
struct GetBatchTraits<std::shared_ptr<SignaledSemaphores::Signal>> {
using Batch = std::shared_ptr<QueueBatchContext>;
static Batch Get(const std::shared_ptr<SignaledSemaphores::Signal> &sig) { return sig ? sig->batch : Batch(); }
};
template <typename BatchSet, typename Map, typename Predicate>
static BatchSet GetQueueBatchSnapshotImpl(const Map &map, Predicate &&pred) {
BatchSet snapshot;
for (auto &entry : map) {
// Intentional copy
auto batch = GetBatchTraits<typename Map::mapped_type>::Get(entry.second);
if (batch && pred(batch)) snapshot.emplace(std::move(batch));
}
return snapshot;
}
template <typename Predicate>
QueueBatchContext::ConstBatchSet SyncValidator::GetQueueLastBatchSnapshot(Predicate &&pred) const {
return GetQueueBatchSnapshotImpl<QueueBatchContext::ConstBatchSet>(queue_sync_states_, std::forward<Predicate>(pred));
}
template <typename Predicate>
QueueBatchContext::BatchSet SyncValidator::GetQueueLastBatchSnapshot(Predicate &&pred) {
return GetQueueBatchSnapshotImpl<QueueBatchContext::BatchSet>(queue_sync_states_, std::forward<Predicate>(pred));
}
QueueBatchContext::BatchSet SyncValidator::GetQueueBatchSnapshot() {
QueueBatchContext::BatchSet snapshot = GetQueueLastBatchSnapshot();
auto append = [&snapshot](const std::shared_ptr<QueueBatchContext> &batch) {
if (batch && !vvl::Contains(snapshot, batch)) {
snapshot.emplace(batch);
}
return false;
};
GetQueueBatchSnapshotImpl<QueueBatchContext::BatchSet>(signaled_semaphores_, append);
return snapshot;
}
struct QueueSubmitCmdState {
std::shared_ptr<const QueueSyncState> queue;
std::shared_ptr<QueueBatchContext> last_batch;
const ErrorObject &error_obj;
SignaledSemaphores signaled;
QueueSubmitCmdState(const ErrorObject &error_obj, const SignaledSemaphores &parent_semaphores)
: error_obj(error_obj), signaled(parent_semaphores) {}
};
bool QueueBatchContext::DoQueueSubmitValidate(const SyncValidator &sync_state, QueueSubmitCmdState &cmd_state,
const VkSubmitInfo2 &batch_info) {
bool skip = false;
// For each submit in the batch...
for (const auto &cb : command_buffers_) {
const auto &cb_access_context = cb.cb->access_context;
if (cb_access_context.GetTagLimit() == 0) {
batch_.cb_index++;
continue; // Skip empty CB's but also skip the unused index for correct reporting
}
skip |= ReplayState(*this, cb_access_context, cmd_state.error_obj, cb.index).ValidateFirstUse();
// The barriers have already been applied in ValidatFirstUse
ResourceUsageRange tag_range = ImportRecordedAccessLog(cb_access_context);
ResolveSubmittedCommandBuffer(*cb_access_context.GetCurrentAccessContext(), tag_range.begin);
}
return skip;
}
bool SignaledSemaphores::SignalSemaphore(const std::shared_ptr<const SEMAPHORE_STATE> &sem_state,
const std::shared_ptr<QueueBatchContext> &batch,
const VkSemaphoreSubmitInfo &signal_info) {
assert(batch);
const SyncExecScope exec_scope =
SyncExecScope::MakeSrc(batch->GetQueueFlags(), signal_info.stageMask, VK_PIPELINE_STAGE_2_HOST_BIT);
std::shared_ptr<Signal> signal = std::make_shared<Signal>(sem_state, batch, exec_scope);
return Insert(sem_state, std::move(signal));
}
bool SignaledSemaphores::Insert(const std::shared_ptr<const SEMAPHORE_STATE> &sem_state, std::shared_ptr<Signal> &&signal) {
const VkSemaphore sem = sem_state->semaphore();
auto signal_it = signaled_.find(sem);
std::shared_ptr<Signal> insert_signal;
if (signal_it == signaled_.end()) {
if (prev_) {
auto prev_sig = GetMapped(prev_->signaled_, sem_state->semaphore(), []() { return std::shared_ptr<Signal>(); });
if (prev_sig) {
// The is an invalid signal, as this semaphore is already signaled... copy the prev state (as prev_ is const)
insert_signal = std::make_shared<Signal>(*prev_sig);
}
}
auto insert_pair = signaled_.emplace(sem, std::move(insert_signal));
signal_it = insert_pair.first;
}
bool success = false;
if (!signal_it->second) {
signal_it->second = std::move(signal);
success = true;
}
return success;
}
bool SignaledSemaphores::SignalSemaphore(const std::shared_ptr<const SEMAPHORE_STATE> &sem_state, const PresentedImage &presented,
ResourceUsageTag acq_tag) {
// Ignore any signal we haven't waited... CoreChecks should have reported this
std::shared_ptr<Signal> signal = std::make_shared<Signal>(sem_state, presented, acq_tag);
return Insert(sem_state, std::move(signal));
}
std::shared_ptr<const SignaledSemaphores::Signal> SignaledSemaphores::Unsignal(VkSemaphore sem) {
std::shared_ptr<const Signal> unsignaled;
const auto found_it = signaled_.find(sem);
if (found_it != signaled_.end()) {
// Move the unsignaled singal out from the signaled list, but keep the shared_ptr as the caller needs the contents for
// a bit.
unsignaled = std::move(found_it->second);
if (!prev_) {
// No parent, not need to keep the entry
// IFF (prev_) leave the entry in the leaf table as we use it to export unsignal to prev_ during record phase
signaled_.erase(found_it);
}
} else if (prev_) {
// We can't unsignal prev_ because it's const * by design.
// We put in an empty placeholder
signaled_.emplace(sem, std::shared_ptr<Signal>());
unsignaled = GetPrev(sem);
}
// NOTE: No else clause. Because if we didn't find it, and there's no previous, this indicates an error,
// but CoreChecks should have reported it
// If unsignaled is null, there was a missing pending semaphore, and that's also issue CoreChecks reports
return unsignaled;
}
void SignaledSemaphores::Resolve(SignaledSemaphores &parent, std::shared_ptr<QueueBatchContext> &last_batch) {
// Must only be called on child objects, with the non-const reference of the parent/previous object passed in
assert(prev_ == &parent);
// The global the semaphores we applied to the cmd_state QueueBatchContexts
// NOTE: All conserved QueueBatchContext's need to have there access logs reset to use the global logger and the only conserved
// QBC's are those referenced by unwaited signals and the last batch.
for (auto &sig_sem : signaled_) {
if (sig_sem.second && sig_sem.second->batch) {
auto &sig_batch = sig_sem.second->batch;
// Batches retained for signalled semaphore don't need to retain event data, unless it's the last batch in the submit
if (sig_batch != last_batch) {
sig_batch->ResetEventsContext();
// Make sure that retained batches are minimal, and trim after the events contexts has been cleared.
sig_batch->Trim();
}
}
// Import clears in the parent any signal waited in the
parent.Import(sig_sem.first, std::move(sig_sem.second));
}
Reset();
}
void SignaledSemaphores::Import(VkSemaphore sem, std::shared_ptr<Signal> &&from) {
// Overwrite the s tate with the last state from this
if (from) {
assert(sem == from->sem_state->semaphore());
signaled_[sem] = std::move(from);
} else {
signaled_.erase(sem);
}
}
void SignaledSemaphores::Reset() {
signaled_.clear();
prev_ = nullptr;
}
syncval_state::CommandBuffer::CommandBuffer(SyncValidator *dev, VkCommandBuffer cb, const VkCommandBufferAllocateInfo *pCreateInfo,
const COMMAND_POOL_STATE *pool)
: CMD_BUFFER_STATE(dev, cb, pCreateInfo, pool), access_context(*dev, this) {}
void syncval_state::CommandBuffer::Destroy() {
access_context.Destroy(); // must be first to clean up self references correctly.
CMD_BUFFER_STATE::Destroy();
}
void syncval_state::CommandBuffer::Reset() {
CMD_BUFFER_STATE::Reset();
access_context.Reset();
}
void syncval_state::CommandBuffer::NotifyInvalidate(const BASE_NODE::NodeList &invalid_nodes, bool unlink) {
for (auto &obj : invalid_nodes) {
switch (obj->Type()) {
case kVulkanObjectTypeEvent:
access_context.RecordDestroyEvent(static_cast<EVENT_STATE *>(obj.get()));
break;
default:
break;
}
CMD_BUFFER_STATE::NotifyInvalidate(invalid_nodes, unlink);
}
}
std::shared_ptr<CMD_BUFFER_STATE> SyncValidator::CreateCmdBufferState(VkCommandBuffer cb,
const VkCommandBufferAllocateInfo *pCreateInfo,
const COMMAND_POOL_STATE *cmd_pool) {
auto cb_state = std::make_shared<syncval_state::CommandBuffer>(this, cb, pCreateInfo, cmd_pool);
if (cb_state) {
cb_state->access_context.SetSelfReference();
}
return std::static_pointer_cast<CMD_BUFFER_STATE>(cb_state);
}
std::shared_ptr<SWAPCHAIN_NODE> SyncValidator::CreateSwapchainState(const VkSwapchainCreateInfoKHR *create_info,
VkSwapchainKHR swapchain) {
return std::static_pointer_cast<SWAPCHAIN_NODE>(std::make_shared<syncval_state::Swapchain>(this, create_info, swapchain));
}
std::shared_ptr<IMAGE_STATE> SyncValidator::CreateImageState(VkImage img, const VkImageCreateInfo *pCreateInfo,
VkFormatFeatureFlags2KHR features) {
return std::make_shared<ImageState>(this, img, pCreateInfo, features);
}
std::shared_ptr<IMAGE_STATE> SyncValidator::CreateImageState(VkImage img, const VkImageCreateInfo *pCreateInfo,
VkSwapchainKHR swapchain, uint32_t swapchain_index,
VkFormatFeatureFlags2KHR features) {
return std::make_shared<ImageState>(this, img, pCreateInfo, swapchain, swapchain_index, features);
}
std::shared_ptr<IMAGE_VIEW_STATE> SyncValidator::CreateImageViewState(
const std::shared_ptr<IMAGE_STATE> &image_state, VkImageView iv, const VkImageViewCreateInfo *ci, VkFormatFeatureFlags2KHR ff,
const VkFilterCubicImageViewImageFormatPropertiesEXT &cubic_props) {
return std::make_shared<ImageViewState>(image_state, iv, ci, ff, cubic_props);
}
bool SyncValidator::PreCallValidateCmdCopyBuffer(VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkBuffer dstBuffer,
uint32_t regionCount, const VkBufferCopy *pRegions,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
const auto *context = cb_context->GetCurrentAccessContext();
// If we have no previous accesses, we have no hazards
auto src_buffer = Get<BUFFER_STATE>(srcBuffer);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (src_buffer) {
const ResourceAccessRange src_range = MakeRange(*src_buffer, copy_region.srcOffset, copy_region.size);
auto hazard = context->DetectHazard(*src_buffer, SYNC_COPY_TRANSFER_READ, src_range);
if (hazard.IsHazard()) {
skip |= LogError(srcBuffer, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdCopyBuffer: Hazard %s for srcBuffer %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(srcBuffer).c_str(), region,
cb_context->FormatHazard(hazard).c_str());
}
}
if (dst_buffer && !skip) {
const ResourceAccessRange dst_range = MakeRange(*dst_buffer, copy_region.dstOffset, copy_region.size);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, dst_range);
if (hazard.IsHazard()) {
skip |= LogError(dstBuffer, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdCopyBuffer: Hazard %s for dstBuffer %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(dstBuffer).c_str(), region,
cb_context->FormatHazard(hazard).c_str());
}
}
if (skip) break;
}
return skip;
}
void SyncValidator::PreCallRecordCmdCopyBuffer(VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkBuffer dstBuffer,
uint32_t regionCount, const VkBufferCopy *pRegions) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
const auto tag = cb_context->NextCommandTag(Func::vkCmdCopyBuffer);
auto *context = cb_context->GetCurrentAccessContext();
auto src_buffer = Get<BUFFER_STATE>(srcBuffer);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (src_buffer) {
const ResourceAccessRange src_range = MakeRange(*src_buffer, copy_region.srcOffset, copy_region.size);
context->UpdateAccessState(*src_buffer, SYNC_COPY_TRANSFER_READ, SyncOrdering::kNonAttachment, src_range, tag);
}
if (dst_buffer) {
const ResourceAccessRange dst_range = MakeRange(*dst_buffer, copy_region.dstOffset, copy_region.size);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, dst_range, tag);
}
}
}
bool SyncValidator::PreCallValidateCmdCopyBuffer2(VkCommandBuffer commandBuffer, const VkCopyBufferInfo2 *pCopyBufferInfo,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
const auto *context = cb_context->GetCurrentAccessContext();
// If we have no previous accesses, we have no hazards
auto src_buffer = Get<BUFFER_STATE>(pCopyBufferInfo->srcBuffer);
auto dst_buffer = Get<BUFFER_STATE>(pCopyBufferInfo->dstBuffer);
for (uint32_t region = 0; region < pCopyBufferInfo->regionCount; region++) {
const auto &copy_region = pCopyBufferInfo->pRegions[region];
if (src_buffer) {
const ResourceAccessRange src_range = MakeRange(*src_buffer, copy_region.srcOffset, copy_region.size);
auto hazard = context->DetectHazard(*src_buffer, SYNC_COPY_TRANSFER_READ, src_range);
if (hazard.IsHazard()) {
// TODO -- add tag information to log msg when useful.
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), pCopyBufferInfo->srcBuffer, error_obj.location,
"Hazard %s for srcBuffer %s, region %" PRIu32 ". Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(pCopyBufferInfo->srcBuffer).c_str(), region, cb_context->FormatHazard(hazard).c_str());
}
}
if (dst_buffer && !skip) {
const ResourceAccessRange dst_range = MakeRange(*dst_buffer, copy_region.dstOffset, copy_region.size);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, dst_range);
if (hazard.IsHazard()) {
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), pCopyBufferInfo->dstBuffer, error_obj.location,
"Hazard %s for dstBuffer %s, region %" PRIu32 ". Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(pCopyBufferInfo->dstBuffer).c_str(), region, cb_context->FormatHazard(hazard).c_str());
}
}
if (skip) break;
}
return skip;
}
bool SyncValidator::PreCallValidateCmdCopyBuffer2KHR(VkCommandBuffer commandBuffer, const VkCopyBufferInfo2KHR *pCopyBufferInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdCopyBuffer2(commandBuffer, pCopyBufferInfo, error_obj);
}
void SyncValidator::RecordCmdCopyBuffer2(VkCommandBuffer commandBuffer, const VkCopyBufferInfo2KHR *pCopyBufferInfo, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
const auto tag = cb_context->NextCommandTag(command);
auto *context = cb_context->GetCurrentAccessContext();
auto src_buffer = Get<BUFFER_STATE>(pCopyBufferInfo->srcBuffer);
auto dst_buffer = Get<BUFFER_STATE>(pCopyBufferInfo->dstBuffer);
for (uint32_t region = 0; region < pCopyBufferInfo->regionCount; region++) {
const auto &copy_region = pCopyBufferInfo->pRegions[region];
if (src_buffer) {
const ResourceAccessRange src_range = MakeRange(*src_buffer, copy_region.srcOffset, copy_region.size);
context->UpdateAccessState(*src_buffer, SYNC_COPY_TRANSFER_READ, SyncOrdering::kNonAttachment, src_range, tag);
}
if (dst_buffer) {
const ResourceAccessRange dst_range = MakeRange(*dst_buffer, copy_region.dstOffset, copy_region.size);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, dst_range, tag);
}
}
}
void SyncValidator::PreCallRecordCmdCopyBuffer2KHR(VkCommandBuffer commandBuffer, const VkCopyBufferInfo2KHR *pCopyBufferInfo) {
RecordCmdCopyBuffer2(commandBuffer, pCopyBufferInfo, Func::vkCmdCopyBuffer2KHR);
}
void SyncValidator::PreCallRecordCmdCopyBuffer2(VkCommandBuffer commandBuffer, const VkCopyBufferInfo2 *pCopyBufferInfo) {
RecordCmdCopyBuffer2(commandBuffer, pCopyBufferInfo, Func::vkCmdCopyBuffer2);
}
bool SyncValidator::PreCallValidateCmdCopyImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const VkImageCopy *pRegions, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto src_image = Get<ImageState>(srcImage);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (src_image) {
auto hazard = context->DetectHazard(*src_image, SYNC_COPY_TRANSFER_READ, copy_region.srcSubresource,
copy_region.srcOffset, copy_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(srcImage, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdCopyImage: Hazard %s for srcImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(srcImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
}
if (dst_image) {
auto hazard = context->DetectHazard(*dst_image, SYNC_COPY_TRANSFER_WRITE, copy_region.dstSubresource,
copy_region.dstOffset, copy_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(dstImage, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdCopyImage: Hazard %s for dstImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(dstImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
if (skip) break;
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdCopyImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const VkImageCopy *pRegions) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdCopyImage);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto src_image = Get<ImageState>(srcImage);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (src_image) {
context->UpdateAccessState(*src_image, SYNC_COPY_TRANSFER_READ, SyncOrdering::kNonAttachment,
copy_region.srcSubresource, copy_region.srcOffset, copy_region.extent, tag);
}
if (dst_image) {
context->UpdateAccessState(*dst_image, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment,
copy_region.dstSubresource, copy_region.dstOffset, copy_region.extent, tag);
}
}
}
bool SyncValidator::PreCallValidateCmdCopyImage2(VkCommandBuffer commandBuffer, const VkCopyImageInfo2 *pCopyImageInfo,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto src_image = Get<ImageState>(pCopyImageInfo->srcImage);
auto dst_image = Get<ImageState>(pCopyImageInfo->dstImage);
for (uint32_t region = 0; region < pCopyImageInfo->regionCount; region++) {
const auto &copy_region = pCopyImageInfo->pRegions[region];
if (src_image) {
auto hazard = context->DetectHazard(*src_image, SYNC_COPY_TRANSFER_READ, copy_region.srcSubresource,
copy_region.srcOffset, copy_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), pCopyImageInfo->srcImage, error_obj.location,
"Hazard %s for srcImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(pCopyImageInfo->srcImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
}
if (dst_image) {
auto hazard = context->DetectHazard(*dst_image, SYNC_COPY_TRANSFER_WRITE, copy_region.dstSubresource,
copy_region.dstOffset, copy_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), pCopyImageInfo->dstImage, error_obj.location,
"Hazard %s for dstImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(pCopyImageInfo->dstImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
if (skip) break;
}
}
return skip;
}
bool SyncValidator::PreCallValidateCmdCopyImage2KHR(VkCommandBuffer commandBuffer, const VkCopyImageInfo2KHR *pCopyImageInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdCopyImage2(commandBuffer, pCopyImageInfo, error_obj);
}
void SyncValidator::RecordCmdCopyImage2(VkCommandBuffer commandBuffer, const VkCopyImageInfo2KHR *pCopyImageInfo, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(command);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto src_image = Get<ImageState>(pCopyImageInfo->srcImage);
auto dst_image = Get<ImageState>(pCopyImageInfo->dstImage);
for (uint32_t region = 0; region < pCopyImageInfo->regionCount; region++) {
const auto &copy_region = pCopyImageInfo->pRegions[region];
if (src_image) {
context->UpdateAccessState(*src_image, SYNC_COPY_TRANSFER_READ, SyncOrdering::kNonAttachment,
copy_region.srcSubresource, copy_region.srcOffset, copy_region.extent, tag);
}
if (dst_image) {
context->UpdateAccessState(*dst_image, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment,
copy_region.dstSubresource, copy_region.dstOffset, copy_region.extent, tag);
}
}
}
void SyncValidator::PreCallRecordCmdCopyImage2KHR(VkCommandBuffer commandBuffer, const VkCopyImageInfo2KHR *pCopyImageInfo) {
RecordCmdCopyImage2(commandBuffer, pCopyImageInfo, Func::vkCmdCopyImage2KHR);
}
void SyncValidator::PreCallRecordCmdCopyImage2(VkCommandBuffer commandBuffer, const VkCopyImageInfo2 *pCopyImageInfo) {
RecordCmdCopyImage2(commandBuffer, pCopyImageInfo, Func::vkCmdCopyImage2);
}
bool SyncValidator::PreCallValidateCmdPipelineBarrier(
VkCommandBuffer commandBuffer, VkPipelineStageFlags srcStageMask, VkPipelineStageFlags dstStageMask,
VkDependencyFlags dependencyFlags, uint32_t memoryBarrierCount, const VkMemoryBarrier *pMemoryBarriers,
uint32_t bufferMemoryBarrierCount, const VkBufferMemoryBarrier *pBufferMemoryBarriers, uint32_t imageMemoryBarrierCount,
const VkImageMemoryBarrier *pImageMemoryBarriers, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
SyncOpPipelineBarrier pipeline_barrier(error_obj.location.function, *this, cb_access_context->GetQueueFlags(), srcStageMask,
dstStageMask, dependencyFlags, memoryBarrierCount, pMemoryBarriers,
bufferMemoryBarrierCount, pBufferMemoryBarriers, imageMemoryBarrierCount,
pImageMemoryBarriers);
skip = pipeline_barrier.Validate(*cb_access_context);
return skip;
}
void SyncValidator::PreCallRecordCmdPipelineBarrier(VkCommandBuffer commandBuffer, VkPipelineStageFlags srcStageMask,
VkPipelineStageFlags dstStageMask, VkDependencyFlags dependencyFlags,
uint32_t memoryBarrierCount, const VkMemoryBarrier *pMemoryBarriers,
uint32_t bufferMemoryBarrierCount,
const VkBufferMemoryBarrier *pBufferMemoryBarriers,
uint32_t imageMemoryBarrierCount,
const VkImageMemoryBarrier *pImageMemoryBarriers) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
cb_access_context->RecordSyncOp<SyncOpPipelineBarrier>(Func::vkCmdPipelineBarrier, *this, cb_access_context->GetQueueFlags(),
srcStageMask, dstStageMask, dependencyFlags, memoryBarrierCount,
pMemoryBarriers, bufferMemoryBarrierCount, pBufferMemoryBarriers,
imageMemoryBarrierCount, pImageMemoryBarriers);
}
bool SyncValidator::PreCallValidateCmdPipelineBarrier2KHR(VkCommandBuffer commandBuffer, const VkDependencyInfoKHR *pDependencyInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdPipelineBarrier2(commandBuffer, pDependencyInfo, error_obj);
}
bool SyncValidator::PreCallValidateCmdPipelineBarrier2(VkCommandBuffer commandBuffer, const VkDependencyInfo *pDependencyInfo,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
SyncOpPipelineBarrier pipeline_barrier(error_obj.location.function, *this, cb_access_context->GetQueueFlags(),
*pDependencyInfo);
skip = pipeline_barrier.Validate(*cb_access_context);
return skip;
}
void SyncValidator::PreCallRecordCmdPipelineBarrier2KHR(VkCommandBuffer commandBuffer, const VkDependencyInfoKHR *pDependencyInfo) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
cb_access_context->RecordSyncOp<SyncOpPipelineBarrier>(Func::vkCmdPipelineBarrier2KHR, *this,
cb_access_context->GetQueueFlags(), *pDependencyInfo);
}
void SyncValidator::PreCallRecordCmdPipelineBarrier2(VkCommandBuffer commandBuffer, const VkDependencyInfo *pDependencyInfo) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
cb_access_context->RecordSyncOp<SyncOpPipelineBarrier>(Func::vkCmdPipelineBarrier2, *this, cb_access_context->GetQueueFlags(),
*pDependencyInfo);
}
void SyncValidator::CreateDevice(const VkDeviceCreateInfo *pCreateInfo) {
// The state tracker sets up the device state
StateTracker::CreateDevice(pCreateInfo);
ForEachShared<QUEUE_STATE>([this](const std::shared_ptr<QUEUE_STATE> &queue_state) {
auto queue_flags = physical_device_state->queue_family_properties[queue_state->queueFamilyIndex].queueFlags;
std::shared_ptr<QueueSyncState> queue_sync_state =
std::make_shared<QueueSyncState>(queue_state, queue_flags, queue_id_limit_++);
queue_sync_states_.emplace(std::make_pair(queue_state->Queue(), std::move(queue_sync_state)));
});
}
bool SyncValidator::ValidateBeginRenderPass(VkCommandBuffer commandBuffer, const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
if (cb_state) {
SyncOpBeginRenderPass sync_op(error_obj.location.function, *this, pRenderPassBegin, pSubpassBeginInfo);
skip = sync_op.Validate(cb_state->access_context);
}
return skip;
}
bool SyncValidator::PreCallValidateCmdBeginRenderPass(VkCommandBuffer commandBuffer, const VkRenderPassBeginInfo *pRenderPassBegin,
VkSubpassContents contents, const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdBeginRenderPass(commandBuffer, pRenderPassBegin, contents, error_obj);
VkSubpassBeginInfo subpass_begin_info = vku::InitStructHelper();
subpass_begin_info.contents = contents;
skip |= ValidateBeginRenderPass(commandBuffer, pRenderPassBegin, &subpass_begin_info, error_obj);
return skip;
}
bool SyncValidator::PreCallValidateCmdBeginRenderPass2(VkCommandBuffer commandBuffer, const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo,
const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdBeginRenderPass2(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, error_obj);
skip |= ValidateBeginRenderPass(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, error_obj);
return skip;
}
bool SyncValidator::PreCallValidateCmdBeginRenderPass2KHR(VkCommandBuffer commandBuffer,
const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdBeginRenderPass2(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, error_obj);
}
void SyncValidator::PostCallRecordBeginCommandBuffer(VkCommandBuffer commandBuffer, const VkCommandBufferBeginInfo *pBeginInfo,
const RecordObject &record_obj) {
// The state tracker sets up the command buffer state
StateTracker::PostCallRecordBeginCommandBuffer(commandBuffer, pBeginInfo, record_obj);
// Create/initialize the structure that trackers accesses at the command buffer scope.
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
cb_state->access_context.Reset();
}
void SyncValidator::RecordCmdBeginRenderPass(VkCommandBuffer commandBuffer, const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
if (cb_state) {
cb_state->access_context.RecordSyncOp<SyncOpBeginRenderPass>(command, *this, pRenderPassBegin, pSubpassBeginInfo);
}
}
void SyncValidator::PostCallRecordCmdBeginRenderPass(VkCommandBuffer commandBuffer, const VkRenderPassBeginInfo *pRenderPassBegin,
VkSubpassContents contents, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdBeginRenderPass(commandBuffer, pRenderPassBegin, contents, record_obj);
VkSubpassBeginInfo subpass_begin_info = vku::InitStructHelper();
subpass_begin_info.contents = contents;
RecordCmdBeginRenderPass(commandBuffer, pRenderPassBegin, &subpass_begin_info, record_obj.location.function);
}
void SyncValidator::PostCallRecordCmdBeginRenderPass2(VkCommandBuffer commandBuffer, const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdBeginRenderPass2(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, record_obj);
RecordCmdBeginRenderPass(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, record_obj.location.function);
}
void SyncValidator::PostCallRecordCmdBeginRenderPass2KHR(VkCommandBuffer commandBuffer,
const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo,
const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdBeginRenderPass2KHR(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, record_obj);
RecordCmdBeginRenderPass(commandBuffer, pRenderPassBegin, pSubpassBeginInfo, record_obj.location.function);
}
bool SyncValidator::ValidateCmdNextSubpass(VkCommandBuffer commandBuffer, const VkSubpassBeginInfo *pSubpassBeginInfo,
const VkSubpassEndInfo *pSubpassEndInfo, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
SyncOpNextSubpass sync_op(error_obj.location.function, *this, pSubpassBeginInfo, pSubpassEndInfo);
return sync_op.Validate(*cb_context);
}
bool SyncValidator::PreCallValidateCmdNextSubpass(VkCommandBuffer commandBuffer, VkSubpassContents contents,
const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdNextSubpass(commandBuffer, contents, error_obj);
// Convert to a NextSubpass2
VkSubpassBeginInfo subpass_begin_info = vku::InitStructHelper();
subpass_begin_info.contents = contents;
VkSubpassEndInfo subpass_end_info = vku::InitStructHelper();
skip |= ValidateCmdNextSubpass(commandBuffer, &subpass_begin_info, &subpass_end_info, error_obj);
return skip;
}
bool SyncValidator::PreCallValidateCmdNextSubpass2KHR(VkCommandBuffer commandBuffer, const VkSubpassBeginInfo *pSubpassBeginInfo,
const VkSubpassEndInfo *pSubpassEndInfo, const ErrorObject &error_obj) const {
return PreCallValidateCmdNextSubpass2(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, error_obj);
}
bool SyncValidator::PreCallValidateCmdNextSubpass2(VkCommandBuffer commandBuffer, const VkSubpassBeginInfo *pSubpassBeginInfo,
const VkSubpassEndInfo *pSubpassEndInfo, const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdNextSubpass2(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, error_obj);
skip |= ValidateCmdNextSubpass(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, error_obj);
return skip;
}
void SyncValidator::RecordCmdNextSubpass(VkCommandBuffer commandBuffer, const VkSubpassBeginInfo *pSubpassBeginInfo,
const VkSubpassEndInfo *pSubpassEndInfo, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpNextSubpass>(command, *this, pSubpassBeginInfo, pSubpassEndInfo);
}
void SyncValidator::PostCallRecordCmdNextSubpass(VkCommandBuffer commandBuffer, VkSubpassContents contents,
const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdNextSubpass(commandBuffer, contents, record_obj);
VkSubpassBeginInfo subpass_begin_info = vku::InitStructHelper();
subpass_begin_info.contents = contents;
RecordCmdNextSubpass(commandBuffer, &subpass_begin_info, nullptr, record_obj.location.function);
}
void SyncValidator::PostCallRecordCmdNextSubpass2(VkCommandBuffer commandBuffer, const VkSubpassBeginInfo *pSubpassBeginInfo,
const VkSubpassEndInfo *pSubpassEndInfo, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdNextSubpass2(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, record_obj);
RecordCmdNextSubpass(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, record_obj.location.function);
}
void SyncValidator::PostCallRecordCmdNextSubpass2KHR(VkCommandBuffer commandBuffer, const VkSubpassBeginInfo *pSubpassBeginInfo,
const VkSubpassEndInfo *pSubpassEndInfo, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdNextSubpass2KHR(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, record_obj);
RecordCmdNextSubpass(commandBuffer, pSubpassBeginInfo, pSubpassEndInfo, record_obj.location.function);
}
bool SyncValidator::ValidateCmdEndRenderPass(VkCommandBuffer commandBuffer, const VkSubpassEndInfo *pSubpassEndInfo,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
auto *cb_context = &cb_state->access_context;
SyncOpEndRenderPass sync_op(error_obj.location.function, *this, pSubpassEndInfo);
skip |= sync_op.Validate(*cb_context);
return skip;
}
bool SyncValidator::PreCallValidateCmdEndRenderPass(VkCommandBuffer commandBuffer, const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdEndRenderPass(commandBuffer, error_obj);
skip |= ValidateCmdEndRenderPass(commandBuffer, nullptr, error_obj);
return skip;
}
bool SyncValidator::PreCallValidateCmdEndRenderPass2(VkCommandBuffer commandBuffer, const VkSubpassEndInfo *pSubpassEndInfo,
const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdEndRenderPass2(commandBuffer, pSubpassEndInfo, error_obj);
skip |= ValidateCmdEndRenderPass(commandBuffer, pSubpassEndInfo, error_obj);
return skip;
}
bool SyncValidator::PreCallValidateCmdEndRenderPass2KHR(VkCommandBuffer commandBuffer, const VkSubpassEndInfo *pSubpassEndInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdEndRenderPass2(commandBuffer, pSubpassEndInfo, error_obj);
}
void SyncValidator::RecordCmdEndRenderPass(VkCommandBuffer commandBuffer, const VkSubpassEndInfo *pSubpassEndInfo, Func command) {
// Resolve the all subpass contexts to the command buffer contexts
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpEndRenderPass>(command, *this, pSubpassEndInfo);
}
// Simple heuristic rule to detect WAW operations representing algorithmically safe or increment
// updates to a resource which do not conflict at the byte level.
// TODO: Revisit this rule to see if it needs to be tighter or looser
// TODO: Add programatic control over suppression heuristics
bool SyncValidator::SupressedBoundDescriptorWAW(const HazardResult &hazard) const {
assert(hazard.IsHazard());
return hazard.IsWAWHazard();
}
void SyncValidator::PostCallRecordCmdEndRenderPass(VkCommandBuffer commandBuffer, const RecordObject &record_obj) {
RecordCmdEndRenderPass(commandBuffer, nullptr, record_obj.location.function);
StateTracker::PostCallRecordCmdEndRenderPass(commandBuffer, record_obj);
}
void SyncValidator::PostCallRecordCmdEndRenderPass2(VkCommandBuffer commandBuffer, const VkSubpassEndInfo *pSubpassEndInfo,
const RecordObject &record_obj) {
RecordCmdEndRenderPass(commandBuffer, pSubpassEndInfo, record_obj.location.function);
StateTracker::PostCallRecordCmdEndRenderPass2(commandBuffer, pSubpassEndInfo, record_obj);
}
void SyncValidator::PostCallRecordCmdEndRenderPass2KHR(VkCommandBuffer commandBuffer, const VkSubpassEndInfo *pSubpassEndInfo,
const RecordObject &record_obj) {
RecordCmdEndRenderPass(commandBuffer, pSubpassEndInfo, record_obj.location.function);
StateTracker::PostCallRecordCmdEndRenderPass2KHR(commandBuffer, pSubpassEndInfo, record_obj);
}
template <typename RegionType>
bool SyncValidator::ValidateCmdCopyBufferToImage(VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkImage dstImage,
VkImageLayout dstImageLayout, uint32_t regionCount, const RegionType *pRegions,
const Location &loc) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto src_buffer = Get<BUFFER_STATE>(srcBuffer);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
HazardResult hazard;
if (dst_image) {
if (src_buffer) {
ResourceAccessRange src_range = MakeRange(
copy_region.bufferOffset,
GetBufferSizeFromCopyImage(copy_region, dst_image->createInfo.format, dst_image->createInfo.arrayLayers));
hazard = context->DetectHazard(*src_buffer, SYNC_COPY_TRANSFER_READ, src_range);
if (hazard.IsHazard()) {
// PHASE1 TODO -- add tag information to log msg when useful.
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), srcBuffer, loc,
"Hazard %s for srcBuffer %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(srcBuffer).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
}
hazard = context->DetectHazard(*dst_image, SYNC_COPY_TRANSFER_WRITE, copy_region.imageSubresource,
copy_region.imageOffset, copy_region.imageExtent, false);
if (hazard.IsHazard()) {
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), dstImage, loc,
"Hazard %s for dstImage %s, region %" PRIu32 ". Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(dstImage).c_str(), region, cb_access_context->FormatHazard(hazard).c_str());
}
if (skip) break;
}
if (skip) break;
}
return skip;
}
bool SyncValidator::PreCallValidateCmdCopyBufferToImage(VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkImage dstImage,
VkImageLayout dstImageLayout, uint32_t regionCount,
const VkBufferImageCopy *pRegions, const ErrorObject &error_obj) const {
return ValidateCmdCopyBufferToImage(commandBuffer, srcBuffer, dstImage, dstImageLayout, regionCount, pRegions,
error_obj.location);
}
bool SyncValidator::PreCallValidateCmdCopyBufferToImage2KHR(VkCommandBuffer commandBuffer,
const VkCopyBufferToImageInfo2KHR *pCopyBufferToImageInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdCopyBufferToImage2(commandBuffer, pCopyBufferToImageInfo, error_obj);
}
bool SyncValidator::PreCallValidateCmdCopyBufferToImage2(VkCommandBuffer commandBuffer,
const VkCopyBufferToImageInfo2 *pCopyBufferToImageInfo,
const ErrorObject &error_obj) const {
return ValidateCmdCopyBufferToImage(commandBuffer, pCopyBufferToImageInfo->srcBuffer, pCopyBufferToImageInfo->dstImage,
pCopyBufferToImageInfo->dstImageLayout, pCopyBufferToImageInfo->regionCount,
pCopyBufferToImageInfo->pRegions, error_obj.location.dot(Field::pCopyBufferToImageInfo));
}
template <typename RegionType>
void SyncValidator::RecordCmdCopyBufferToImage(VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkImage dstImage,
VkImageLayout dstImageLayout, uint32_t regionCount, const RegionType *pRegions,
Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(command);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto src_buffer = Get<BUFFER_STATE>(srcBuffer);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (dst_image) {
if (src_buffer) {
ResourceAccessRange src_range = MakeRange(
copy_region.bufferOffset,
GetBufferSizeFromCopyImage(copy_region, dst_image->createInfo.format, dst_image->createInfo.arrayLayers));
context->UpdateAccessState(*src_buffer, SYNC_COPY_TRANSFER_READ, SyncOrdering::kNonAttachment, src_range, tag);
}
context->UpdateAccessState(*dst_image, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment,
copy_region.imageSubresource, copy_region.imageOffset, copy_region.imageExtent, tag);
}
}
}
void SyncValidator::PreCallRecordCmdCopyBufferToImage(VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkImage dstImage,
VkImageLayout dstImageLayout, uint32_t regionCount,
const VkBufferImageCopy *pRegions) {
StateTracker::PreCallRecordCmdCopyBufferToImage(commandBuffer, srcBuffer, dstImage, dstImageLayout, regionCount, pRegions);
RecordCmdCopyBufferToImage(commandBuffer, srcBuffer, dstImage, dstImageLayout, regionCount, pRegions,
Func::vkCmdCopyBufferToImage);
}
void SyncValidator::PreCallRecordCmdCopyBufferToImage2KHR(VkCommandBuffer commandBuffer,
const VkCopyBufferToImageInfo2KHR *pCopyBufferToImageInfo) {
StateTracker::PreCallRecordCmdCopyBufferToImage2KHR(commandBuffer, pCopyBufferToImageInfo);
RecordCmdCopyBufferToImage(commandBuffer, pCopyBufferToImageInfo->srcBuffer, pCopyBufferToImageInfo->dstImage,
pCopyBufferToImageInfo->dstImageLayout, pCopyBufferToImageInfo->regionCount,
pCopyBufferToImageInfo->pRegions, Func::vkCmdCopyBufferToImage2KHR);
}
void SyncValidator::PreCallRecordCmdCopyBufferToImage2(VkCommandBuffer commandBuffer,
const VkCopyBufferToImageInfo2 *pCopyBufferToImageInfo) {
StateTracker::PreCallRecordCmdCopyBufferToImage2(commandBuffer, pCopyBufferToImageInfo);
RecordCmdCopyBufferToImage(commandBuffer, pCopyBufferToImageInfo->srcBuffer, pCopyBufferToImageInfo->dstImage,
pCopyBufferToImageInfo->dstImageLayout, pCopyBufferToImageInfo->regionCount,
pCopyBufferToImageInfo->pRegions, Func::vkCmdCopyBufferToImage2);
}
template <typename RegionType>
bool SyncValidator::ValidateCmdCopyImageToBuffer(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkBuffer dstBuffer, uint32_t regionCount, const RegionType *pRegions,
const Location &loc) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto src_image = Get<ImageState>(srcImage);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
const auto dst_mem = (dst_buffer && !dst_buffer->sparse) ? dst_buffer->MemState()->deviceMemory() : VK_NULL_HANDLE;
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (src_image) {
auto hazard = context->DetectHazard(*src_image, SYNC_COPY_TRANSFER_READ, copy_region.imageSubresource,
copy_region.imageOffset, copy_region.imageExtent, false);
if (hazard.IsHazard()) {
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), srcImage, loc,
"Hazard %s for srcImage %s, region %" PRIu32 ". Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(srcImage).c_str(), region, cb_access_context->FormatHazard(hazard).c_str());
}
if (dst_mem) {
ResourceAccessRange dst_range = MakeRange(
copy_region.bufferOffset,
GetBufferSizeFromCopyImage(copy_region, src_image->createInfo.format, src_image->createInfo.arrayLayers));
hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, dst_range);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), dstBuffer, loc,
"Hazard %s for dstBuffer %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(dstBuffer).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
}
}
if (skip) break;
}
return skip;
}
bool SyncValidator::PreCallValidateCmdCopyImageToBuffer(VkCommandBuffer commandBuffer, VkImage srcImage,
VkImageLayout srcImageLayout, VkBuffer dstBuffer, uint32_t regionCount,
const VkBufferImageCopy *pRegions, const ErrorObject &error_obj) const {
return ValidateCmdCopyImageToBuffer(commandBuffer, srcImage, srcImageLayout, dstBuffer, regionCount, pRegions,
error_obj.location);
}
bool SyncValidator::PreCallValidateCmdCopyImageToBuffer2KHR(VkCommandBuffer commandBuffer,
const VkCopyImageToBufferInfo2KHR *pCopyImageToBufferInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdCopyImageToBuffer2(commandBuffer, pCopyImageToBufferInfo, error_obj);
}
bool SyncValidator::PreCallValidateCmdCopyImageToBuffer2(VkCommandBuffer commandBuffer,
const VkCopyImageToBufferInfo2 *pCopyImageToBufferInfo,
const ErrorObject &error_obj) const {
return ValidateCmdCopyImageToBuffer(commandBuffer, pCopyImageToBufferInfo->srcImage, pCopyImageToBufferInfo->srcImageLayout,
pCopyImageToBufferInfo->dstBuffer, pCopyImageToBufferInfo->regionCount,
pCopyImageToBufferInfo->pRegions, error_obj.location.dot(Field::pCopyImageToBufferInfo));
}
template <typename RegionType>
void SyncValidator::RecordCmdCopyImageToBuffer(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkBuffer dstBuffer, uint32_t regionCount, const RegionType *pRegions, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(command);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto src_image = Get<ImageState>(srcImage);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
const auto dst_mem = (dst_buffer && !dst_buffer->sparse) ? dst_buffer->MemState()->deviceMemory() : VK_NULL_HANDLE;
const VulkanTypedHandle dst_handle(dst_mem, kVulkanObjectTypeDeviceMemory);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &copy_region = pRegions[region];
if (src_image) {
context->UpdateAccessState(*src_image, SYNC_COPY_TRANSFER_READ, SyncOrdering::kNonAttachment,
copy_region.imageSubresource, copy_region.imageOffset, copy_region.imageExtent, tag);
if (dst_buffer) {
ResourceAccessRange dst_range = MakeRange(
copy_region.bufferOffset,
GetBufferSizeFromCopyImage(copy_region, src_image->createInfo.format, src_image->createInfo.arrayLayers));
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, dst_range, tag);
}
}
}
}
void SyncValidator::PreCallRecordCmdCopyImageToBuffer(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkBuffer dstBuffer, uint32_t regionCount, const VkBufferImageCopy *pRegions) {
StateTracker::PreCallRecordCmdCopyImageToBuffer(commandBuffer, srcImage, srcImageLayout, dstBuffer, regionCount, pRegions);
RecordCmdCopyImageToBuffer(commandBuffer, srcImage, srcImageLayout, dstBuffer, regionCount, pRegions,
Func::vkCmdCopyImageToBuffer);
}
void SyncValidator::PreCallRecordCmdCopyImageToBuffer2KHR(VkCommandBuffer commandBuffer,
const VkCopyImageToBufferInfo2KHR *pCopyImageToBufferInfo) {
StateTracker::PreCallRecordCmdCopyImageToBuffer2KHR(commandBuffer, pCopyImageToBufferInfo);
RecordCmdCopyImageToBuffer(commandBuffer, pCopyImageToBufferInfo->srcImage, pCopyImageToBufferInfo->srcImageLayout,
pCopyImageToBufferInfo->dstBuffer, pCopyImageToBufferInfo->regionCount,
pCopyImageToBufferInfo->pRegions, Func::vkCmdCopyImageToBuffer2KHR);
}
void SyncValidator::PreCallRecordCmdCopyImageToBuffer2(VkCommandBuffer commandBuffer,
const VkCopyImageToBufferInfo2 *pCopyImageToBufferInfo) {
StateTracker::PreCallRecordCmdCopyImageToBuffer2(commandBuffer, pCopyImageToBufferInfo);
RecordCmdCopyImageToBuffer(commandBuffer, pCopyImageToBufferInfo->srcImage, pCopyImageToBufferInfo->srcImageLayout,
pCopyImageToBufferInfo->dstBuffer, pCopyImageToBufferInfo->regionCount,
pCopyImageToBufferInfo->pRegions, Func::vkCmdCopyImageToBuffer2);
}
template <typename RegionType>
bool SyncValidator::ValidateCmdBlitImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const RegionType *pRegions, VkFilter filter, const Location &loc) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto src_image = Get<ImageState>(srcImage);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &blit_region = pRegions[region];
if (src_image) {
VkOffset3D offset = {std::min(blit_region.srcOffsets[0].x, blit_region.srcOffsets[1].x),
std::min(blit_region.srcOffsets[0].y, blit_region.srcOffsets[1].y),
std::min(blit_region.srcOffsets[0].z, blit_region.srcOffsets[1].z)};
VkExtent3D extent = {static_cast<uint32_t>(abs(blit_region.srcOffsets[1].x - blit_region.srcOffsets[0].x)),
static_cast<uint32_t>(abs(blit_region.srcOffsets[1].y - blit_region.srcOffsets[0].y)),
static_cast<uint32_t>(abs(blit_region.srcOffsets[1].z - blit_region.srcOffsets[0].z))};
auto hazard =
context->DetectHazard(*src_image, SYNC_BLIT_TRANSFER_READ, blit_region.srcSubresource, offset, extent, false);
if (hazard.IsHazard()) {
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), srcImage, loc,
"Hazard %s for srcImage %s, region %" PRIu32 ". Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(srcImage).c_str(), region, cb_access_context->FormatHazard(hazard).c_str());
}
}
if (dst_image) {
VkOffset3D offset = {std::min(blit_region.dstOffsets[0].x, blit_region.dstOffsets[1].x),
std::min(blit_region.dstOffsets[0].y, blit_region.dstOffsets[1].y),
std::min(blit_region.dstOffsets[0].z, blit_region.dstOffsets[1].z)};
VkExtent3D extent = {static_cast<uint32_t>(abs(blit_region.dstOffsets[1].x - blit_region.dstOffsets[0].x)),
static_cast<uint32_t>(abs(blit_region.dstOffsets[1].y - blit_region.dstOffsets[0].y)),
static_cast<uint32_t>(abs(blit_region.dstOffsets[1].z - blit_region.dstOffsets[0].z))};
auto hazard =
context->DetectHazard(*dst_image, SYNC_BLIT_TRANSFER_WRITE, blit_region.dstSubresource, offset, extent, false);
if (hazard.IsHazard()) {
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), dstImage, loc,
"Hazard %s for dstImage %s, region %" PRIu32 ". Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(dstImage).c_str(), region, cb_access_context->FormatHazard(hazard).c_str());
}
if (skip) break;
}
}
return skip;
}
bool SyncValidator::PreCallValidateCmdBlitImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const VkImageBlit *pRegions, VkFilter filter, const ErrorObject &error_obj) const {
return ValidateCmdBlitImage(commandBuffer, srcImage, srcImageLayout, dstImage, dstImageLayout, regionCount, pRegions, filter,
error_obj.location);
}
bool SyncValidator::PreCallValidateCmdBlitImage2KHR(VkCommandBuffer commandBuffer, const VkBlitImageInfo2KHR *pBlitImageInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdBlitImage2(commandBuffer, pBlitImageInfo, error_obj);
}
bool SyncValidator::PreCallValidateCmdBlitImage2(VkCommandBuffer commandBuffer, const VkBlitImageInfo2 *pBlitImageInfo,
const ErrorObject &error_obj) const {
return ValidateCmdBlitImage(commandBuffer, pBlitImageInfo->srcImage, pBlitImageInfo->srcImageLayout, pBlitImageInfo->dstImage,
pBlitImageInfo->dstImageLayout, pBlitImageInfo->regionCount, pBlitImageInfo->pRegions,
pBlitImageInfo->filter, error_obj.location.dot(Field::pBlitImageInfo));
}
template <typename RegionType>
void SyncValidator::RecordCmdBlitImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const RegionType *pRegions, VkFilter filter, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
const auto tag = cb_state->access_context.NextCommandTag(command);
auto *context = cb_state->access_context.GetCurrentAccessContext();
assert(context);
auto src_image = Get<ImageState>(srcImage);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &blit_region = pRegions[region];
if (src_image) {
VkOffset3D offset = {std::min(blit_region.srcOffsets[0].x, blit_region.srcOffsets[1].x),
std::min(blit_region.srcOffsets[0].y, blit_region.srcOffsets[1].y),
std::min(blit_region.srcOffsets[0].z, blit_region.srcOffsets[1].z)};
VkExtent3D extent = {static_cast<uint32_t>(abs(blit_region.srcOffsets[1].x - blit_region.srcOffsets[0].x)),
static_cast<uint32_t>(abs(blit_region.srcOffsets[1].y - blit_region.srcOffsets[0].y)),
static_cast<uint32_t>(abs(blit_region.srcOffsets[1].z - blit_region.srcOffsets[0].z))};
context->UpdateAccessState(*src_image, SYNC_BLIT_TRANSFER_READ, SyncOrdering::kNonAttachment,
blit_region.srcSubresource, offset, extent, tag);
}
if (dst_image) {
VkOffset3D offset = {std::min(blit_region.dstOffsets[0].x, blit_region.dstOffsets[1].x),
std::min(blit_region.dstOffsets[0].y, blit_region.dstOffsets[1].y),
std::min(blit_region.dstOffsets[0].z, blit_region.dstOffsets[1].z)};
VkExtent3D extent = {static_cast<uint32_t>(abs(blit_region.dstOffsets[1].x - blit_region.dstOffsets[0].x)),
static_cast<uint32_t>(abs(blit_region.dstOffsets[1].y - blit_region.dstOffsets[0].y)),
static_cast<uint32_t>(abs(blit_region.dstOffsets[1].z - blit_region.dstOffsets[0].z))};
context->UpdateAccessState(*dst_image, SYNC_BLIT_TRANSFER_WRITE, SyncOrdering::kNonAttachment,
blit_region.dstSubresource, offset, extent, tag);
}
}
}
void SyncValidator::PreCallRecordCmdBlitImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const VkImageBlit *pRegions, VkFilter filter) {
StateTracker::PreCallRecordCmdBlitImage(commandBuffer, srcImage, srcImageLayout, dstImage, dstImageLayout, regionCount,
pRegions, filter);
RecordCmdBlitImage(commandBuffer, srcImage, srcImageLayout, dstImage, dstImageLayout, regionCount, pRegions, filter,
Func::vkCmdBlitImage);
}
void SyncValidator::PreCallRecordCmdBlitImage2KHR(VkCommandBuffer commandBuffer, const VkBlitImageInfo2KHR *pBlitImageInfo) {
StateTracker::PreCallRecordCmdBlitImage2KHR(commandBuffer, pBlitImageInfo);
RecordCmdBlitImage(commandBuffer, pBlitImageInfo->srcImage, pBlitImageInfo->srcImageLayout, pBlitImageInfo->dstImage,
pBlitImageInfo->dstImageLayout, pBlitImageInfo->regionCount, pBlitImageInfo->pRegions,
pBlitImageInfo->filter, Func::vkCmdBlitImage2KHR);
}
void SyncValidator::PreCallRecordCmdBlitImage2(VkCommandBuffer commandBuffer, const VkBlitImageInfo2 *pBlitImageInfo) {
StateTracker::PreCallRecordCmdBlitImage2KHR(commandBuffer, pBlitImageInfo);
RecordCmdBlitImage(commandBuffer, pBlitImageInfo->srcImage, pBlitImageInfo->srcImageLayout, pBlitImageInfo->dstImage,
pBlitImageInfo->dstImageLayout, pBlitImageInfo->regionCount, pBlitImageInfo->pRegions,
pBlitImageInfo->filter, Func::vkCmdBlitImage2);
}
bool SyncValidator::ValidateIndirectBuffer(const CommandBufferAccessContext &cb_context, const AccessContext &context,
VkCommandBuffer commandBuffer, const VkDeviceSize struct_size, const VkBuffer buffer,
const VkDeviceSize offset, const uint32_t drawCount, const uint32_t stride,
const Location &loc) const {
bool skip = false;
if (drawCount == 0) return skip;
auto buf_state = Get<BUFFER_STATE>(buffer);
VkDeviceSize size = struct_size;
if (drawCount == 1 || stride == size) {
if (drawCount > 1) size *= drawCount;
const ResourceAccessRange range = MakeRange(offset, size);
auto hazard = context.DetectHazard(*buf_state, SYNC_DRAW_INDIRECT_INDIRECT_COMMAND_READ, range);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), buf_state->buffer(), loc,
"Hazard %s for indirect %s in %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(buffer).c_str(), FormatHandle(commandBuffer).c_str(),
cb_context.FormatHazard(hazard).c_str());
}
} else {
for (uint32_t i = 0; i < drawCount; ++i) {
const ResourceAccessRange range = MakeRange(offset + i * stride, size);
auto hazard = context.DetectHazard(*buf_state, SYNC_DRAW_INDIRECT_INDIRECT_COMMAND_READ, range);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), buf_state->buffer(), loc,
"Hazard %s for indirect %s in %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(buffer).c_str(), FormatHandle(commandBuffer).c_str(),
cb_context.FormatHazard(hazard).c_str());
break;
}
}
}
return skip;
}
void SyncValidator::RecordIndirectBuffer(AccessContext &context, const ResourceUsageTag tag, const VkDeviceSize struct_size,
const VkBuffer buffer, const VkDeviceSize offset, const uint32_t drawCount,
uint32_t stride) {
auto buf_state = Get<BUFFER_STATE>(buffer);
VkDeviceSize size = struct_size;
if (drawCount == 1 || stride == size) {
if (drawCount > 1) size *= drawCount;
const ResourceAccessRange range = MakeRange(offset, size);
context.UpdateAccessState(*buf_state, SYNC_DRAW_INDIRECT_INDIRECT_COMMAND_READ, SyncOrdering::kNonAttachment, range, tag);
} else {
for (uint32_t i = 0; i < drawCount; ++i) {
const ResourceAccessRange range = MakeRange(offset + i * stride, size);
context.UpdateAccessState(*buf_state, SYNC_DRAW_INDIRECT_INDIRECT_COMMAND_READ, SyncOrdering::kNonAttachment, range,
tag);
}
}
}
bool SyncValidator::ValidateCountBuffer(const CommandBufferAccessContext &cb_context, const AccessContext &context,
VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
const Location &loc) const {
bool skip = false;
auto count_buf_state = Get<BUFFER_STATE>(buffer);
const ResourceAccessRange range = MakeRange(offset, 4);
auto hazard = context.DetectHazard(*count_buf_state, SYNC_DRAW_INDIRECT_INDIRECT_COMMAND_READ, range);
if (hazard.IsHazard()) {
skip |=
LogError(string_SyncHazardVUID(hazard.Hazard()), count_buf_state->buffer(), loc,
"Hazard %s for countBuffer %s in %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(buffer).c_str(), FormatHandle(commandBuffer).c_str(), cb_context.FormatHazard(hazard).c_str());
}
return skip;
}
void SyncValidator::RecordCountBuffer(AccessContext &context, const ResourceUsageTag tag, VkBuffer buffer, VkDeviceSize offset) {
auto count_buf_state = Get<BUFFER_STATE>(buffer);
const ResourceAccessRange range = MakeRange(offset, 4);
context.UpdateAccessState(*count_buf_state, SYNC_DRAW_INDIRECT_INDIRECT_COMMAND_READ, SyncOrdering::kNonAttachment, range, tag);
}
bool SyncValidator::PreCallValidateCmdDispatch(VkCommandBuffer commandBuffer, uint32_t x, uint32_t y, uint32_t z,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
skip |= cb_state->access_context.ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_COMPUTE, error_obj.location);
return skip;
}
void SyncValidator::PreCallRecordCmdDispatch(VkCommandBuffer commandBuffer, uint32_t x, uint32_t y, uint32_t z) {
StateTracker::PreCallRecordCmdDispatch(commandBuffer, x, y, z);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdDispatch);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_COMPUTE, tag);
}
bool SyncValidator::PreCallValidateCmdDispatchIndirect(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *context = cb_state->access_context.GetCurrentAccessContext();
assert(context);
if (!context) return skip;
skip |= cb_state->access_context.ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_COMPUTE, error_obj.location);
skip |= ValidateIndirectBuffer(cb_state->access_context, *context, commandBuffer, sizeof(VkDispatchIndirectCommand), buffer,
offset, 1, sizeof(VkDispatchIndirectCommand), error_obj.location);
return skip;
}
void SyncValidator::PreCallRecordCmdDispatchIndirect(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset) {
StateTracker::PreCallRecordCmdDispatchIndirect(commandBuffer, buffer, offset);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdDispatchIndirect);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_COMPUTE, tag);
RecordIndirectBuffer(*context, tag, sizeof(VkDispatchIndirectCommand), buffer, offset, 1, sizeof(VkDispatchIndirectCommand));
}
bool SyncValidator::PreCallValidateCmdDraw(VkCommandBuffer commandBuffer, uint32_t vertexCount, uint32_t instanceCount,
uint32_t firstVertex, uint32_t firstInstance, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
skip |= cb_access_context->ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, error_obj.location);
skip |= cb_access_context->ValidateDrawVertex(vertexCount, firstVertex, error_obj.location);
skip |= cb_access_context->ValidateDrawSubpassAttachment(error_obj.location);
return skip;
}
void SyncValidator::PreCallRecordCmdDraw(VkCommandBuffer commandBuffer, uint32_t vertexCount, uint32_t instanceCount,
uint32_t firstVertex, uint32_t firstInstance) {
StateTracker::PreCallRecordCmdDraw(commandBuffer, vertexCount, instanceCount, firstVertex, firstInstance);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdDraw);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, tag);
cb_access_context->RecordDrawVertex(vertexCount, firstVertex, tag);
cb_access_context->RecordDrawSubpassAttachment(tag);
}
bool SyncValidator::PreCallValidateCmdDrawIndexed(VkCommandBuffer commandBuffer, uint32_t indexCount, uint32_t instanceCount,
uint32_t firstIndex, int32_t vertexOffset, uint32_t firstInstance,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
skip |= cb_access_context->ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, error_obj.location);
skip |= cb_access_context->ValidateDrawVertexIndex(indexCount, firstIndex, error_obj.location);
skip |= cb_access_context->ValidateDrawSubpassAttachment(error_obj.location);
return skip;
}
void SyncValidator::PreCallRecordCmdDrawIndexed(VkCommandBuffer commandBuffer, uint32_t indexCount, uint32_t instanceCount,
uint32_t firstIndex, int32_t vertexOffset, uint32_t firstInstance) {
StateTracker::PreCallRecordCmdDrawIndexed(commandBuffer, indexCount, instanceCount, firstIndex, vertexOffset, firstInstance);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdDrawIndexed);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, tag);
cb_access_context->RecordDrawVertexIndex(indexCount, firstIndex, tag);
cb_access_context->RecordDrawSubpassAttachment(tag);
}
bool SyncValidator::PreCallValidateCmdDrawIndirect(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
uint32_t drawCount, uint32_t stride, const ErrorObject &error_obj) const {
bool skip = false;
if (drawCount == 0) return skip;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
skip |= cb_access_context->ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, error_obj.location);
skip |= cb_access_context->ValidateDrawSubpassAttachment(error_obj.location);
skip |= ValidateIndirectBuffer(*cb_access_context, *context, commandBuffer, sizeof(VkDrawIndirectCommand), buffer, offset,
drawCount, stride, error_obj.location);
// TODO: For now, we validate the whole vertex buffer. It might cause some false positive.
// VkDrawIndirectCommand buffer could be changed until SubmitQueue.
// We will validate the vertex buffer in SubmitQueue in the future.
skip |= cb_access_context->ValidateDrawVertex(std::optional<uint32_t>(), 0, error_obj.location);
return skip;
}
void SyncValidator::PreCallRecordCmdDrawIndirect(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
uint32_t drawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndirect(commandBuffer, buffer, offset, drawCount, stride);
if (drawCount == 0) return;
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdDrawIndirect);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, tag);
cb_access_context->RecordDrawSubpassAttachment(tag);
RecordIndirectBuffer(*context, tag, sizeof(VkDrawIndirectCommand), buffer, offset, drawCount, stride);
// TODO: For now, we record the whole vertex buffer. It might cause some false positive.
// VkDrawIndirectCommand buffer could be changed until SubmitQueue.
// We will record the vertex buffer in SubmitQueue in the future.
cb_access_context->RecordDrawVertex(std::optional<uint32_t>(), 0, tag);
}
bool SyncValidator::PreCallValidateCmdDrawIndexedIndirect(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
uint32_t drawCount, uint32_t stride, const ErrorObject &error_obj) const {
bool skip = false;
if (drawCount == 0) return skip;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
skip |= cb_access_context->ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, error_obj.location);
skip |= cb_access_context->ValidateDrawSubpassAttachment(error_obj.location);
skip |= ValidateIndirectBuffer(*cb_access_context, *context, commandBuffer, sizeof(VkDrawIndexedIndirectCommand), buffer,
offset, drawCount, stride, error_obj.location);
// TODO: For now, we validate the whole index and vertex buffer. It might cause some false positive.
// VkDrawIndexedIndirectCommand buffer could be changed until SubmitQueue.
// We will validate the index and vertex buffer in SubmitQueue in the future.
skip |= cb_access_context->ValidateDrawVertexIndex(std::optional<uint32_t>(), 0, error_obj.location);
return skip;
}
void SyncValidator::PreCallRecordCmdDrawIndexedIndirect(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
uint32_t drawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndexedIndirect(commandBuffer, buffer, offset, drawCount, stride);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdDrawIndexedIndirect);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, tag);
cb_access_context->RecordDrawSubpassAttachment(tag);
RecordIndirectBuffer(*context, tag, sizeof(VkDrawIndexedIndirectCommand), buffer, offset, drawCount, stride);
// TODO: For now, we record the whole index and vertex buffer. It might cause some false positive.
// VkDrawIndexedIndirectCommand buffer could be changed until SubmitQueue.
// We will record the index and vertex buffer in SubmitQueue in the future.
cb_access_context->RecordDrawVertexIndex(std::optional<uint32_t>(), 0, tag);
}
bool SyncValidator::PreCallValidateCmdDrawIndirectCount(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset, uint32_t maxDrawCount,
uint32_t stride, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
skip |= cb_access_context->ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, error_obj.location);
skip |= cb_access_context->ValidateDrawSubpassAttachment(error_obj.location);
skip |= ValidateIndirectBuffer(*cb_access_context, *context, commandBuffer, sizeof(VkDrawIndirectCommand), buffer, offset,
maxDrawCount, stride, error_obj.location);
skip |= ValidateCountBuffer(*cb_access_context, *context, commandBuffer, countBuffer, countBufferOffset, error_obj.location);
// TODO: For now, we validate the whole vertex buffer. It might cause some false positive.
// VkDrawIndirectCommand buffer could be changed until SubmitQueue.
// We will validate the vertex buffer in SubmitQueue in the future.
skip |= cb_access_context->ValidateDrawVertex(std::optional<uint32_t>(), 0, error_obj.location);
return skip;
}
void SyncValidator::RecordCmdDrawIndirectCount(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset, uint32_t maxDrawCount,
uint32_t stride, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(command);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, tag);
cb_access_context->RecordDrawSubpassAttachment(tag);
RecordIndirectBuffer(*context, tag, sizeof(VkDrawIndirectCommand), buffer, offset, 1, stride);
RecordCountBuffer(*context, tag, countBuffer, countBufferOffset);
// TODO: For now, we record the whole vertex buffer. It might cause some false positive.
// VkDrawIndirectCommand buffer could be changed until SubmitQueue.
// We will record the vertex buffer in SubmitQueue in the future.
cb_access_context->RecordDrawVertex(std::optional<uint32_t>(), 0, tag);
}
void SyncValidator::PreCallRecordCmdDrawIndirectCount(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset, uint32_t maxDrawCount,
uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount,
stride);
RecordCmdDrawIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
Func::vkCmdDrawIndirectCount);
}
bool SyncValidator::PreCallValidateCmdDrawIndirectCountKHR(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride,
const ErrorObject &error_obj) const {
return PreCallValidateCmdDrawIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
error_obj);
}
void SyncValidator::PreCallRecordCmdDrawIndirectCountKHR(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndirectCountKHR(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount,
stride);
RecordCmdDrawIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
Func::vkCmdDrawIndirectCountKHR);
}
bool SyncValidator::PreCallValidateCmdDrawIndirectCountAMD(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride,
const ErrorObject &error_obj) const {
return PreCallValidateCmdDrawIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
error_obj);
}
void SyncValidator::PreCallRecordCmdDrawIndirectCountAMD(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndirectCountAMD(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount,
stride);
RecordCmdDrawIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
Func::vkCmdDrawIndirectCountAMD);
}
bool SyncValidator::PreCallValidateCmdDrawIndexedIndirectCount(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
skip |= cb_access_context->ValidateDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, error_obj.location);
skip |= cb_access_context->ValidateDrawSubpassAttachment(error_obj.location);
skip |= ValidateIndirectBuffer(*cb_access_context, *context, commandBuffer, sizeof(VkDrawIndexedIndirectCommand), buffer,
offset, maxDrawCount, stride, error_obj.location);
skip |= ValidateCountBuffer(*cb_access_context, *context, commandBuffer, countBuffer, countBufferOffset, error_obj.location);
// TODO: For now, we validate the whole index and vertex buffer. It might cause some false positive.
// VkDrawIndexedIndirectCommand buffer could be changed until SubmitQueue.
// We will validate the index and vertex buffer in SubmitQueue in the future.
skip |= cb_access_context->ValidateDrawVertexIndex(std::optional<uint32_t>(), 0, error_obj.location);
return skip;
}
void SyncValidator::RecordCmdDrawIndexedIndirectCount(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset, uint32_t maxDrawCount,
uint32_t stride, Func command) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(command);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
cb_access_context->RecordDispatchDrawDescriptorSet(VK_PIPELINE_BIND_POINT_GRAPHICS, tag);
cb_access_context->RecordDrawSubpassAttachment(tag);
RecordIndirectBuffer(*context, tag, sizeof(VkDrawIndexedIndirectCommand), buffer, offset, 1, stride);
RecordCountBuffer(*context, tag, countBuffer, countBufferOffset);
// TODO: For now, we record the whole index and vertex buffer. It might cause some false positive.
// VkDrawIndexedIndirectCommand buffer could be changed until SubmitQueue.
// We will update the index and vertex buffer in SubmitQueue in the future.
cb_access_context->RecordDrawVertexIndex(std::optional<uint32_t>(), 0, tag);
}
void SyncValidator::PreCallRecordCmdDrawIndexedIndirectCount(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndexedIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset,
maxDrawCount, stride);
RecordCmdDrawIndexedIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
Func::vkCmdDrawIndexedIndirectCount);
}
bool SyncValidator::PreCallValidateCmdDrawIndexedIndirectCountKHR(VkCommandBuffer commandBuffer, VkBuffer buffer,
VkDeviceSize offset, VkBuffer countBuffer,
VkDeviceSize countBufferOffset, uint32_t maxDrawCount,
uint32_t stride, const ErrorObject &error_obj) const {
return PreCallValidateCmdDrawIndexedIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount,
stride, error_obj);
}
void SyncValidator::PreCallRecordCmdDrawIndexedIndirectCountKHR(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndexedIndirectCountKHR(commandBuffer, buffer, offset, countBuffer, countBufferOffset,
maxDrawCount, stride);
RecordCmdDrawIndexedIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
Func::vkCmdDrawIndexedIndirectCountKHR);
}
bool SyncValidator::PreCallValidateCmdDrawIndexedIndirectCountAMD(VkCommandBuffer commandBuffer, VkBuffer buffer,
VkDeviceSize offset, VkBuffer countBuffer,
VkDeviceSize countBufferOffset, uint32_t maxDrawCount,
uint32_t stride, const ErrorObject &error_obj) const {
return PreCallValidateCmdDrawIndexedIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount,
stride, error_obj);
}
void SyncValidator::PreCallRecordCmdDrawIndexedIndirectCountAMD(VkCommandBuffer commandBuffer, VkBuffer buffer, VkDeviceSize offset,
VkBuffer countBuffer, VkDeviceSize countBufferOffset,
uint32_t maxDrawCount, uint32_t stride) {
StateTracker::PreCallRecordCmdDrawIndexedIndirectCountAMD(commandBuffer, buffer, offset, countBuffer, countBufferOffset,
maxDrawCount, stride);
RecordCmdDrawIndexedIndirectCount(commandBuffer, buffer, offset, countBuffer, countBufferOffset, maxDrawCount, stride,
Func::vkCmdDrawIndexedIndirectCountAMD);
}
bool SyncValidator::PreCallValidateCmdClearColorImage(VkCommandBuffer commandBuffer, VkImage image, VkImageLayout imageLayout,
const VkClearColorValue *pColor, uint32_t rangeCount,
const VkImageSubresourceRange *pRanges, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto image_state = Get<ImageState>(image);
for (uint32_t index = 0; index < rangeCount; index++) {
const auto &range = pRanges[index];
if (image_state) {
auto hazard = context->DetectHazard(*image_state, SYNC_CLEAR_TRANSFER_WRITE, range, false);
if (hazard.IsHazard()) {
skip |= LogError(image, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdClearColorImage: Hazard %s for %s, range index %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(image).c_str(), index,
cb_access_context->FormatHazard(hazard).c_str());
}
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdClearColorImage(VkCommandBuffer commandBuffer, VkImage image, VkImageLayout imageLayout,
const VkClearColorValue *pColor, uint32_t rangeCount,
const VkImageSubresourceRange *pRanges) {
StateTracker::PreCallRecordCmdClearColorImage(commandBuffer, image, imageLayout, pColor, rangeCount, pRanges);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdClearColorImage);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto image_state = Get<ImageState>(image);
for (uint32_t index = 0; index < rangeCount; index++) {
const auto &range = pRanges[index];
if (image_state) {
context->UpdateAccessState(*image_state, SYNC_CLEAR_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
}
}
bool SyncValidator::PreCallValidateCmdClearDepthStencilImage(VkCommandBuffer commandBuffer, VkImage image,
VkImageLayout imageLayout,
const VkClearDepthStencilValue *pDepthStencil, uint32_t rangeCount,
const VkImageSubresourceRange *pRanges,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto image_state = Get<ImageState>(image);
for (uint32_t index = 0; index < rangeCount; index++) {
const auto &range = pRanges[index];
if (image_state) {
auto hazard = context->DetectHazard(*image_state, SYNC_CLEAR_TRANSFER_WRITE, range, false);
if (hazard.IsHazard()) {
skip |= LogError(image, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdClearDepthStencilImage: Hazard %s for %s, range index %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(image).c_str(), index,
cb_access_context->FormatHazard(hazard).c_str());
}
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdClearDepthStencilImage(VkCommandBuffer commandBuffer, VkImage image, VkImageLayout imageLayout,
const VkClearDepthStencilValue *pDepthStencil, uint32_t rangeCount,
const VkImageSubresourceRange *pRanges) {
StateTracker::PreCallRecordCmdClearDepthStencilImage(commandBuffer, image, imageLayout, pDepthStencil, rangeCount, pRanges);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdClearDepthStencilImage);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto image_state = Get<ImageState>(image);
for (uint32_t index = 0; index < rangeCount; index++) {
const auto &range = pRanges[index];
if (image_state) {
context->UpdateAccessState(*image_state, SYNC_CLEAR_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
}
}
bool SyncValidator::PreCallValidateCmdClearAttachments(VkCommandBuffer commandBuffer, uint32_t attachmentCount,
const VkClearAttachment *pAttachments, uint32_t rectCount,
const VkClearRect *pRects, const ErrorObject &error_obj) const {
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
const auto cb_access_context = &cb_state->access_context;
const auto rp_access_context = cb_access_context->GetCurrentRenderPassContext();
if (!rp_access_context) return false;
bool skip = false;
for (const auto &attachment : vvl::make_span(pAttachments, attachmentCount)) {
for (const auto &rect : vvl::make_span(pRects, rectCount)) {
const auto rect_index = static_cast<uint32_t>(&rect - pRects);
skip |= rp_access_context->ValidateClearAttachment(cb_access_context->GetExecutionContext(), *cb_state,
error_obj.location, attachment, rect, rect_index);
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdClearAttachments(VkCommandBuffer commandBuffer, uint32_t attachmentCount,
const VkClearAttachment *pAttachments, uint32_t rectCount,
const VkClearRect *pRects) {
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
auto cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdClearAttachments);
const auto rp_access_context = cb_access_context->GetCurrentRenderPassContext();
if (!rp_access_context) return;
for (const auto &attachment : vvl::make_span(pAttachments, attachmentCount)) {
for (const auto &rect : vvl::make_span(pRects, rectCount)) {
rp_access_context->RecordClearAttachment(*cb_state, tag, attachment, rect);
}
}
}
bool SyncValidator::PreCallValidateCmdCopyQueryPoolResults(VkCommandBuffer commandBuffer, VkQueryPool queryPool,
uint32_t firstQuery, uint32_t queryCount, VkBuffer dstBuffer,
VkDeviceSize dstOffset, VkDeviceSize stride, VkQueryResultFlags flags,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(dstOffset, stride * queryCount);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, range);
if (hazard.IsHazard()) {
skip |= LogError(dstBuffer, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdCopyQueryPoolResults: Hazard %s for dstBuffer %s. Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(dstBuffer).c_str(),
cb_access_context->FormatHazard(hazard).c_str());
}
}
// TODO:Track VkQueryPool
return skip;
}
void SyncValidator::PreCallRecordCmdCopyQueryPoolResults(VkCommandBuffer commandBuffer, VkQueryPool queryPool, uint32_t firstQuery,
uint32_t queryCount, VkBuffer dstBuffer, VkDeviceSize dstOffset,
VkDeviceSize stride, VkQueryResultFlags flags) {
StateTracker::PreCallRecordCmdCopyQueryPoolResults(commandBuffer, queryPool, firstQuery, queryCount, dstBuffer, dstOffset,
stride, flags);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdCopyQueryPoolResults);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(dstOffset, stride * queryCount);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
// TODO:Track VkQueryPool
}
bool SyncValidator::PreCallValidateCmdFillBuffer(VkCommandBuffer commandBuffer, VkBuffer dstBuffer, VkDeviceSize dstOffset,
VkDeviceSize size, uint32_t data, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(*dst_buffer, dstOffset, size);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, range);
if (hazard.IsHazard()) {
skip |= LogError(dstBuffer, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdFillBuffer: Hazard %s for dstBuffer %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(dstBuffer).c_str(), cb_access_context->FormatHazard(hazard).c_str());
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdFillBuffer(VkCommandBuffer commandBuffer, VkBuffer dstBuffer, VkDeviceSize dstOffset,
VkDeviceSize size, uint32_t data) {
StateTracker::PreCallRecordCmdFillBuffer(commandBuffer, dstBuffer, dstOffset, size, data);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdFillBuffer);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(*dst_buffer, dstOffset, size);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
}
bool SyncValidator::PreCallValidateCmdResolveImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const VkImageResolve *pRegions, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto src_image = Get<ImageState>(srcImage);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &resolve_region = pRegions[region];
if (src_image) {
auto hazard = context->DetectHazard(*src_image, SYNC_RESOLVE_TRANSFER_READ, resolve_region.srcSubresource,
resolve_region.srcOffset, resolve_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(srcImage, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdResolveImage: Hazard %s for srcImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(srcImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
}
if (dst_image) {
auto hazard = context->DetectHazard(*dst_image, SYNC_RESOLVE_TRANSFER_WRITE, resolve_region.dstSubresource,
resolve_region.dstOffset, resolve_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(dstImage, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdResolveImage: Hazard %s for dstImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(dstImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
if (skip) break;
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdResolveImage(VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout,
VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount,
const VkImageResolve *pRegions) {
StateTracker::PreCallRecordCmdResolveImage(commandBuffer, srcImage, srcImageLayout, dstImage, dstImageLayout, regionCount,
pRegions);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdResolveImage);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto src_image = Get<ImageState>(srcImage);
auto dst_image = Get<ImageState>(dstImage);
for (uint32_t region = 0; region < regionCount; region++) {
const auto &resolve_region = pRegions[region];
if (src_image) {
context->UpdateAccessState(*src_image, SYNC_RESOLVE_TRANSFER_READ, SyncOrdering::kNonAttachment,
resolve_region.srcSubresource, resolve_region.srcOffset, resolve_region.extent, tag);
}
if (dst_image) {
context->UpdateAccessState(*dst_image, SYNC_RESOLVE_TRANSFER_WRITE, SyncOrdering::kNonAttachment,
resolve_region.dstSubresource, resolve_region.dstOffset, resolve_region.extent, tag);
}
}
}
bool SyncValidator::PreCallValidateCmdResolveImage2(VkCommandBuffer commandBuffer, const VkResolveImageInfo2KHR *pResolveImageInfo,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
const Location image_info_loc = error_obj.location.dot(Field::pResolveImageInfo);
auto src_image = Get<ImageState>(pResolveImageInfo->srcImage);
auto dst_image = Get<ImageState>(pResolveImageInfo->dstImage);
for (uint32_t region = 0; region < pResolveImageInfo->regionCount; region++) {
const Location region_loc = image_info_loc.dot(Field::pRegions, region);
const auto &resolve_region = pResolveImageInfo->pRegions[region];
if (src_image) {
auto hazard = context->DetectHazard(*src_image, SYNC_RESOLVE_TRANSFER_READ, resolve_region.srcSubresource,
resolve_region.srcOffset, resolve_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), pResolveImageInfo->srcImage, region_loc,
"Hazard %s for srcImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(pResolveImageInfo->srcImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
}
if (dst_image) {
auto hazard = context->DetectHazard(*dst_image, SYNC_RESOLVE_TRANSFER_WRITE, resolve_region.dstSubresource,
resolve_region.dstOffset, resolve_region.extent, false);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), pResolveImageInfo->dstImage, region_loc,
"Hazard %s for dstImage %s, region %" PRIu32 ". Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(pResolveImageInfo->dstImage).c_str(), region,
cb_access_context->FormatHazard(hazard).c_str());
}
if (skip) break;
}
}
return skip;
}
bool SyncValidator::PreCallValidateCmdResolveImage2KHR(VkCommandBuffer commandBuffer,
const VkResolveImageInfo2KHR *pResolveImageInfo,
const ErrorObject &error_obj) const {
return PreCallValidateCmdResolveImage2(commandBuffer, pResolveImageInfo, error_obj);
}
void SyncValidator::RecordCmdResolveImage2(VkCommandBuffer commandBuffer, const VkResolveImageInfo2KHR *pResolveImageInfo,
Func command) {
StateTracker::PreCallRecordCmdResolveImage2KHR(commandBuffer, pResolveImageInfo);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(command);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto src_image = Get<ImageState>(pResolveImageInfo->srcImage);
auto dst_image = Get<ImageState>(pResolveImageInfo->dstImage);
for (uint32_t region = 0; region < pResolveImageInfo->regionCount; region++) {
const auto &resolve_region = pResolveImageInfo->pRegions[region];
if (src_image) {
context->UpdateAccessState(*src_image, SYNC_RESOLVE_TRANSFER_READ, SyncOrdering::kNonAttachment,
resolve_region.srcSubresource, resolve_region.srcOffset, resolve_region.extent, tag);
}
if (dst_image) {
context->UpdateAccessState(*dst_image, SYNC_RESOLVE_TRANSFER_WRITE, SyncOrdering::kNonAttachment,
resolve_region.dstSubresource, resolve_region.dstOffset, resolve_region.extent, tag);
}
}
}
void SyncValidator::PreCallRecordCmdResolveImage2KHR(VkCommandBuffer commandBuffer,
const VkResolveImageInfo2KHR *pResolveImageInfo) {
RecordCmdResolveImage2(commandBuffer, pResolveImageInfo, Func::vkCmdResolveImage2KHR);
}
void SyncValidator::PreCallRecordCmdResolveImage2(VkCommandBuffer commandBuffer, const VkResolveImageInfo2 *pResolveImageInfo) {
RecordCmdResolveImage2(commandBuffer, pResolveImageInfo, Func::vkCmdResolveImage2);
}
bool SyncValidator::PreCallValidateCmdUpdateBuffer(VkCommandBuffer commandBuffer, VkBuffer dstBuffer, VkDeviceSize dstOffset,
VkDeviceSize dataSize, const void *pData, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
// VK_WHOLE_SIZE not allowed
const ResourceAccessRange range = MakeRange(dstOffset, dataSize);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, range);
if (hazard.IsHazard()) {
skip |= LogError(dstBuffer, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdUpdateBuffer: Hazard %s for dstBuffer %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(dstBuffer).c_str(), cb_access_context->FormatHazard(hazard).c_str());
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdUpdateBuffer(VkCommandBuffer commandBuffer, VkBuffer dstBuffer, VkDeviceSize dstOffset,
VkDeviceSize dataSize, const void *pData) {
StateTracker::PreCallRecordCmdUpdateBuffer(commandBuffer, dstBuffer, dstOffset, dataSize, pData);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdUpdateBuffer);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
// VK_WHOLE_SIZE not allowed
const ResourceAccessRange range = MakeRange(dstOffset, dataSize);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
}
bool SyncValidator::PreCallValidateCmdWriteBufferMarkerAMD(VkCommandBuffer commandBuffer, VkPipelineStageFlagBits pipelineStage,
VkBuffer dstBuffer, VkDeviceSize dstOffset, uint32_t marker,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(dstOffset, 4);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, range);
if (hazard.IsHazard()) {
skip |= LogError(string_SyncHazardVUID(hazard.Hazard()), dstBuffer, error_obj.location,
"Hazard %s for dstBuffer %s. Access info %s.", string_SyncHazard(hazard.Hazard()),
FormatHandle(dstBuffer).c_str(), cb_access_context->FormatHazard(hazard).c_str());
}
}
return skip;
}
void SyncValidator::PreCallRecordCmdWriteBufferMarkerAMD(VkCommandBuffer commandBuffer, VkPipelineStageFlagBits pipelineStage,
VkBuffer dstBuffer, VkDeviceSize dstOffset, uint32_t marker) {
StateTracker::PreCallRecordCmdWriteBufferMarkerAMD(commandBuffer, pipelineStage, dstBuffer, dstOffset, marker);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdWriteBufferMarkerAMD);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(dstOffset, 4);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
}
bool SyncValidator::PreCallValidateCmdSetEvent(VkCommandBuffer commandBuffer, VkEvent event, VkPipelineStageFlags stageMask,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
const auto *access_context = cb_context->GetCurrentAccessContext();
assert(access_context);
if (!access_context) return skip;
SyncOpSetEvent set_event_op(error_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask, nullptr);
return set_event_op.Validate(*cb_context);
}
void SyncValidator::PostCallRecordCmdSetEvent(VkCommandBuffer commandBuffer, VkEvent event, VkPipelineStageFlags stageMask,
const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdSetEvent(commandBuffer, event, stageMask, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpSetEvent>(record_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask,
cb_context->GetCurrentAccessContext());
}
bool SyncValidator::PreCallValidateCmdSetEvent2KHR(VkCommandBuffer commandBuffer, VkEvent event,
const VkDependencyInfoKHR *pDependencyInfo, const ErrorObject &error_obj) const {
return PreCallValidateCmdSetEvent2(commandBuffer, event, pDependencyInfo, error_obj);
}
bool SyncValidator::PreCallValidateCmdSetEvent2(VkCommandBuffer commandBuffer, VkEvent event,
const VkDependencyInfo *pDependencyInfo, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
if (!pDependencyInfo) return skip;
const auto *access_context = cb_context->GetCurrentAccessContext();
assert(access_context);
if (!access_context) return skip;
SyncOpSetEvent set_event_op(error_obj.location.function, *this, cb_context->GetQueueFlags(), event, *pDependencyInfo, nullptr);
return set_event_op.Validate(*cb_context);
}
void SyncValidator::PostCallRecordCmdSetEvent2KHR(VkCommandBuffer commandBuffer, VkEvent event,
const VkDependencyInfoKHR *pDependencyInfo, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdSetEvent2KHR(commandBuffer, event, pDependencyInfo, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
if (!pDependencyInfo) return;
cb_context->RecordSyncOp<SyncOpSetEvent>(record_obj.location.function, *this, cb_context->GetQueueFlags(), event,
*pDependencyInfo, cb_context->GetCurrentAccessContext());
}
void SyncValidator::PostCallRecordCmdSetEvent2(VkCommandBuffer commandBuffer, VkEvent event,
const VkDependencyInfo *pDependencyInfo, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdSetEvent2(commandBuffer, event, pDependencyInfo, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
if (!pDependencyInfo) return;
cb_context->RecordSyncOp<SyncOpSetEvent>(record_obj.location.function, *this, cb_context->GetQueueFlags(), event,
*pDependencyInfo, cb_context->GetCurrentAccessContext());
}
bool SyncValidator::PreCallValidateCmdResetEvent(VkCommandBuffer commandBuffer, VkEvent event, VkPipelineStageFlags stageMask,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
SyncOpResetEvent reset_event_op(error_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask);
return reset_event_op.Validate(*cb_context);
}
void SyncValidator::PostCallRecordCmdResetEvent(VkCommandBuffer commandBuffer, VkEvent event, VkPipelineStageFlags stageMask,
const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdResetEvent(commandBuffer, event, stageMask, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpResetEvent>(record_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask);
}
bool SyncValidator::PreCallValidateCmdResetEvent2(VkCommandBuffer commandBuffer, VkEvent event, VkPipelineStageFlags2 stageMask,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
SyncOpResetEvent reset_event_op(error_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask);
return reset_event_op.Validate(*cb_context);
return PreCallValidateCmdResetEvent2(commandBuffer, event, stageMask, error_obj);
}
bool SyncValidator::PreCallValidateCmdResetEvent2KHR(VkCommandBuffer commandBuffer, VkEvent event,
VkPipelineStageFlags2KHR stageMask, const ErrorObject &error_obj) const {
return PreCallValidateCmdResetEvent2(commandBuffer, event, stageMask, error_obj);
}
void SyncValidator::PostCallRecordCmdResetEvent2KHR(VkCommandBuffer commandBuffer, VkEvent event,
VkPipelineStageFlags2KHR stageMask, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdResetEvent2KHR(commandBuffer, event, stageMask, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpResetEvent>(record_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask);
}
void SyncValidator::PostCallRecordCmdResetEvent2(VkCommandBuffer commandBuffer, VkEvent event, VkPipelineStageFlags2 stageMask,
const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdResetEvent2(commandBuffer, event, stageMask, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpResetEvent>(record_obj.location.function, *this, cb_context->GetQueueFlags(), event, stageMask);
}
bool SyncValidator::PreCallValidateCmdWaitEvents(VkCommandBuffer commandBuffer, uint32_t eventCount, const VkEvent *pEvents,
VkPipelineStageFlags srcStageMask, VkPipelineStageFlags dstStageMask,
uint32_t memoryBarrierCount, const VkMemoryBarrier *pMemoryBarriers,
uint32_t bufferMemoryBarrierCount,
const VkBufferMemoryBarrier *pBufferMemoryBarriers,
uint32_t imageMemoryBarrierCount, const VkImageMemoryBarrier *pImageMemoryBarriers,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
SyncOpWaitEvents wait_events_op(error_obj.location.function, *this, cb_context->GetQueueFlags(), eventCount, pEvents,
srcStageMask, dstStageMask, memoryBarrierCount, pMemoryBarriers, bufferMemoryBarrierCount,
pBufferMemoryBarriers, imageMemoryBarrierCount, pImageMemoryBarriers);
return wait_events_op.Validate(*cb_context);
}
void SyncValidator::PostCallRecordCmdWaitEvents(VkCommandBuffer commandBuffer, uint32_t eventCount, const VkEvent *pEvents,
VkPipelineStageFlags srcStageMask, VkPipelineStageFlags dstStageMask,
uint32_t memoryBarrierCount, const VkMemoryBarrier *pMemoryBarriers,
uint32_t bufferMemoryBarrierCount,
const VkBufferMemoryBarrier *pBufferMemoryBarriers,
uint32_t imageMemoryBarrierCount, const VkImageMemoryBarrier *pImageMemoryBarriers,
const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdWaitEvents(commandBuffer, eventCount, pEvents, srcStageMask, dstStageMask, memoryBarrierCount,
pMemoryBarriers, bufferMemoryBarrierCount, pBufferMemoryBarriers,
imageMemoryBarrierCount, pImageMemoryBarriers, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpWaitEvents>(record_obj.location.function, *this, cb_context->GetQueueFlags(), eventCount,
pEvents, srcStageMask, dstStageMask, memoryBarrierCount, pMemoryBarriers,
bufferMemoryBarrierCount, pBufferMemoryBarriers, imageMemoryBarrierCount,
pImageMemoryBarriers);
}
bool SyncValidator::PreCallValidateCmdWaitEvents2KHR(VkCommandBuffer commandBuffer, uint32_t eventCount, const VkEvent *pEvents,
const VkDependencyInfoKHR *pDependencyInfos,
const ErrorObject &error_obj) const {
return PreCallValidateCmdWaitEvents2(commandBuffer, eventCount, pEvents, pDependencyInfos, error_obj);
}
void SyncValidator::PostCallRecordCmdWaitEvents2KHR(VkCommandBuffer commandBuffer, uint32_t eventCount, const VkEvent *pEvents,
const VkDependencyInfoKHR *pDependencyInfos, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdWaitEvents2KHR(commandBuffer, eventCount, pEvents, pDependencyInfos, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpWaitEvents>(record_obj.location.function, *this, cb_context->GetQueueFlags(), eventCount,
pEvents, pDependencyInfos);
}
bool SyncValidator::PreCallValidateCmdWaitEvents2(VkCommandBuffer commandBuffer, uint32_t eventCount, const VkEvent *pEvents,
const VkDependencyInfo *pDependencyInfos, const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
SyncOpWaitEvents wait_events_op(error_obj.location.function, *this, cb_context->GetQueueFlags(), eventCount, pEvents,
pDependencyInfos);
skip |= wait_events_op.Validate(*cb_context);
return skip;
}
void SyncValidator::PostCallRecordCmdWaitEvents2(VkCommandBuffer commandBuffer, uint32_t eventCount, const VkEvent *pEvents,
const VkDependencyInfo *pDependencyInfos, const RecordObject &record_obj) {
StateTracker::PostCallRecordCmdWaitEvents2KHR(commandBuffer, eventCount, pEvents, pDependencyInfos, record_obj);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
cb_context->RecordSyncOp<SyncOpWaitEvents>(record_obj.location.function, *this, cb_context->GetQueueFlags(), eventCount,
pEvents, pDependencyInfos);
}
void SyncEventState::ResetFirstScope() {
first_scope.reset();
scope = SyncExecScope();
first_scope_tag = 0;
}
// Keep the "ignore this event" logic in same place for ValidateWait and RecordWait to use
SyncEventState::IgnoreReason SyncEventState::IsIgnoredByWait(vvl::Func command, VkPipelineStageFlags2KHR srcStageMask) const {
IgnoreReason reason = NotIgnored;
if ((vvl::Func::vkCmdWaitEvents2KHR == command || vvl::Func::vkCmdWaitEvents2 == command) &&
(vvl::Func::vkCmdSetEvent == last_command)) {
reason = SetVsWait2;
} else if ((last_command == vvl::Func::vkCmdResetEvent || last_command == vvl::Func::vkCmdResetEvent2KHR) &&
!HasBarrier(0U, 0U)) {
reason = (last_command == vvl::Func::vkCmdResetEvent) ? ResetWaitRace : Reset2WaitRace;
} else if (unsynchronized_set != vvl::Func::Empty) {
reason = SetRace;
} else if (first_scope) {
const VkPipelineStageFlags2KHR missing_bits = scope.mask_param & ~srcStageMask;
// Note it is the "not missing bits" path that is the only "NotIgnored" path
if (missing_bits) reason = MissingStageBits;
} else {
reason = MissingSetEvent;
}
return reason;
}
bool SyncEventState::HasBarrier(VkPipelineStageFlags2KHR stageMask, VkPipelineStageFlags2KHR exec_scope_arg) const {
return (last_command == vvl::Func::Empty) || (stageMask & VK_PIPELINE_STAGE_ALL_COMMANDS_BIT) || (barriers & exec_scope_arg) ||
(barriers & VK_PIPELINE_STAGE_ALL_COMMANDS_BIT);
}
void SyncEventState::AddReferencedTags(ResourceUsageTagSet &referenced) const {
if (first_scope) {
first_scope->AddReferencedTags(referenced);
}
}
SyncOpBarriers::SyncOpBarriers(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags,
VkPipelineStageFlags srcStageMask, VkPipelineStageFlags dstStageMask,
VkDependencyFlags dependencyFlags, uint32_t memoryBarrierCount,
const VkMemoryBarrier *pMemoryBarriers, uint32_t bufferMemoryBarrierCount,
const VkBufferMemoryBarrier *pBufferMemoryBarriers, uint32_t imageMemoryBarrierCount,
const VkImageMemoryBarrier *pImageMemoryBarriers)
: SyncOpBase(command), barriers_(1) {
auto &barrier_set = barriers_[0];
barrier_set.dependency_flags = dependencyFlags;
barrier_set.src_exec_scope = SyncExecScope::MakeSrc(queue_flags, srcStageMask);
barrier_set.dst_exec_scope = SyncExecScope::MakeDst(queue_flags, dstStageMask);
// Translate the API parameters into structures SyncVal understands directly, and dehandle for safer/faster replay.
barrier_set.MakeMemoryBarriers(barrier_set.src_exec_scope, barrier_set.dst_exec_scope, dependencyFlags, memoryBarrierCount,
pMemoryBarriers);
barrier_set.MakeBufferMemoryBarriers(sync_state, barrier_set.src_exec_scope, barrier_set.dst_exec_scope, dependencyFlags,
bufferMemoryBarrierCount, pBufferMemoryBarriers);
barrier_set.MakeImageMemoryBarriers(sync_state, barrier_set.src_exec_scope, barrier_set.dst_exec_scope, dependencyFlags,
imageMemoryBarrierCount, pImageMemoryBarriers);
}
SyncOpBarriers::SyncOpBarriers(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags, uint32_t event_count,
const VkDependencyInfoKHR *dep_infos)
: SyncOpBase(command), barriers_(event_count) {
for (uint32_t i = 0; i < event_count; i++) {
const auto &dep_info = dep_infos[i];
auto &barrier_set = barriers_[i];
barrier_set.dependency_flags = dep_info.dependencyFlags;
auto stage_masks = sync_utils::GetGlobalStageMasks(dep_info);
barrier_set.src_exec_scope = SyncExecScope::MakeSrc(queue_flags, stage_masks.src);
barrier_set.dst_exec_scope = SyncExecScope::MakeDst(queue_flags, stage_masks.dst);
// Translate the API parameters into structures SyncVal understands directly, and dehandle for safer/faster replay.
barrier_set.MakeMemoryBarriers(queue_flags, dep_info.dependencyFlags, dep_info.memoryBarrierCount,
dep_info.pMemoryBarriers);
barrier_set.MakeBufferMemoryBarriers(sync_state, queue_flags, dep_info.dependencyFlags, dep_info.bufferMemoryBarrierCount,
dep_info.pBufferMemoryBarriers);
barrier_set.MakeImageMemoryBarriers(sync_state, queue_flags, dep_info.dependencyFlags, dep_info.imageMemoryBarrierCount,
dep_info.pImageMemoryBarriers);
}
}
SyncOpPipelineBarrier::SyncOpPipelineBarrier(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags,
VkPipelineStageFlags srcStageMask, VkPipelineStageFlags dstStageMask,
VkDependencyFlags dependencyFlags, uint32_t memoryBarrierCount,
const VkMemoryBarrier *pMemoryBarriers, uint32_t bufferMemoryBarrierCount,
const VkBufferMemoryBarrier *pBufferMemoryBarriers, uint32_t imageMemoryBarrierCount,
const VkImageMemoryBarrier *pImageMemoryBarriers)
: SyncOpBarriers(command, sync_state, queue_flags, srcStageMask, dstStageMask, dependencyFlags, memoryBarrierCount,
pMemoryBarriers, bufferMemoryBarrierCount, pBufferMemoryBarriers, imageMemoryBarrierCount,
pImageMemoryBarriers) {}
SyncOpPipelineBarrier::SyncOpPipelineBarrier(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags,
const VkDependencyInfoKHR &dep_info)
: SyncOpBarriers(command, sync_state, queue_flags, 1, &dep_info) {}
bool SyncOpPipelineBarrier::Validate(const CommandBufferAccessContext &cb_context) const {
bool skip = false;
const auto *context = cb_context.GetCurrentAccessContext();
assert(context);
if (!context) return skip;
assert(barriers_.size() == 1); // PipelineBarriers only support a single barrier set.
// Validate Image Layout transitions
const auto &barrier_set = barriers_[0];
for (const auto &image_barrier : barrier_set.image_memory_barriers) {
if (image_barrier.new_layout == image_barrier.old_layout) continue; // Only interested in layout transitions at this point.
const auto *image_state = image_barrier.image.get();
if (!image_state) continue;
const auto hazard = context->DetectImageBarrierHazard(image_barrier);
if (hazard.IsHazard()) {
// PHASE1 TODO -- add tag information to log msg when useful.
const auto &sync_state = cb_context.GetSyncState();
const auto image_handle = image_state->image();
skip |= sync_state.LogError(image_handle, string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s for image barrier %" PRIu32 " %s. Access info %s.", CmdName(),
string_SyncHazard(hazard.Hazard()), image_barrier.index,
sync_state.FormatHandle(image_handle).c_str(), cb_context.FormatHazard(hazard).c_str());
}
}
return skip;
}
struct SyncOpPipelineBarrierFunctorFactory {
using BarrierOpFunctor = PipelineBarrierOp;
using ApplyFunctor = ApplyBarrierFunctor<BarrierOpFunctor>;
using GlobalBarrierOpFunctor = PipelineBarrierOp;
using GlobalApplyFunctor = ApplyBarrierOpsFunctor<GlobalBarrierOpFunctor>;
using BufferRange = SingleRangeGenerator<ResourceAccessRange>;
using ImageRange = subresource_adapter::ImageRangeGenerator;
using GlobalRange = SingleRangeGenerator<ResourceAccessRange>;
using ImageState = syncval_state::ImageState;
ApplyFunctor MakeApplyFunctor(QueueId queue_id, const SyncBarrier &barrier, bool layout_transition) const {
return ApplyFunctor(BarrierOpFunctor(queue_id, barrier, layout_transition));
}
GlobalApplyFunctor MakeGlobalApplyFunctor(size_t size_hint, ResourceUsageTag tag) const {
return GlobalApplyFunctor(true /* resolve */, size_hint, tag);
}
GlobalBarrierOpFunctor MakeGlobalBarrierOpFunctor(QueueId queue_id, const SyncBarrier &barrier) const {
return GlobalBarrierOpFunctor(queue_id, barrier, false);
}
BufferRange MakeRangeGen(const BUFFER_STATE &buffer, const ResourceAccessRange &range) const {
if (!SimpleBinding(buffer)) return ResourceAccessRange();
const auto base_address = ResourceBaseAddress(buffer);
return (range + base_address);
}
ImageRange MakeRangeGen(const ImageState &image, const VkImageSubresourceRange &subresource_range) const {
return image.MakeImageRangeGen(subresource_range, false);
}
GlobalRange MakeGlobalRangeGen() const { return kFullRange; }
};
template <typename Barriers, typename FunctorFactory>
void SyncOpBarriers::ApplyBarriers(const Barriers &barriers, const FunctorFactory &factory, const QueueId queue_id,
const ResourceUsageTag tag, AccessContext *context) {
for (const auto &barrier : barriers) {
const auto *state = barrier.GetState();
if (state) {
auto update_action = factory.MakeApplyFunctor(queue_id, barrier.barrier, barrier.IsLayoutTransition());
auto range_gen = factory.MakeRangeGen(*state, barrier.Range());
UpdateMemoryAccessState(context->GetAccessStateMap(), update_action, range_gen);
}
}
}
template <typename Barriers, typename FunctorFactory>
void SyncOpBarriers::ApplyGlobalBarriers(const Barriers &barriers, const FunctorFactory &factory, const QueueId queue_id,
const ResourceUsageTag tag, AccessContext *access_context) {
auto barriers_functor = factory.MakeGlobalApplyFunctor(barriers.size(), tag);
for (const auto &barrier : barriers) {
barriers_functor.EmplaceBack(factory.MakeGlobalBarrierOpFunctor(queue_id, barrier));
}
auto range_gen = factory.MakeGlobalRangeGen();
UpdateMemoryAccessState(access_context->GetAccessStateMap(), barriers_functor, range_gen);
}
ResourceUsageTag SyncOpPipelineBarrier::Record(CommandBufferAccessContext *cb_context) {
const auto tag = cb_context->NextCommandTag(command_);
ReplayRecord(*cb_context, tag);
return tag;
}
void SyncOpPipelineBarrier::ReplayRecord(CommandExecutionContext &exec_context, const ResourceUsageTag exec_tag) const {
SyncOpPipelineBarrierFunctorFactory factory;
// Pipeline barriers only have a single barrier set, unlike WaitEvents2
assert(barriers_.size() == 1);
const auto &barrier_set = barriers_[0];
if (!exec_context.ValidForSyncOps()) return;
SyncEventsContext *events_context = exec_context.GetCurrentEventsContext();
AccessContext *access_context = exec_context.GetCurrentAccessContext();
const auto queue_id = exec_context.GetQueueId();
ApplyBarriers(barrier_set.buffer_memory_barriers, factory, queue_id, exec_tag, access_context);
ApplyBarriers(barrier_set.image_memory_barriers, factory, queue_id, exec_tag, access_context);
ApplyGlobalBarriers(barrier_set.memory_barriers, factory, queue_id, exec_tag, access_context);
if (barrier_set.single_exec_scope) {
events_context->ApplyBarrier(barrier_set.src_exec_scope, barrier_set.dst_exec_scope, exec_tag);
} else {
for (const auto &barrier : barrier_set.memory_barriers) {
events_context->ApplyBarrier(barrier.src_exec_scope, barrier.dst_exec_scope, exec_tag);
}
}
}
bool SyncOpPipelineBarrier::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
// The layout transitions happen at the replay tag
ResourceUsageRange first_use_range = {recorded_tag, recorded_tag + 1};
return replay.DetectFirstUseHazard(first_use_range);
}
void SyncOpBarriers::BarrierSet::MakeMemoryBarriers(const SyncExecScope &src, const SyncExecScope &dst,
VkDependencyFlags dependency_flags, uint32_t memory_barrier_count,
const VkMemoryBarrier *barriers) {
memory_barriers.reserve(std::max<uint32_t>(1, memory_barrier_count));
for (uint32_t barrier_index = 0; barrier_index < memory_barrier_count; barrier_index++) {
const auto &barrier = barriers[barrier_index];
SyncBarrier sync_barrier(barrier, src, dst);
memory_barriers.emplace_back(sync_barrier);
}
if (0 == memory_barrier_count) {
// If there are no global memory barriers, force an exec barrier
memory_barriers.emplace_back(SyncBarrier(src, dst));
}
single_exec_scope = true;
}
void SyncOpBarriers::BarrierSet::MakeBufferMemoryBarriers(const SyncValidator &sync_state, const SyncExecScope &src,
const SyncExecScope &dst, VkDependencyFlags dependencyFlags,
uint32_t barrier_count, const VkBufferMemoryBarrier *barriers) {
buffer_memory_barriers.reserve(barrier_count);
for (uint32_t index = 0; index < barrier_count; index++) {
const auto &barrier = barriers[index];
auto buffer = sync_state.Get<BUFFER_STATE>(barrier.buffer);
if (buffer) {
const auto range = MakeRange(*buffer, barrier.offset, barrier.size);
const SyncBarrier sync_barrier(barrier, src, dst);
buffer_memory_barriers.emplace_back(buffer, sync_barrier, range);
} else {
buffer_memory_barriers.emplace_back();
}
}
}
void SyncOpBarriers::BarrierSet::MakeMemoryBarriers(VkQueueFlags queue_flags, VkDependencyFlags dependency_flags,
uint32_t memory_barrier_count, const VkMemoryBarrier2 *barriers) {
memory_barriers.reserve(memory_barrier_count);
for (uint32_t barrier_index = 0; barrier_index < memory_barrier_count; barrier_index++) {
const auto &barrier = barriers[barrier_index];
auto src = SyncExecScope::MakeSrc(queue_flags, barrier.srcStageMask);
auto dst = SyncExecScope::MakeDst(queue_flags, barrier.dstStageMask);
SyncBarrier sync_barrier(barrier, src, dst);
memory_barriers.emplace_back(sync_barrier);
}
single_exec_scope = false;
}
void SyncOpBarriers::BarrierSet::MakeBufferMemoryBarriers(const SyncValidator &sync_state, VkQueueFlags queue_flags,
VkDependencyFlags dependencyFlags, uint32_t barrier_count,
const VkBufferMemoryBarrier2 *barriers) {
buffer_memory_barriers.reserve(barrier_count);
for (uint32_t index = 0; index < barrier_count; index++) {
const auto &barrier = barriers[index];
auto src = SyncExecScope::MakeSrc(queue_flags, barrier.srcStageMask);
auto dst = SyncExecScope::MakeDst(queue_flags, barrier.dstStageMask);
auto buffer = sync_state.Get<BUFFER_STATE>(barrier.buffer);
if (buffer) {
const auto range = MakeRange(*buffer, barrier.offset, barrier.size);
const SyncBarrier sync_barrier(barrier, src, dst);
buffer_memory_barriers.emplace_back(buffer, sync_barrier, range);
} else {
buffer_memory_barriers.emplace_back();
}
}
}
void SyncOpBarriers::BarrierSet::MakeImageMemoryBarriers(const SyncValidator &sync_state, const SyncExecScope &src,
const SyncExecScope &dst, VkDependencyFlags dependencyFlags,
uint32_t barrier_count, const VkImageMemoryBarrier *barriers) {
image_memory_barriers.reserve(barrier_count);
for (uint32_t index = 0; index < barrier_count; index++) {
const auto &barrier = barriers[index];
auto image = sync_state.Get<ImageState>(barrier.image);
if (image) {
auto subresource_range = NormalizeSubresourceRange(image->createInfo, barrier.subresourceRange);
const SyncBarrier sync_barrier(barrier, src, dst);
image_memory_barriers.emplace_back(image, index, sync_barrier, barrier.oldLayout, barrier.newLayout, subresource_range);
} else {
image_memory_barriers.emplace_back();
image_memory_barriers.back().index = index; // Just in case we're interested in the ones we skipped.
}
}
}
void SyncOpBarriers::BarrierSet::MakeImageMemoryBarriers(const SyncValidator &sync_state, VkQueueFlags queue_flags,
VkDependencyFlags dependencyFlags, uint32_t barrier_count,
const VkImageMemoryBarrier2 *barriers) {
image_memory_barriers.reserve(barrier_count);
for (uint32_t index = 0; index < barrier_count; index++) {
const auto &barrier = barriers[index];
auto src = SyncExecScope::MakeSrc(queue_flags, barrier.srcStageMask);
auto dst = SyncExecScope::MakeDst(queue_flags, barrier.dstStageMask);
auto image = sync_state.Get<ImageState>(barrier.image);
if (image) {
auto subresource_range = NormalizeSubresourceRange(image->createInfo, barrier.subresourceRange);
const SyncBarrier sync_barrier(barrier, src, dst);
image_memory_barriers.emplace_back(image, index, sync_barrier, barrier.oldLayout, barrier.newLayout, subresource_range);
} else {
image_memory_barriers.emplace_back();
image_memory_barriers.back().index = index; // Just in case we're interested in the ones we skipped.
}
}
}
SyncOpWaitEvents::SyncOpWaitEvents(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags,
uint32_t eventCount, const VkEvent *pEvents, VkPipelineStageFlags srcStageMask,
VkPipelineStageFlags dstStageMask, uint32_t memoryBarrierCount,
const VkMemoryBarrier *pMemoryBarriers, uint32_t bufferMemoryBarrierCount,
const VkBufferMemoryBarrier *pBufferMemoryBarriers, uint32_t imageMemoryBarrierCount,
const VkImageMemoryBarrier *pImageMemoryBarriers)
: SyncOpBarriers(command, sync_state, queue_flags, srcStageMask, dstStageMask, VkDependencyFlags(0U), memoryBarrierCount,
pMemoryBarriers, bufferMemoryBarrierCount, pBufferMemoryBarriers, imageMemoryBarrierCount,
pImageMemoryBarriers) {
MakeEventsList(sync_state, eventCount, pEvents);
}
SyncOpWaitEvents::SyncOpWaitEvents(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags,
uint32_t eventCount, const VkEvent *pEvents, const VkDependencyInfoKHR *pDependencyInfo)
: SyncOpBarriers(command, sync_state, queue_flags, eventCount, pDependencyInfo) {
MakeEventsList(sync_state, eventCount, pEvents);
assert(events_.size() == barriers_.size()); // Just so nobody gets clever and decides to cull the event or barrier arrays
}
const char *const SyncOpWaitEvents::kIgnored = "Wait operation is ignored for this event.";
bool SyncOpWaitEvents::Validate(const CommandBufferAccessContext &cb_context) const {
bool skip = false;
const auto &sync_state = cb_context.GetSyncState();
const auto command_buffer_handle = cb_context.GetCBState().commandBuffer();
// This is only interesting at record and not replay (Execute/Submit) time.
for (size_t barrier_set_index = 0; barrier_set_index < barriers_.size(); barrier_set_index++) {
const auto &barrier_set = barriers_[barrier_set_index];
if (barrier_set.single_exec_scope) {
const Location loc(Cmd());
if (barrier_set.src_exec_scope.mask_param & VK_PIPELINE_STAGE_HOST_BIT) {
const std::string vuid = std::string("SYNC-") + std::string(CmdName()) + std::string("-hostevent-unsupported");
sync_state.LogInfo(vuid, command_buffer_handle, loc,
"srcStageMask includes %s, unsupported by synchronization validation.",
string_VkPipelineStageFlagBits(VK_PIPELINE_STAGE_HOST_BIT));
} else {
const auto &barriers = barrier_set.memory_barriers;
for (size_t barrier_index = 0; barrier_index < barriers.size(); barrier_index++) {
const auto &barrier = barriers[barrier_index];
if (barrier.src_exec_scope.mask_param & VK_PIPELINE_STAGE_HOST_BIT) {
const std::string vuid =
std::string("SYNC-") + std::string(CmdName()) + std::string("-hostevent-unsupported");
sync_state.LogInfo(vuid, command_buffer_handle, loc,
"srcStageMask %s of %s %zu, %s %zu, unsupported by synchronization validation.",
string_VkPipelineStageFlagBits(VK_PIPELINE_STAGE_HOST_BIT), "pDependencyInfo",
barrier_set_index, "pMemoryBarriers", barrier_index);
}
}
}
}
}
// The rest is common to record time and replay time.
skip |= DoValidate(cb_context, ResourceUsageRecord::kMaxIndex);
return skip;
}
bool SyncOpWaitEvents::DoValidate(const CommandExecutionContext &exec_context, const ResourceUsageTag base_tag) const {
bool skip = false;
const auto &sync_state = exec_context.GetSyncState();
const QueueId queue_id = exec_context.GetQueueId();
VkPipelineStageFlags2KHR event_stage_masks = 0U;
VkPipelineStageFlags2KHR barrier_mask_params = 0U;
bool events_not_found = false;
const auto *events_context = exec_context.GetCurrentEventsContext();
assert(events_context);
size_t barrier_set_index = 0;
size_t barrier_set_incr = (barriers_.size() == 1) ? 0 : 1;
for (const auto &event : events_) {
const auto *sync_event = events_context->Get(event.get());
const auto &barrier_set = barriers_[barrier_set_index];
if (!sync_event) {
// NOTE PHASE2: This is where we'll need queue submit time validation to come back and check the srcStageMask bits
// or solve this with replay creating the SyncEventState in the queue context... also this will be a
// new validation error... wait without previously submitted set event...
events_not_found = true; // Demote "extra_stage_bits" error to warning, to avoid false positives at *record time*
barrier_set_index += barrier_set_incr;
continue; // Core, Lifetimes, or Param check needs to catch invalid events.
}
// For replay calls, don't revalidate "same command buffer" events
if (sync_event->last_command_tag >= base_tag) continue;
const auto event_handle = sync_event->event->event();
// TODO add "destroyed" checks
if (sync_event->first_scope) {
// Only accumulate barrier and event stages if there is a pending set in the current context
barrier_mask_params |= barrier_set.src_exec_scope.mask_param;
event_stage_masks |= sync_event->scope.mask_param;
}
const auto &src_exec_scope = barrier_set.src_exec_scope;
const auto ignore_reason = sync_event->IsIgnoredByWait(command_, src_exec_scope.mask_param);
if (ignore_reason) {
switch (ignore_reason) {
case SyncEventState::ResetWaitRace:
case SyncEventState::Reset2WaitRace: {
// Four permuations of Reset and Wait calls...
const char *vuid = (command_ == vvl::Func::vkCmdWaitEvents) ? "VUID-vkCmdResetEvent-event-03834"
: "VUID-vkCmdResetEvent-event-03835";
if (ignore_reason == SyncEventState::Reset2WaitRace) {
vuid = (command_ == vvl::Func::vkCmdWaitEvents) ? "VUID-vkCmdResetEvent2-event-03831"
: "VUID-vkCmdResetEvent2-event-03832";
}
const char *const message =
"%s: %s %s operation following %s without intervening execution barrier, may cause race condition. %s";
skip |=
sync_state.LogError(event_handle, vuid, message, CmdName(), sync_state.FormatHandle(event_handle).c_str(),
CmdName(), vvl::String(sync_event->last_command), kIgnored);
break;
}
case SyncEventState::SetRace: {
// Issue error message that Wait is waiting on an signal subject to race condition, and is thus ignored for
// this event
const char *const vuid = "SYNC-vkCmdWaitEvents-unsynchronized-setops";
const char *const message =
"%s: %s Unsychronized %s calls result in race conditions w.r.t. event signalling, %s %s";
const char *const reason = "First synchronization scope is undefined.";
skip |=
sync_state.LogError(event_handle, vuid, message, CmdName(), sync_state.FormatHandle(event_handle).c_str(),
vvl::String(sync_event->last_command), reason, kIgnored);
break;
}
case SyncEventState::MissingStageBits: {
const auto missing_bits = sync_event->scope.mask_param & ~src_exec_scope.mask_param;
// Issue error message that event waited for is not in wait events scope
const char *const vuid = "VUID-vkCmdWaitEvents-srcStageMask-01158";
const char *const message = "%s: %s stageMask %" PRIx64 " includes bits not present in srcStageMask 0x%" PRIx64
". Bits missing from srcStageMask %s. %s";
skip |=
sync_state.LogError(event_handle, vuid, message, CmdName(), sync_state.FormatHandle(event_handle).c_str(),
sync_event->scope.mask_param, src_exec_scope.mask_param,
sync_utils::StringPipelineStageFlags(missing_bits).c_str(), kIgnored);
break;
}
case SyncEventState::SetVsWait2: {
skip |= sync_state.LogError(
event_handle, "VUID-vkCmdWaitEvents2-pEvents-03837", "%s: Follows set of %s by %s. Disallowed.", CmdName(),
sync_state.FormatHandle(event_handle).c_str(), vvl::String(sync_event->last_command));
break;
}
case SyncEventState::MissingSetEvent: {
// TODO: There are conditions at queue submit time where we can definitively say that
// a missing set event is an error. Add those if not captured in CoreChecks
break;
}
default:
assert(ignore_reason == SyncEventState::NotIgnored);
}
} else if (barrier_set.image_memory_barriers.size()) {
const auto &image_memory_barriers = barrier_set.image_memory_barriers;
const auto *context = exec_context.GetCurrentAccessContext();
assert(context);
for (const auto &image_memory_barrier : image_memory_barriers) {
if (image_memory_barrier.old_layout == image_memory_barrier.new_layout) continue;
const auto *image_state = image_memory_barrier.image.get();
if (!image_state) continue;
const auto &subresource_range = image_memory_barrier.range;
const auto &src_access_scope = image_memory_barrier.barrier.src_access_scope;
const auto hazard = context->DetectImageBarrierHazard(*image_state, subresource_range, sync_event->scope.exec_scope,
src_access_scope, queue_id, *sync_event,
AccessContext::DetectOptions::kDetectAll);
if (hazard.IsHazard()) {
skip |= sync_state.LogError(image_state->image(), string_SyncHazardVUID(hazard.Hazard()),
"%s: Hazard %s for image barrier %" PRIu32 " %s. Access info %s.", CmdName(),
string_SyncHazard(hazard.Hazard()), image_memory_barrier.index,
sync_state.FormatHandle(image_state->image()).c_str(),
exec_context.FormatHazard(hazard).c_str());
break;
}
}
}
// TODO: Add infrastructure for checking pDependencyInfo's vs. CmdSetEvent2 VUID - vkCmdWaitEvents2KHR - pEvents -
// 03839
barrier_set_index += barrier_set_incr;
}
// Note that we can't check for HOST in pEvents as we don't track that set event type
const auto extra_stage_bits = (barrier_mask_params & ~VK_PIPELINE_STAGE_2_HOST_BIT_KHR) & ~event_stage_masks;
if (extra_stage_bits) {
// Issue error message that event waited for is not in wait events scope
// NOTE: This isn't exactly the right VUID for WaitEvents2, but it's as close as we currently have support for
const char *const vuid = (vvl::Func::vkCmdWaitEvents == command_) ? "VUID-vkCmdWaitEvents-srcStageMask-01158"
: "VUID-vkCmdWaitEvents2-pEvents-03838";
const char *const message =
"srcStageMask 0x%" PRIx64 " contains stages not present in pEvents stageMask. Extra stages are %s.%s";
const auto handle = exec_context.Handle();
const Location loc(Cmd());
if (events_not_found) {
sync_state.LogInfo(vuid, handle, loc, message, barrier_mask_params,
sync_utils::StringPipelineStageFlags(extra_stage_bits).c_str(),
" vkCmdSetEvent may be in previously submitted command buffer.");
} else {
skip |= sync_state.LogError(vuid, handle, loc, message, barrier_mask_params,
sync_utils::StringPipelineStageFlags(extra_stage_bits).c_str(), "");
}
}
return skip;
}
struct SyncOpWaitEventsFunctorFactory {
using BarrierOpFunctor = WaitEventBarrierOp;
using ApplyFunctor = ApplyBarrierFunctor<BarrierOpFunctor>;
using GlobalBarrierOpFunctor = WaitEventBarrierOp;
using GlobalApplyFunctor = ApplyBarrierOpsFunctor<GlobalBarrierOpFunctor>;
using BufferRange = EventSimpleRangeGenerator;
using ImageRange = EventImageRangeGenerator;
using GlobalRange = EventSimpleRangeGenerator;
using ImageState = syncval_state::ImageState;
// Need to restrict to only valid exec and access scope for this event
// Pass by value is intentional to get a copy we can change without modifying the passed barrier
SyncBarrier RestrictToEvent(SyncBarrier barrier) const {
barrier.src_exec_scope.exec_scope = sync_event->scope.exec_scope & barrier.src_exec_scope.exec_scope;
barrier.src_access_scope = sync_event->scope.valid_accesses & barrier.src_access_scope;
return barrier;
}
ApplyFunctor MakeApplyFunctor(QueueId queue_id, const SyncBarrier &barrier_arg, bool layout_transition) const {
auto barrier = RestrictToEvent(barrier_arg);
return ApplyFunctor(BarrierOpFunctor(queue_id, sync_event->first_scope_tag, barrier, layout_transition));
}
GlobalApplyFunctor MakeGlobalApplyFunctor(size_t size_hint, ResourceUsageTag tag) const {
return GlobalApplyFunctor(false /* don't resolve */, size_hint, tag);
}
GlobalBarrierOpFunctor MakeGlobalBarrierOpFunctor(const QueueId queue_id, const SyncBarrier &barrier_arg) const {
auto barrier = RestrictToEvent(barrier_arg);
return GlobalBarrierOpFunctor(queue_id, sync_event->first_scope_tag, barrier, false);
}
BufferRange MakeRangeGen(const BUFFER_STATE &buffer, const ResourceAccessRange &range_arg) const {
const auto base_address = ResourceBaseAddress(buffer);
ResourceAccessRange range = SimpleBinding(buffer) ? (range_arg + base_address) : ResourceAccessRange();
EventSimpleRangeGenerator filtered_range_gen(sync_event->FirstScope(), range);
return filtered_range_gen;
}
ImageRange MakeRangeGen(const ImageState &image, const VkImageSubresourceRange &subresource_range) const {
ImageRangeGen image_range_gen = image.MakeImageRangeGen(subresource_range, false);
EventImageRangeGenerator filtered_range_gen(sync_event->FirstScope(), image_range_gen);
return filtered_range_gen;
}
GlobalRange MakeGlobalRangeGen() const { return EventSimpleRangeGenerator(sync_event->FirstScope(), kFullRange); }
SyncOpWaitEventsFunctorFactory(SyncEventState *sync_event_) : sync_event(sync_event_) { assert(sync_event); }
SyncEventState *sync_event;
};
ResourceUsageTag SyncOpWaitEvents::Record(CommandBufferAccessContext *cb_context) {
const auto tag = cb_context->NextCommandTag(command_);
ReplayRecord(*cb_context, tag);
return tag;
}
void SyncOpWaitEvents::ReplayRecord(CommandExecutionContext &exec_context, ResourceUsageTag exec_tag) const {
// Unlike PipelineBarrier, WaitEvent is *not* limited to accesses within the current subpass (if any) and thus needs to import
// all accesses. Can instead import for all first_scopes, or a union of them, if this becomes a performance/memory issue,
// but with no idea of the performance of the union, nor of whether it even matters... take the simplest approach here,
if (!exec_context.ValidForSyncOps()) return;
AccessContext *access_context = exec_context.GetCurrentAccessContext();
SyncEventsContext *events_context = exec_context.GetCurrentEventsContext();
const QueueId queue_id = exec_context.GetQueueId();
access_context->ResolvePreviousAccesses();
size_t barrier_set_index = 0;
size_t barrier_set_incr = (barriers_.size() == 1) ? 0 : 1;
assert(barriers_.size() == 1 || (barriers_.size() == events_.size()));
for (auto &event_shared : events_) {
if (!event_shared.get()) continue;
auto *sync_event = events_context->GetFromShared(event_shared);
sync_event->last_command = command_;
sync_event->last_command_tag = exec_tag;
const auto &barrier_set = barriers_[barrier_set_index];
const auto &dst = barrier_set.dst_exec_scope;
if (!sync_event->IsIgnoredByWait(command_, barrier_set.src_exec_scope.mask_param)) {
// These apply barriers one at a time as the are restricted to the resource ranges specified per each barrier,
// but do not update the dependency chain information (but set the "pending" state) // s.t. the order independence
// of the barriers is maintained.
SyncOpWaitEventsFunctorFactory factory(sync_event);
ApplyBarriers(barrier_set.buffer_memory_barriers, factory, queue_id, exec_tag, access_context);
ApplyBarriers(barrier_set.image_memory_barriers, factory, queue_id, exec_tag, access_context);
ApplyGlobalBarriers(barrier_set.memory_barriers, factory, queue_id, exec_tag, access_context);
// Apply the global barrier to the event itself (for race condition tracking)
// Events don't happen at a stage, so we need to store the unexpanded ALL_COMMANDS if set for inter-event-calls
sync_event->barriers = dst.mask_param & VK_PIPELINE_STAGE_ALL_COMMANDS_BIT;
sync_event->barriers |= dst.exec_scope;
} else {
// We ignored this wait, so we don't have any effective synchronization barriers for it.
sync_event->barriers = 0U;
}
barrier_set_index += barrier_set_incr;
}
// Apply the pending barriers
ResolvePendingBarrierFunctor apply_pending_action(exec_tag);
access_context->ApplyToContext(apply_pending_action);
}
bool SyncOpWaitEvents::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
return DoValidate(replay.GetExecutionContext(), replay.GetBaseTag() + recorded_tag);
}
bool SyncValidator::PreCallValidateCmdWriteBufferMarker2AMD(VkCommandBuffer commandBuffer, VkPipelineStageFlags2KHR pipelineStage,
VkBuffer dstBuffer, VkDeviceSize dstOffset, uint32_t marker,
const ErrorObject &error_obj) const {
bool skip = false;
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_access_context = &cb_state->access_context;
const auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
if (!context) return skip;
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(dstOffset, 4);
auto hazard = context->DetectHazard(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, range);
if (hazard.IsHazard()) {
skip |= LogError(dstBuffer, string_SyncHazardVUID(hazard.Hazard()),
"vkCmdWriteBufferMarkerAMD2: Hazard %s for dstBuffer %s. Access info %s.",
string_SyncHazard(hazard.Hazard()), FormatHandle(dstBuffer).c_str(),
cb_access_context->FormatHazard(hazard).c_str());
}
}
return skip;
}
void SyncOpWaitEvents::MakeEventsList(const SyncValidator &sync_state, uint32_t event_count, const VkEvent *events) {
events_.reserve(event_count);
for (uint32_t event_index = 0; event_index < event_count; event_index++) {
events_.emplace_back(sync_state.Get<EVENT_STATE>(events[event_index]));
}
}
SyncOpResetEvent::SyncOpResetEvent(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags, VkEvent event,
VkPipelineStageFlags2KHR stageMask)
: SyncOpBase(command),
event_(sync_state.Get<EVENT_STATE>(event)),
exec_scope_(SyncExecScope::MakeSrc(queue_flags, stageMask)) {}
bool SyncOpResetEvent::Validate(const CommandBufferAccessContext &cb_context) const {
return DoValidate(cb_context, ResourceUsageRecord::kMaxIndex);
}
bool SyncOpResetEvent::DoValidate(const CommandExecutionContext &exec_context, const ResourceUsageTag base_tag) const {
auto *events_context = exec_context.GetCurrentEventsContext();
assert(events_context);
bool skip = false;
if (!events_context) return skip;
const auto &sync_state = exec_context.GetSyncState();
const auto *sync_event = events_context->Get(event_);
if (!sync_event) return skip; // Core, Lifetimes, or Param check needs to catch invalid events.
if (sync_event->last_command_tag > base_tag) return skip; // if we validated this in recording of the secondary, don't repeat
const char *const set_wait =
"%s: %s %s operation following %s without intervening execution barrier, is a race condition and may result in data "
"hazards.";
const char *message = set_wait; // Only one message this call.
if (!sync_event->HasBarrier(exec_scope_.mask_param, exec_scope_.exec_scope)) {
const char *vuid = nullptr;
switch (sync_event->last_command) {
case vvl::Func::vkCmdSetEvent:
case vvl::Func::vkCmdSetEvent2KHR:
case vvl::Func::vkCmdSetEvent2:
// Needs a barrier between set and reset
vuid = "SYNC-vkCmdResetEvent-missingbarrier-set";
break;
case vvl::Func::vkCmdWaitEvents:
case vvl::Func::vkCmdWaitEvents2KHR:
case vvl::Func::vkCmdWaitEvents2: {
// Needs to be in the barriers chain (either because of a barrier, or because of dstStageMask
vuid = "SYNC-vkCmdResetEvent-missingbarrier-wait";
break;
}
default:
// The only other valid last command that wasn't one.
assert((sync_event->last_command == vvl::Func::Empty) || (sync_event->last_command == vvl::Func::vkCmdResetEvent) ||
(sync_event->last_command == vvl::Func::vkCmdResetEvent2KHR));
break;
}
if (vuid) {
skip |= sync_state.LogError(event_->event(), vuid, message, CmdName(), sync_state.FormatHandle(event_->event()).c_str(),
CmdName(), vvl::String(sync_event->last_command));
}
}
return skip;
}
ResourceUsageTag SyncOpResetEvent::Record(CommandBufferAccessContext *cb_context) {
const auto tag = cb_context->NextCommandTag(command_);
ReplayRecord(*cb_context, tag);
return tag;
}
bool SyncOpResetEvent::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
return DoValidate(replay.GetExecutionContext(), replay.GetBaseTag() + recorded_tag);
}
void SyncOpResetEvent::ReplayRecord(CommandExecutionContext &exec_context, ResourceUsageTag exec_tag) const {
if (!exec_context.ValidForSyncOps()) return;
SyncEventsContext *events_context = exec_context.GetCurrentEventsContext();
auto *sync_event = events_context->GetFromShared(event_);
if (!sync_event) return; // Core, Lifetimes, or Param check needs to catch invalid events.
// Update the event state
sync_event->last_command = command_;
sync_event->last_command_tag = exec_tag;
sync_event->unsynchronized_set = vvl::Func::Empty;
sync_event->ResetFirstScope();
sync_event->barriers = 0U;
}
SyncOpSetEvent::SyncOpSetEvent(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags, VkEvent event,
VkPipelineStageFlags2KHR stageMask, const AccessContext *access_context)
: SyncOpBase(command),
event_(sync_state.Get<EVENT_STATE>(event)),
recorded_context_(),
src_exec_scope_(SyncExecScope::MakeSrc(queue_flags, stageMask)),
dep_info_() {
// Snapshot the current access_context for later inspection at wait time.
// NOTE: This appears brute force, but given that we only save a "first-last" model of access history, the current
// access context (include barrier state for chaining) won't necessarily contain the needed information at Wait
// or Submit time reference.
if (access_context) {
recorded_context_ = std::make_shared<const AccessContext>(*access_context);
}
}
SyncOpSetEvent::SyncOpSetEvent(vvl::Func command, const SyncValidator &sync_state, VkQueueFlags queue_flags, VkEvent event,
const VkDependencyInfoKHR &dep_info, const AccessContext *access_context)
: SyncOpBase(command),
event_(sync_state.Get<EVENT_STATE>(event)),
recorded_context_(),
src_exec_scope_(SyncExecScope::MakeSrc(queue_flags, sync_utils::GetGlobalStageMasks(dep_info).src)),
dep_info_(new safe_VkDependencyInfo(&dep_info)) {
if (access_context) {
recorded_context_ = std::make_shared<const AccessContext>(*access_context);
}
}
bool SyncOpSetEvent::Validate(const CommandBufferAccessContext &cb_context) const {
return DoValidate(cb_context, ResourceUsageRecord::kMaxIndex);
}
bool SyncOpSetEvent::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
return DoValidate(replay.GetExecutionContext(), replay.GetBaseTag() + recorded_tag);
}
bool SyncOpSetEvent::DoValidate(const CommandExecutionContext &exec_context, const ResourceUsageTag base_tag) const {
bool skip = false;
const auto &sync_state = exec_context.GetSyncState();
auto *events_context = exec_context.GetCurrentEventsContext();
assert(events_context);
if (!events_context) return skip;
const auto *sync_event = events_context->Get(event_);
if (!sync_event) return skip; // Core, Lifetimes, or Param check needs to catch invalid events.
if (sync_event->last_command_tag >= base_tag) return skip; // for replay we don't want to revalidate internal "last commmand"
const char *const reset_set =
"%s: %s %s operation following %s without intervening execution barrier, is a race condition and may result in data "
"hazards.";
const char *const wait =
"%s: %s %s operation following %s without intervening vkCmdResetEvent, may result in data hazard and is ignored.";
if (!sync_event->HasBarrier(src_exec_scope_.mask_param, src_exec_scope_.exec_scope)) {
const char *vuid_stem = nullptr;
const char *message = nullptr;
switch (sync_event->last_command) {
case vvl::Func::vkCmdResetEvent:
case vvl::Func::vkCmdResetEvent2KHR:
case vvl::Func::vkCmdResetEvent2:
// Needs a barrier between reset and set
vuid_stem = "-missingbarrier-reset";
message = reset_set;
break;
case vvl::Func::vkCmdSetEvent:
case vvl::Func::vkCmdSetEvent2KHR:
case vvl::Func::vkCmdSetEvent2:
// Needs a barrier between set and set
vuid_stem = "-missingbarrier-set";
message = reset_set;
break;
case vvl::Func::vkCmdWaitEvents:
case vvl::Func::vkCmdWaitEvents2KHR:
case vvl::Func::vkCmdWaitEvents2:
// Needs a barrier or is in second execution scope
vuid_stem = "-missingbarrier-wait";
message = wait;
break;
default:
// The only other valid last command that wasn't one.
assert(sync_event->last_command == vvl::Func::Empty);
break;
}
if (vuid_stem) {
assert(nullptr != message);
std::string vuid("SYNC-");
vuid.append(CmdName()).append(vuid_stem);
skip |= sync_state.LogError(event_->event(), vuid.c_str(), message, CmdName(),
sync_state.FormatHandle(event_->event()).c_str(), CmdName(),
vvl::String(sync_event->last_command));
}
}
return skip;
}
ResourceUsageTag SyncOpSetEvent::Record(CommandBufferAccessContext *cb_context) {
const auto tag = cb_context->NextCommandTag(command_);
auto *events_context = cb_context->GetCurrentEventsContext();
const QueueId queue_id = cb_context->GetQueueId();
assert(recorded_context_);
if (recorded_context_ && events_context) {
DoRecord(queue_id, tag, recorded_context_, events_context);
}
return tag;
}
void SyncOpSetEvent::ReplayRecord(CommandExecutionContext &exec_context, ResourceUsageTag exec_tag) const {
// Create a copy of the current context, and merge in the state snapshot at record set event time
// Note: we mustn't change the recorded context copy, as a given CB could be submitted more than once (in generaL)
if (!exec_context.ValidForSyncOps()) return;
SyncEventsContext *events_context = exec_context.GetCurrentEventsContext();
AccessContext *access_context = exec_context.GetCurrentAccessContext();
const QueueId queue_id = exec_context.GetQueueId();
// Note: merged_context is a copy of the access_context, combined with the recorded context
auto merged_context = std::make_shared<AccessContext>(*access_context);
merged_context->ResolveFromContext(QueueTagOffsetBarrierAction(queue_id, exec_tag), *recorded_context_);
merged_context->Trim(); // Ensure the copy is minimal and normalized
DoRecord(queue_id, exec_tag, merged_context, events_context);
}
void SyncOpSetEvent::DoRecord(QueueId queue_id, ResourceUsageTag tag, const std::shared_ptr<const AccessContext> &access_context,
SyncEventsContext *events_context) const {
auto *sync_event = events_context->GetFromShared(event_);
if (!sync_event) return; // Core, Lifetimes, or Param check needs to catch invalid events.
// NOTE: We're going to simply record the sync scope here, as anything else would be implementation defined/undefined
// and we're issuing errors re: missing barriers between event commands, which if the user fixes would fix
// any issues caused by naive scope setting here.
// What happens with two SetEvent is that one cannot know what group of operations will be waited for.
// Given:
// Stuff1; SetEvent; Stuff2; SetEvent; WaitEvents;
// WaitEvents cannot know which of Stuff1, Stuff2, or both has completed execution.
if (!sync_event->HasBarrier(src_exec_scope_.mask_param, src_exec_scope_.exec_scope)) {
sync_event->unsynchronized_set = sync_event->last_command;
sync_event->ResetFirstScope();
} else if (!sync_event->first_scope) {
// We only set the scope if there isn't one
sync_event->scope = src_exec_scope_;
// Save the shared_ptr to copy of the access_context present at set time (sent us by the caller)
sync_event->first_scope = access_context;
sync_event->unsynchronized_set = vvl::Func::Empty;
sync_event->first_scope_tag = tag;
}
// TODO: Store dep_info_ shared ptr in sync_state for WaitEvents2 validation
sync_event->last_command = command_;
sync_event->last_command_tag = tag;
sync_event->barriers = 0U;
}
SyncOpBeginRenderPass::SyncOpBeginRenderPass(vvl::Func command, const SyncValidator &sync_state,
const VkRenderPassBeginInfo *pRenderPassBegin,
const VkSubpassBeginInfo *pSubpassBeginInfo)
: SyncOpBase(command), rp_context_(nullptr) {
if (pRenderPassBegin) {
rp_state_ = sync_state.Get<RENDER_PASS_STATE>(pRenderPassBegin->renderPass);
renderpass_begin_info_ = safe_VkRenderPassBeginInfo(pRenderPassBegin);
auto fb_state = sync_state.Get<FRAMEBUFFER_STATE>(pRenderPassBegin->framebuffer);
if (fb_state) {
shared_attachments_ = sync_state.GetAttachmentViews(*renderpass_begin_info_.ptr(), *fb_state);
// TODO: Revisit this when all attachment validation is through SyncOps to see if we can discard the plain pointer copy
// Note that this a safe to presist as long as shared_attachments is not cleared
attachments_.reserve(shared_attachments_.size());
for (const auto &attachment : shared_attachments_) {
attachments_.emplace_back(static_cast<const syncval_state::ImageViewState *>(attachment.get()));
}
}
if (pSubpassBeginInfo) {
subpass_begin_info_ = safe_VkSubpassBeginInfo(pSubpassBeginInfo);
}
}
}
bool SyncOpBeginRenderPass::Validate(const CommandBufferAccessContext &cb_context) const {
// Check if any of the layout transitions are hazardous.... but we don't have the renderpass context to work with, so we
bool skip = false;
assert(rp_state_.get());
if (nullptr == rp_state_.get()) return skip;
auto &rp_state = *rp_state_.get();
const uint32_t subpass = 0;
// Construct the state we can use to validate against... (since validation is const and RecordCmdBeginRenderPass
// hasn't happened yet)
const std::vector<AccessContext> empty_context_vector;
AccessContext temp_context(subpass, cb_context.GetQueueFlags(), rp_state.subpass_dependencies, empty_context_vector,
cb_context.GetCurrentAccessContext());
// Validate attachment operations
if (attachments_.size() == 0) return skip;
const auto &render_area = renderpass_begin_info_.renderArea;
// Since the isn't a valid RenderPassAccessContext until Record, needs to create the view/generator list... we could limit this
// by predicating on whether subpass 0 uses the attachment if it is too expensive to create the full list redundantly here.
// More broadly we could look at thread specific state shared between Validate and Record as is done for other heavyweight
// operations (though it's currently a messy approach)
AttachmentViewGenVector view_gens = RenderPassAccessContext::CreateAttachmentViewGen(render_area, attachments_);
skip |= temp_context.ValidateLayoutTransitions(cb_context, rp_state, render_area, subpass, view_gens, command_);
// Validate load operations if there were no layout transition hazards
if (!skip) {
temp_context.RecordLayoutTransitions(rp_state, subpass, view_gens, kInvalidTag);
skip |= temp_context.ValidateLoadOperation(cb_context, rp_state, render_area, subpass, view_gens, command_);
}
return skip;
}
ResourceUsageTag SyncOpBeginRenderPass::Record(CommandBufferAccessContext *cb_context) {
assert(rp_state_.get());
if (nullptr == rp_state_.get()) return cb_context->NextCommandTag(command_);
const ResourceUsageTag begin_tag =
cb_context->RecordBeginRenderPass(command_, *rp_state_.get(), renderpass_begin_info_.renderArea, attachments_);
// Note: this state update must be after RecordBeginRenderPass as there is no current render pass until that function runs
rp_context_ = cb_context->GetCurrentRenderPassContext();
return begin_tag;
}
bool SyncOpBeginRenderPass::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
// Need to update the exec_contexts state (which for RenderPass operations *must* be a QueueBatchContext, as
// render pass operations are not allowed in secondary command buffers.
replay.BeginRenderPassReplaySetup(*this);
// Only the layout transitions happen at the replay tag, loadOp's happen at a subsequent tag
ResourceUsageRange first_use_range = {recorded_tag, recorded_tag + 1};
return replay.DetectFirstUseHazard(first_use_range);
}
void SyncOpBeginRenderPass::ReplayRecord(CommandExecutionContext &exec_context, ResourceUsageTag exec_tag) const {
// All the needed replay state changes (for the layout transition, and context update) have to happen in ReplayValidate
}
SyncOpNextSubpass::SyncOpNextSubpass(vvl::Func command, const SyncValidator &sync_state,
const VkSubpassBeginInfo *pSubpassBeginInfo, const VkSubpassEndInfo *pSubpassEndInfo)
: SyncOpBase(command) {
if (pSubpassBeginInfo) {
subpass_begin_info_.initialize(pSubpassBeginInfo);
}
if (pSubpassEndInfo) {
subpass_end_info_.initialize(pSubpassEndInfo);
}
}
bool SyncOpNextSubpass::Validate(const CommandBufferAccessContext &cb_context) const {
bool skip = false;
const auto *renderpass_context = cb_context.GetCurrentRenderPassContext();
if (!renderpass_context) return skip;
skip |= renderpass_context->ValidateNextSubpass(cb_context.GetExecutionContext(), command_);
return skip;
}
ResourceUsageTag SyncOpNextSubpass::Record(CommandBufferAccessContext *cb_context) {
return cb_context->RecordNextSubpass(command_);
}
bool SyncOpNextSubpass::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
// Any store/resolve operations happen before the NextSubpass tag so we can advance to the next subpass state
replay.NextSubpassReplaySetup();
// Only the layout transitions happen at the replay tag, loadOp's happen at a subsequent tag
ResourceUsageRange first_use_range = {recorded_tag, recorded_tag + 1};
return replay.DetectFirstUseHazard(first_use_range);
}
void SyncOpNextSubpass::ReplayRecord(CommandExecutionContext &exec_context, ResourceUsageTag exec_tag) const {
// All the needed replay state changes (for the layout transition, and context update) have to happen in ReplayValidate
}
SyncOpEndRenderPass::SyncOpEndRenderPass(vvl::Func command, const SyncValidator &sync_state,
const VkSubpassEndInfo *pSubpassEndInfo)
: SyncOpBase(command) {
if (pSubpassEndInfo) {
subpass_end_info_.initialize(pSubpassEndInfo);
}
}
bool SyncOpEndRenderPass::Validate(const CommandBufferAccessContext &cb_context) const {
bool skip = false;
const auto *renderpass_context = cb_context.GetCurrentRenderPassContext();
if (!renderpass_context) return skip;
skip |= renderpass_context->ValidateEndRenderPass(cb_context.GetExecutionContext(), command_);
return skip;
}
ResourceUsageTag SyncOpEndRenderPass::Record(CommandBufferAccessContext *cb_context) {
return cb_context->RecordEndRenderPass(command_);
}
bool SyncOpEndRenderPass::ReplayValidate(ReplayState &replay, ResourceUsageTag recorded_tag) const {
// Any store/resolve operations happen before the EndRenderPass tag so we can ignore them
// Only the layout transitions happen at the replay tag
ResourceUsageRange first_use_range = {recorded_tag, recorded_tag + 1};
bool skip = replay.DetectFirstUseHazard(first_use_range);
// We can cleanup here as the recorded tag represents the final layout transition (which is the last operation or the RP
replay.EndRenderPassReplayCleanup();
return skip;
}
void SyncOpEndRenderPass::ReplayRecord(CommandExecutionContext &exec_context, ResourceUsageTag exec_tag) const {
}
void SyncValidator::PreCallRecordCmdWriteBufferMarker2AMD(VkCommandBuffer commandBuffer, VkPipelineStageFlags2KHR pipelineStage,
VkBuffer dstBuffer, VkDeviceSize dstOffset, uint32_t marker) {
StateTracker::PreCallRecordCmdWriteBufferMarker2AMD(commandBuffer, pipelineStage, dstBuffer, dstOffset, marker);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_access_context = &cb_state->access_context;
const auto tag = cb_access_context->NextCommandTag(Func::vkCmdWriteBufferMarker2AMD);
auto *context = cb_access_context->GetCurrentAccessContext();
assert(context);
auto dst_buffer = Get<BUFFER_STATE>(dstBuffer);
if (dst_buffer) {
const ResourceAccessRange range = MakeRange(dstOffset, 4);
context->UpdateAccessState(*dst_buffer, SYNC_COPY_TRANSFER_WRITE, SyncOrdering::kNonAttachment, range, tag);
}
}
bool SyncValidator::PreCallValidateCmdExecuteCommands(VkCommandBuffer commandBuffer, uint32_t commandBufferCount,
const VkCommandBuffer *pCommandBuffers, const ErrorObject &error_obj) const {
bool skip = StateTracker::PreCallValidateCmdExecuteCommands(commandBuffer, commandBufferCount, pCommandBuffers, error_obj);
const auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return skip;
const auto *cb_context = &cb_state->access_context;
// Heavyweight, but we need a proxy copy of the active command buffer access context
CommandBufferAccessContext proxy_cb_context(*cb_context, CommandBufferAccessContext::AsProxyContext());
// Make working copies of the access and events contexts
for (uint32_t cb_index = 0; cb_index < commandBufferCount; ++cb_index) {
proxy_cb_context.NextIndexedCommandTag(error_obj.location.function, cb_index);
const auto recorded_cb = Get<syncval_state::CommandBuffer>(pCommandBuffers[cb_index]);
if (!recorded_cb) continue;
const auto *recorded_cb_context = &recorded_cb->access_context;
assert(recorded_cb_context);
skip |= ReplayState(proxy_cb_context, *recorded_cb_context, error_obj, cb_index).ValidateFirstUse();
// The barriers have already been applied in ValidatFirstUse
ResourceUsageRange tag_range = proxy_cb_context.ImportRecordedAccessLog(*recorded_cb_context);
proxy_cb_context.ResolveExecutedCommandBuffer(*recorded_cb_context->GetCurrentAccessContext(), tag_range.begin);
}
return skip;
}
void SyncValidator::PreCallRecordCmdExecuteCommands(VkCommandBuffer commandBuffer, uint32_t commandBufferCount,
const VkCommandBuffer *pCommandBuffers) {
StateTracker::PreCallRecordCmdExecuteCommands(commandBuffer, commandBufferCount, pCommandBuffers);
auto cb_state = Get<syncval_state::CommandBuffer>(commandBuffer);
assert(cb_state);
if (!cb_state) return;
auto *cb_context = &cb_state->access_context;
for (uint32_t cb_index = 0; cb_index < commandBufferCount; ++cb_index) {
const ResourceUsageTag cb_tag = cb_context->NextIndexedCommandTag(vvl::Func::vkCmdExecuteCommands, cb_index);
const auto recorded_cb = Get<syncval_state::CommandBuffer>(pCommandBuffers[cb_index]);
if (!recorded_cb) continue;
cb_context->AddHandle(cb_tag, "pCommandBuffers", recorded_cb->Handle(), cb_index);
const auto *recorded_cb_context = &recorded_cb->access_context;
cb_context->RecordExecutedCommandBuffer(*recorded_cb_context);
}
}
void SyncValidator::PostCallRecordBindImageMemory(VkDevice device, VkImage image, VkDeviceMemory mem, VkDeviceSize memoryOffset,
const RecordObject &record_obj) {
StateTracker::PostCallRecordBindImageMemory(device, image, mem, memoryOffset, record_obj);
if (VK_SUCCESS != record_obj.result) return;
const VkBindImageMemoryInfo bind_info = ConvertImageMemoryInfo(device, image, mem, memoryOffset);
UpdateSyncImageMemoryBindState(1, &bind_info);
}
void SyncValidator::PostCallRecordBindImageMemory2(VkDevice device, uint32_t bindInfoCount, const VkBindImageMemoryInfo *pBindInfos,
const RecordObject &record_obj) {
StateTracker::PostCallRecordBindImageMemory2(device, bindInfoCount, pBindInfos, record_obj);
if (VK_SUCCESS != record_obj.result) return;
UpdateSyncImageMemoryBindState(bindInfoCount, pBindInfos);
}
void SyncValidator::PostCallRecordBindImageMemory2KHR(VkDevice device, uint32_t bindInfoCount,
const VkBindImageMemoryInfo *pBindInfos, const RecordObject &record_obj) {
StateTracker::PostCallRecordBindImageMemory2KHR(device, bindInfoCount, pBindInfos, record_obj);
if (VK_SUCCESS != record_obj.result) return;
UpdateSyncImageMemoryBindState(bindInfoCount, pBindInfos);
}
void SyncValidator::PostCallRecordQueueWaitIdle(VkQueue queue, const RecordObject &record_obj) {
StateTracker::PostCallRecordQueueWaitIdle(queue, record_obj);
if ((record_obj.result != VK_SUCCESS) || (!enabled[sync_validation_queue_submit]) || (queue == VK_NULL_HANDLE)) return;
const auto queue_state = GetQueueSyncStateShared(queue);
if (!queue_state) return; // Invalid queue
QueueId waited_queue = queue_state->GetQueueId();
ApplyTaggedWait(waited_queue, ResourceUsageRecord::kMaxIndex);
// Eliminate waitable fences from the current queue.
vvl::EraseIf(waitable_fences_, [waited_queue](const SignaledFence &sf) { return sf.second.queue_id == waited_queue; });
}
void SyncValidator::PostCallRecordDeviceWaitIdle(VkDevice device, const RecordObject &record_obj) {
StateTracker::PostCallRecordDeviceWaitIdle(device, record_obj);
// We need to treat this a fence waits for all queues... noting that present engine ops will be preserved.
ForAllQueueBatchContexts([](const std::shared_ptr<QueueBatchContext> &batch) {
batch->ApplyTaggedWait(QueueSyncState::kQueueAny, ResourceUsageRecord::kMaxIndex);
});
// As we we've waited for everything on device, any waits are mooted. (except for acquires)
vvl::EraseIf(waitable_fences_, [](SignaledFences::value_type &waitable) { return waitable.second.acquired.Invalid(); });
}
struct QueuePresentCmdState {
std::shared_ptr<const QueueSyncState> queue;
std::shared_ptr<QueueBatchContext> present_batch;
SignaledSemaphores signaled;
PresentedImages presented_images;
QueuePresentCmdState(const SignaledSemaphores &parent_semaphores) : signaled(parent_semaphores) {}
};
bool SyncValidator::PreCallValidateQueuePresentKHR(VkQueue queue, const VkPresentInfoKHR *pPresentInfo,
const ErrorObject &error_obj) const {
bool skip = false;
// Since this early return is above the TlsGuard, the Record phase must also be.
if (!enabled[sync_validation_queue_submit]) return skip;
vvl::TlsGuard<QueuePresentCmdState> cmd_state(&skip, signaled_semaphores_);
cmd_state->queue = GetQueueSyncStateShared(queue);
if (!cmd_state->queue) return skip; // Invalid Queue
// The submit id is a mutable automic which is not recoverable on a skip == true condition
uint64_t submit_id = cmd_state->queue->ReserveSubmitId();
std::shared_ptr<const QueueBatchContext> last_batch = cmd_state->queue->LastBatch();
std::shared_ptr<QueueBatchContext> batch(std::make_shared<QueueBatchContext>(*this, *cmd_state->queue, submit_id, 0));
ResourceUsageRange tag_range = SetupPresentInfo(*pPresentInfo, batch, cmd_state->presented_images);
batch->SetupAccessContext(last_batch, *pPresentInfo, cmd_state->presented_images, cmd_state->signaled);
batch->SetupBatchTags(tag_range);
// Update the present tags
for (auto &presented : cmd_state->presented_images) {
presented.tag += batch->GetTagRange().begin;
}
skip |= batch->DoQueuePresentValidate(error_obj.location, cmd_state->presented_images);
batch->DoPresentOperations(cmd_state->presented_images);
batch->LogPresentOperations(cmd_state->presented_images);
batch->Cleanup();
if (!skip) {
cmd_state->present_batch = std::move(batch);
}
return skip;
}
ResourceUsageRange SyncValidator::SetupPresentInfo(const VkPresentInfoKHR &present_info, std::shared_ptr<QueueBatchContext> &batch,
PresentedImages &presented_images) const {
const VkSwapchainKHR *const swapchains = present_info.pSwapchains;
const uint32_t *const image_indices = present_info.pImageIndices;
const uint32_t swap_count = present_info.swapchainCount;
// Create the working list of presented images
presented_images.reserve(swap_count);
for (uint32_t present_index = 0; present_index < swap_count; present_index++) {
// Note: Given the "EraseIf" implementation for acquire fence waits, each presentation needs a unique tag.
const ResourceUsageTag tag = presented_images.size();
presented_images.emplace_back(*this, batch, swapchains[present_index], image_indices[present_index], present_index, tag);
if (presented_images.back().Invalid()) {
presented_images.pop_back();
}
}
// Present is tagged for each swap.
return ResourceUsageRange(0, presented_images.size());
}
void SyncValidator::PostCallRecordQueuePresentKHR(VkQueue queue, const VkPresentInfoKHR *pPresentInfo,
const RecordObject &record_obj) {
StateTracker::PostCallRecordQueuePresentKHR(queue, pPresentInfo, record_obj);
if (!enabled[sync_validation_queue_submit]) return;
// The earliest return (when enabled), must be *after* the TlsGuard, as it is the TlsGuard that cleans up the cmd_state
// static payload
vvl::TlsGuard<QueuePresentCmdState> cmd_state;
// See ValidationStateTracker::PostCallRecordQueuePresentKHR for spec excerpt supporting
if (record_obj.result == VK_ERROR_OUT_OF_HOST_MEMORY || record_obj.result == VK_ERROR_OUT_OF_DEVICE_MEMORY ||
record_obj.result == VK_ERROR_DEVICE_LOST) {
return;
}
// Update the state with the data from the validate phase
cmd_state->signaled.Resolve(signaled_semaphores_, cmd_state->present_batch);
std::shared_ptr<QueueSyncState> queue_state = std::const_pointer_cast<QueueSyncState>(std::move(cmd_state->queue));
for (auto &presented : cmd_state->presented_images) {
presented.ExportToSwapchain(*this);
}
queue_state->UpdateLastBatch(std::move(cmd_state->present_batch));
}
void SyncValidator::PostCallRecordAcquireNextImageKHR(VkDevice device, VkSwapchainKHR swapchain, uint64_t timeout,
VkSemaphore semaphore, VkFence fence, uint32_t *pImageIndex,
const RecordObject &record_obj) {
StateTracker::PostCallRecordAcquireNextImageKHR(device, swapchain, timeout, semaphore, fence, pImageIndex, record_obj);
if (!enabled[sync_validation_queue_submit]) return;
RecordAcquireNextImageState(device, swapchain, timeout, semaphore, fence, pImageIndex, record_obj);
}
void SyncValidator::PostCallRecordAcquireNextImage2KHR(VkDevice device, const VkAcquireNextImageInfoKHR *pAcquireInfo,
uint32_t *pImageIndex, const RecordObject &record_obj) {
StateTracker::PostCallRecordAcquireNextImage2KHR(device, pAcquireInfo, pImageIndex, record_obj);
if (!enabled[sync_validation_queue_submit]) return;
RecordAcquireNextImageState(device, pAcquireInfo->swapchain, pAcquireInfo->timeout, pAcquireInfo->semaphore,
pAcquireInfo->fence, pImageIndex, record_obj);
}
void SyncValidator::RecordAcquireNextImageState(VkDevice device, VkSwapchainKHR swapchain, uint64_t timeout, VkSemaphore semaphore,
VkFence fence, uint32_t *pImageIndex, const RecordObject &record_obj) {
if ((VK_SUCCESS != record_obj.result) && (VK_SUBOPTIMAL_KHR != record_obj.result)) return;
// Get the image out of the presented list and create apppropriate fences/semaphores.
auto swapchain_state = Get<syncval_state::Swapchain>(swapchain);
if (BASE_NODE::Invalid(swapchain_state)) return; // Invalid acquire calls to be caught in CoreCheck/Parameter validation
PresentedImage presented = swapchain_state->MovePresentedImage(*pImageIndex);
if (presented.Invalid()) return;
// No way to make access safe, so nothing to record
if ((semaphore == VK_NULL_HANDLE) && (fence == VK_NULL_HANDLE)) return;
// We create a queue-less QBC for the Semaphore and fences to wait on
// Note: this is a heavyweight way to deal with the fact that all operation logs live in the QueueBatchContext... and
// acquire doesn't happen on a queue, but we need a place to put the acquire operation access record.
auto batch = std::make_shared<QueueBatchContext>(*this);
batch->SetupAccessContext(presented);
ResourceUsageRange acquire_tag_range(0, 1);
batch->SetupBatchTags(ResourceUsageRange(0, 1));
const ResourceUsageTag acquire_tag = batch->GetTagRange().begin;
batch->DoAcquireOperation(presented);
batch->LogAcquireOperation(presented, record_obj.location.function);
// Now swap out the present queue batch with the acquired one.
// Note that fence and signal will read the acquire batch from presented, so this needs to be done before
// setting up the synchronization
presented.batch = std::move(batch);
if (semaphore != VK_NULL_HANDLE) {
std::shared_ptr<const SEMAPHORE_STATE> sem_state = Get<SEMAPHORE_STATE>(semaphore);
if (bool(sem_state)) {
signaled_semaphores_.SignalSemaphore(sem_state, presented, acquire_tag);
}
}
if (fence != VK_NULL_HANDLE) {
UpdateFenceWaitInfo(fence, presented, acquire_tag);
}
}
bool SyncValidator::PreCallValidateQueueSubmit(VkQueue queue, uint32_t submitCount, const VkSubmitInfo *pSubmits, VkFence fence,
const ErrorObject &error_obj) const {
auto queue_state = GetQueueSyncStateShared(queue);
if (!bool(queue_state)) return false;
SubmitInfoConverter submit_info(submitCount, pSubmits, queue_state->GetQueueFlags());
return ValidateQueueSubmit(queue, submitCount, submit_info.info2s.data(), fence, error_obj);
}
bool SyncValidator::ValidateQueueSubmit(VkQueue queue, uint32_t submitCount, const VkSubmitInfo2 *pSubmits, VkFence fence,
const ErrorObject &error_obj) const {
bool skip = false;
// Since this early return is above the TlsGuard, the Record phase must also be.
if (!enabled[sync_validation_queue_submit]) return skip;
vvl::TlsGuard<QueueSubmitCmdState> cmd_state(&skip, error_obj, signaled_semaphores_);
cmd_state->queue = GetQueueSyncStateShared(queue);
if (!cmd_state->queue) return skip; // Invalid Queue
// The submit id is a mutable automic which is not recoverable on a skip == true condition
uint64_t submit_id = cmd_state->queue->ReserveSubmitId();
// verify each submit batch
// Since the last batch from the queue state is const, we need to track the last_batch separately from the
// most recently created batch
std::shared_ptr<const QueueBatchContext> last_batch = cmd_state->queue->LastBatch();
std::shared_ptr<QueueBatchContext> batch;
for (uint32_t batch_idx = 0; batch_idx < submitCount; batch_idx++) {
const VkSubmitInfo2 &submit = pSubmits[batch_idx];
batch = std::make_shared<QueueBatchContext>(*this, *cmd_state->queue, submit_id, batch_idx);
batch->SetupCommandBufferInfo(submit);
batch->SetupAccessContext(last_batch, submit, cmd_state->signaled);
// Skip import and validation of empty batches
if (batch->GetTagRange().size()) {
batch->SetupBatchTags();
skip |= batch->DoQueueSubmitValidate(*this, *cmd_state, submit);
}
// Empty batches could have semaphores, though.
for (uint32_t sem_idx = 0; sem_idx < submit.signalSemaphoreInfoCount; ++sem_idx) {
const VkSemaphoreSubmitInfo &semaphore_info = submit.pSignalSemaphoreInfos[sem_idx];
// Make a copy of the state, signal the copy and pend it...
auto sem_state = Get<SEMAPHORE_STATE>(semaphore_info.semaphore);
if (!sem_state) continue;
cmd_state->signaled.SignalSemaphore(sem_state, batch, semaphore_info);
}
// Unless the previous batch was referenced by a signal, the QueueBatchContext will self destruct, but as
// we ResolvePrevious as we can let any contexts we've fully referenced go.
batch->Cleanup(); // Clear the temporaries that the batch holds.
last_batch = batch;
}
// The most recently created batch will become the queue's "last batch" in the record phase
if (batch) {
cmd_state->last_batch = std::move(batch);
}
// Note that if we skip, guard cleans up for us, but cannot release the reserved tag range
return skip;
}
void SyncValidator::PostCallRecordQueueSubmit(VkQueue queue, uint32_t submitCount, const VkSubmitInfo *pSubmits, VkFence fence,
const RecordObject &record_obj) {
StateTracker::PostCallRecordQueueSubmit(queue, submitCount, pSubmits, fence, record_obj);
RecordQueueSubmit(queue, fence, record_obj);
}
void SyncValidator::RecordQueueSubmit(VkQueue queue, VkFence fence, const RecordObject &record_obj) {
// If this return is above the TlsGuard, then the Validate phase return must also be.
if (!enabled[sync_validation_queue_submit]) return; // Queue submit validation must be affirmatively enabled
// The earliest return (when enabled), must be *after* the TlsGuard, as it is the TlsGuard that cleans up the cmd_state
// static payload
vvl::TlsGuard<QueueSubmitCmdState> cmd_state;
if (VK_SUCCESS != record_obj.result) return; // dispatched QueueSubmit failed
if (!cmd_state->queue) return; // Validation couldn't find a valid queue object
// Don't need to look up the queue state again, but we need a non-const version
std::shared_ptr<QueueSyncState> queue_state = std::const_pointer_cast<QueueSyncState>(std::move(cmd_state->queue));
cmd_state->signaled.Resolve(signaled_semaphores_, cmd_state->last_batch);
queue_state->UpdateLastBatch(std::move(cmd_state->last_batch));
ResourceUsageRange fence_tag_range = ReserveGlobalTagRange(1U);
UpdateFenceWaitInfo(fence, queue_state->GetQueueId(), fence_tag_range.begin);
}
bool SyncValidator::PreCallValidateQueueSubmit2KHR(VkQueue queue, uint32_t submitCount, const VkSubmitInfo2KHR *pSubmits,
VkFence fence, const ErrorObject &error_obj) const {
return PreCallValidateQueueSubmit2(queue, submitCount, pSubmits, fence, error_obj);
}
bool SyncValidator::PreCallValidateQueueSubmit2(VkQueue queue, uint32_t submitCount, const VkSubmitInfo2KHR *pSubmits,
VkFence fence, const ErrorObject &error_obj) const {
return ValidateQueueSubmit(queue, submitCount, pSubmits, fence, error_obj);
}
void SyncValidator::PostCallRecordQueueSubmit2KHR(VkQueue queue, uint32_t submitCount, const VkSubmitInfo2KHR *pSubmits,
VkFence fence, const RecordObject &record_obj) {
StateTracker::PostCallRecordQueueSubmit2KHR(queue, submitCount, pSubmits, fence, record_obj);
RecordQueueSubmit(queue, fence, record_obj);
}
void SyncValidator::PostCallRecordQueueSubmit2(VkQueue queue, uint32_t submitCount, const VkSubmitInfo2KHR *pSubmits, VkFence fence,
const RecordObject &record_obj) {
StateTracker::PostCallRecordQueueSubmit2(queue, submitCount, pSubmits, fence, record_obj);
RecordQueueSubmit(queue, fence, record_obj);
}
void SyncValidator::PostCallRecordGetFenceStatus(VkDevice device, VkFence fence, const RecordObject &record_obj) {
StateTracker::PostCallRecordGetFenceStatus(device, fence, record_obj);
if (!enabled[sync_validation_queue_submit]) return;
if (record_obj.result == VK_SUCCESS) {
// fence is signalled, mark it as waited for
WaitForFence(fence);
}
}
void SyncValidator::PostCallRecordWaitForFences(VkDevice device, uint32_t fenceCount, const VkFence *pFences, VkBool32 waitAll,
uint64_t timeout, const RecordObject &record_obj) {
StateTracker::PostCallRecordWaitForFences(device, fenceCount, pFences, waitAll, timeout, record_obj);
if (!enabled[sync_validation_queue_submit]) return;
if ((record_obj.result == VK_SUCCESS) && ((VK_TRUE == waitAll) || (1 == fenceCount))) {
// We can only know the pFences have signal if we waited for all of them, or there was only one of them
for (uint32_t i = 0; i < fenceCount; i++) {
WaitForFence(pFences[i]);
}
}
}
void SyncValidator::PostCallRecordGetSwapchainImagesKHR(VkDevice device, VkSwapchainKHR swapchain, uint32_t *pSwapchainImageCount,
VkImage *pSwapchainImages, const RecordObject &record_obj) {
StateTracker::PostCallRecordGetSwapchainImagesKHR(device, swapchain, pSwapchainImageCount, pSwapchainImages, record_obj);
if ((record_obj.result != VK_SUCCESS) && (record_obj.result != VK_INCOMPLETE)) return;
auto swapchain_state = Get<SWAPCHAIN_NODE>(swapchain);
if (pSwapchainImages) {
for (uint32_t i = 0; i < *pSwapchainImageCount; ++i) {
SWAPCHAIN_IMAGE &swapchain_image = swapchain_state->images[i];
if (swapchain_image.image_state) {
auto *sync_image = static_cast<ImageState *>(swapchain_image.image_state);
assert(sync_image->IsTiled()); // This is the assumption from the spec, and the implementation relies on it
sync_image->SetOpaqueBaseAddress(*this);
}
}
}
}
AttachmentViewGen::AttachmentViewGen(const syncval_state::ImageViewState *image_view, const VkOffset3D &offset,
const VkExtent3D &extent)
: view_(image_view), view_mask_(image_view->normalized_subresource_range.aspectMask), gen_store_() {
gen_store_[Gen::kViewSubresource].emplace(image_view->GetFullViewImageRangeGen());
gen_store_[Gen::kRenderArea].emplace(image_view->MakeImageRangeGen(offset, extent));
const auto depth = view_mask_ & VK_IMAGE_ASPECT_DEPTH_BIT;
if (depth && (depth != view_mask_)) {
gen_store_[Gen::kDepthOnlyRenderArea].emplace(image_view->MakeImageRangeGen(offset, extent, depth));
}
const auto stencil = view_mask_ & VK_IMAGE_ASPECT_STENCIL_BIT;
if (stencil && (stencil != view_mask_)) {
gen_store_[Gen::kStencilOnlyRenderArea].emplace(image_view->MakeImageRangeGen(offset, extent, stencil));
}
}
const std::optional<ImageRangeGen> &AttachmentViewGen::GetRangeGen(AttachmentViewGen::Gen type) const {
static_assert(Gen::kGenSize == 4, "Function written with this assumption");
// If the view is a depth only view, then the depth only portion of the render area is simply the render area.
// If the view is a depth stencil view, then the depth only portion of the render area will be a subset,
// and thus needs the generator function that will produce the address ranges of that subset
const bool depth_only = (type == kDepthOnlyRenderArea) && (view_mask_ == VK_IMAGE_ASPECT_DEPTH_BIT);
const bool stencil_only = (type == kStencilOnlyRenderArea) && (view_mask_ == VK_IMAGE_ASPECT_STENCIL_BIT);
if (depth_only || stencil_only) {
type = Gen::kRenderArea;
}
return gen_store_[type];
}
AttachmentViewGen::Gen AttachmentViewGen::GetDepthStencilRenderAreaGenType(bool depth_op, bool stencil_op) const {
assert(IsValid());
assert(view_mask_ & (VK_IMAGE_ASPECT_DEPTH_BIT | VK_IMAGE_ASPECT_STENCIL_BIT));
if (depth_op) {
assert(view_mask_ & VK_IMAGE_ASPECT_DEPTH_BIT);
if (stencil_op) {
assert(view_mask_ & VK_IMAGE_ASPECT_STENCIL_BIT);
return kRenderArea;
}
return kDepthOnlyRenderArea;
}
if (stencil_op) {
assert(view_mask_ & VK_IMAGE_ASPECT_STENCIL_BIT);
return kStencilOnlyRenderArea;
}
assert(depth_op || stencil_op);
return kRenderArea;
}
void SyncEventsContext::ApplyBarrier(const SyncExecScope &src, const SyncExecScope &dst, ResourceUsageTag tag) {
const bool all_commands_bit = 0 != (src.mask_param & VK_PIPELINE_STAGE_ALL_COMMANDS_BIT);
for (auto &event_pair : map_) {
assert(event_pair.second); // Shouldn't be storing empty
auto &sync_event = *event_pair.second;
// Events don't happen at a stage, so we need to check and store the unexpanded ALL_COMMANDS if set for inter-event-calls
// But only if occuring before the tag
if (((sync_event.barriers & src.exec_scope) || all_commands_bit) && (sync_event.last_command_tag <= tag)) {
sync_event.barriers |= dst.exec_scope;
sync_event.barriers |= dst.mask_param & VK_PIPELINE_STAGE_ALL_COMMANDS_BIT;
}
}
}
void SyncEventsContext::ApplyTaggedWait(VkQueueFlags queue_flags, ResourceUsageTag tag) {
const SyncExecScope src_scope =
SyncExecScope::MakeSrc(queue_flags, VK_PIPELINE_STAGE_2_ALL_COMMANDS_BIT, VK_PIPELINE_STAGE_2_HOST_BIT);
const SyncExecScope dst_scope = SyncExecScope::MakeDst(queue_flags, VK_PIPELINE_STAGE_2_ALL_COMMANDS_BIT);
ApplyBarrier(src_scope, dst_scope, tag);
}
SyncEventsContext &SyncEventsContext::DeepCopy(const SyncEventsContext &from) {
// We need a deep copy of the const context to update during validation phase
for (const auto &event : from.map_) {
map_.emplace(event.first, std::make_shared<SyncEventState>(*event.second));
}
return *this;
}
void SyncEventsContext::AddReferencedTags(ResourceUsageTagSet &referenced) const {
for (const auto &event : map_) {
const std::shared_ptr<const SyncEventState> &event_state = event.second;
if (event_state) {
event_state->AddReferencedTags(referenced);
}
}
}
QueueBatchContext::QueueBatchContext(const SyncValidator &sync_state, const QueueSyncState &queue_state, uint64_t submit_index,
uint32_t batch_index)
: CommandExecutionContext(&sync_state),
queue_state_(&queue_state),
tag_range_(0, 0),
current_access_context_(&access_context_),
batch_log_(),
queue_sync_tag_(sync_state.GetQueueIdLimit(), ResourceUsageTag(0)),
batch_(queue_state, submit_index, batch_index) {}
QueueBatchContext::QueueBatchContext(const SyncValidator &sync_state)
: CommandExecutionContext(&sync_state),
queue_state_(),
tag_range_(0, 0),
current_access_context_(&access_context_),
batch_log_(),
queue_sync_tag_(sync_state.GetQueueIdLimit(), ResourceUsageTag(0)),
batch_() {}
void QueueBatchContext::Trim() {
// Clean up unneeded access context contents and log information
access_context_.Trim();
ResourceUsageTagSet used_tags;
access_context_.AddReferencedTags(used_tags);
// Note: AccessContexts in the SyncEventsState are trimmed when created.
events_context_.AddReferencedTags(used_tags);
// Only conserve AccessLog references that are referenced by used_tags
batch_log_.Trim(used_tags);
}
void QueueBatchContext::ResolveSubmittedCommandBuffer(const AccessContext &recorded_context, ResourceUsageTag offset) {
GetCurrentAccessContext()->ResolveFromContext(QueueTagOffsetBarrierAction(GetQueueId(), offset), recorded_context);
}
VulkanTypedHandle QueueBatchContext::Handle() const { return queue_state_->Handle(); }
template <typename Predicate>
void QueueBatchContext::ApplyPredicatedWait(Predicate &predicate) {
access_context_.EraseIf([&predicate](ResourceAccessRangeMap::value_type &access) {
// Apply..Wait returns true if the waited access is empty...
return access.second.ApplyPredicatedWait<Predicate>(predicate);
});
}
void QueueBatchContext::ApplyTaggedWait(QueueId queue_id, ResourceUsageTag tag) {
const bool any_queue = (queue_id == QueueSyncState::kQueueAny);
if (any_queue) {
// This isn't just avoid an unneeded test, but to allow *all* queues to to be waited in a single pass
// (and it does avoid doing the same test for every access, as well as avoiding the need for the predicate
// to grok Queue/Device/Wait differences.
ResourceAccessState::WaitTagPredicate predicate{tag};
ApplyPredicatedWait(predicate);
} else {
ResourceAccessState::WaitQueueTagPredicate predicate{queue_id, tag};
ApplyPredicatedWait(predicate);
}
// SwapChain acquire QBC's have no queue, but also, events are always empty.
if (queue_state_ && (queue_id == GetQueueId() || any_queue)) {
events_context_.ApplyTaggedWait(GetQueueFlags(), tag);
}
}
void QueueBatchContext::ApplyAcquireWait(const AcquiredImage &acquired) {
ResourceAccessState::WaitAcquirePredicate predicate{acquired.present_tag, acquired.acquire_tag};
ApplyPredicatedWait(predicate);
}
void QueueBatchContext::BeginRenderPassReplaySetup(ReplayState &replay, const SyncOpBeginRenderPass &begin_op) {
current_access_context_ = replay.ReplayStateRenderPassBegin(GetQueueFlags(), begin_op, access_context_);
}
void QueueBatchContext::NextSubpassReplaySetup(ReplayState &replay) {
current_access_context_ = replay.ReplayStateRenderPassNext();
}
void QueueBatchContext::EndRenderPassReplayCleanup(ReplayState &replay) {
replay.ReplayStateRenderPassEnd(access_context_);
current_access_context_ = &access_context_;
}
void QueueBatchContext::Cleanup() {
// Clear these after validation and import, not valid after.
batch_ = BatchAccessLog::BatchRecord();
command_buffers_.clear();
async_batches_.clear();
}
AccessContext *ReplayState::RenderPassReplayState::Begin(VkQueueFlags queue_flags, const SyncOpBeginRenderPass &begin_op_,
const AccessContext &external_context) {
Reset();
begin_op = &begin_op_;
subpass = 0;
const RenderPassAccessContext *rp_context = begin_op->GetRenderPassAccessContext();
assert(rp_context);
replay_context = &rp_context->GetContexts()[0];
InitSubpassContexts(queue_flags, *rp_context->GetRenderPassState(), &external_context, subpass_contexts);
// Replace the Async contexts with the the async context of the "external" context
// For replay we don't care about async subpasses, just async queue batches
for (auto &context : subpass_contexts) {
context.ClearAsyncContexts();
context.ImportAsyncContexts(external_context);
}
return &subpass_contexts[0];
}
AccessContext *ReplayState::RenderPassReplayState::Next() {
subpass++;
const RenderPassAccessContext *rp_context = begin_op->GetRenderPassAccessContext();
replay_context = &rp_context->GetContexts()[subpass];
return &subpass_contexts[subpass];
}
void ReplayState::RenderPassReplayState::End(AccessContext &external_context) {
external_context.ResolveChildContexts(subpass_contexts);
Reset();
}
class ApplySemaphoreBarrierAction {
public:
ApplySemaphoreBarrierAction(const SemaphoreScope &signal, const SemaphoreScope &wait) : signal_(signal), wait_(wait) {}
void operator()(ResourceAccessState *access) const { access->ApplySemaphore(signal_, wait_); }
private:
const SemaphoreScope &signal_;
const SemaphoreScope wait_;
};
class ApplyAcquireNextSemaphoreAction {
public:
ApplyAcquireNextSemaphoreAction(const SyncExecScope &wait_scope, ResourceUsageTag acquire_tag)
: barrier_(1, SyncBarrier(getPresentSrcScope(), getPresentValidAccesses(), wait_scope, SyncStageAccessFlags())),
acq_tag_(acquire_tag) {}
void operator()(ResourceAccessState *access) const {
// Note that the present operations may or may not be present, given that the fence wait may have cleared them out.
// Also, if a subsequent present has happened, we *don't* want to protect that...
if (access->LastWriteTag() <= acq_tag_) {
access->ApplyBarriersImmediate(barrier_);
}
}
private:
// kPresentSrcScope/kPresentValidAccesses cannot be regular global variables, because they use global
// variables from another compilation unit (through syncStageAccessMaskByStageBit() call) for initialization,
// and initialization of globals between compilation units is undefined. Instead they get initialized
// on the first use (it's important to ensure this first use is also not initialization of some global!).
const SyncExecScope &getPresentSrcScope() const {
static const SyncExecScope kPresentSrcScope =
SyncExecScope(VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL, // mask_param (unused)
VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL, // expanded_mask
VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL, // exec_scope
getPresentValidAccesses()); // valid_accesses
return kPresentSrcScope;
}
const SyncStageAccessFlags &getPresentValidAccesses() const {
static const SyncStageAccessFlags kPresentValidAccesses =
SyncStageAccessFlags(SyncStageAccess::AccessScopeByStage(VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL));
return kPresentValidAccesses;
}
private:
std::vector<SyncBarrier> barrier_;
ResourceUsageTag acq_tag_;
};
// Overload for QueuePresent semaphore waiting. Not applicable to QueueSubmit semaphores
std::shared_ptr<QueueBatchContext> QueueBatchContext::ResolveOneWaitSemaphore(VkSemaphore sem,
const PresentedImages &presented_images,
SignaledSemaphores &signaled) {
auto sem_state = sync_state_->Get<SEMAPHORE_STATE>(sem);
if (!sem_state) return nullptr; // Semaphore validity is handled by CoreChecks
// When signal_state goes out of scope, the signal information will be dropped, as Unsignal has released ownership.
auto signal_state = signaled.Unsignal(sem);
if (!signal_state) return nullptr; // Invalid signal, skip it.
assert(signal_state->batch);
const AccessContext &from_context = signal_state->batch->access_context_;
const SemaphoreScope &signal_scope = signal_state->first_scope;
const QueueId queue_id = GetQueueId();
const auto queue_flags = queue_state_->GetQueueFlags();
SemaphoreScope wait_scope{queue_id, SyncExecScope::MakeDst(queue_flags, VK_PIPELINE_STAGE_2_PRESENT_ENGINE_BIT_SYNCVAL)};
// If signal queue == wait queue, signal is treated as a memory barrier with an access scope equal to the present accesses
SyncBarrier sem_barrier(signal_scope, wait_scope, SyncBarrier::AllAccess());
const BatchBarrierOp sem_same_queue_op(wait_scope.queue, sem_barrier);
// Need to import the rest of the same queue contents without modification
SyncBarrier noop_barrier;
const BatchBarrierOp noop_barrier_op(wait_scope.queue, noop_barrier);
// Otherwise apply semaphore rules apply
const ApplySemaphoreBarrierAction sem_not_same_queue_op(signal_scope, wait_scope);
const SemaphoreScope noop_semaphore_scope(queue_id, noop_barrier.dst_exec_scope);
const ApplySemaphoreBarrierAction noop_sem_op(signal_scope, noop_semaphore_scope);
// For each presented image
for (const auto &presented : presented_images) {
// Need a copy that can be used as the pseudo-iterator...
subresource_adapter::ImageRangeGenerator range_gen(presented.range_gen);
if (signal_scope.queue == wait_scope.queue) {
// If signal queue == wait queue, signal is treated as a memory barrier with an access scope equal to the
// valid accesses for the sync scope.
access_context_.ResolveFromContext(sem_same_queue_op, from_context, range_gen);
access_context_.ResolveFromContext(noop_barrier_op, from_context);
} else {
access_context_.ResolveFromContext(sem_not_same_queue_op, from_context, range_gen);
access_context_.ResolveFromContext(noop_sem_op, from_context);
}
}
return signal_state->batch;
}
std::shared_ptr<QueueBatchContext> QueueBatchContext::ResolveOneWaitSemaphore(VkSemaphore sem, VkPipelineStageFlags2 wait_mask,
SignaledSemaphores &signaled) {
auto sem_state = sync_state_->Get<SEMAPHORE_STATE>(sem);
if (!sem_state) return nullptr; // Semaphore validity is handled by CoreChecks
// When signal state goes out of scope, the signal information will be dropped, as Unsignal has released ownership.
auto signal_state = signaled.Unsignal(sem);
if (!signal_state) return nullptr; // Invalid signal, skip it.
assert(signal_state->batch);
const SemaphoreScope &signal_scope = signal_state->first_scope;
const auto queue_flags = queue_state_->GetQueueFlags();
SemaphoreScope wait_scope{GetQueueId(), SyncExecScope::MakeDst(queue_flags, wait_mask)};
const AccessContext &from_context = signal_state->batch->access_context_;
if (signal_state->acquired.image) {
// Import the *presenting* batch, but replacing presenting with acquired.
ApplyAcquireNextSemaphoreAction apply_acq(wait_scope, signal_state->acquired.acquire_tag);
access_context_.ResolveFromContext(apply_acq, from_context, signal_state->acquired.generator);
// Grab the reset of the presenting QBC, with no effective barrier, won't overwrite the acquire, as the tag is newer
SyncBarrier noop_barrier;
const BatchBarrierOp noop_barrier_op(wait_scope.queue, noop_barrier);
access_context_.ResolveFromContext(noop_barrier_op, from_context);
} else {
if (signal_scope.queue == wait_scope.queue) {
// If signal queue == wait queue, signal is treated as a memory barrier with an access scope equal to the
// valid accesses for the sync scope.
SyncBarrier sem_barrier(signal_scope, wait_scope, SyncBarrier::AllAccess());
const BatchBarrierOp sem_barrier_op(wait_scope.queue, sem_barrier);
access_context_.ResolveFromContext(sem_barrier_op, from_context);
events_context_.ApplyBarrier(sem_barrier.src_exec_scope, sem_barrier.dst_exec_scope, ResourceUsageRecord::kMaxIndex);
} else {
ApplySemaphoreBarrierAction sem_op(signal_scope, wait_scope);
access_context_.ResolveFromContext(sem_op, signal_state->batch->access_context_);
}
}
// Cannot move from the signal state because it could be from the const global state, and C++ doesn't
// enforce deep constness.
return signal_state->batch;
}
void QueueBatchContext::ImportSyncTags(const QueueBatchContext &from) {
// NOTE: Assumes that from has set it's tag limit in it's own queue_id slot.
size_t q_limit = queue_sync_tag_.size();
assert(q_limit == from.queue_sync_tag_.size());
for (size_t q = 0; q < q_limit; q++) {
queue_sync_tag_[q] = std::max(queue_sync_tag_[q], from.queue_sync_tag_[q]);
}
}
void QueueBatchContext::SetupAccessContext(const std::shared_ptr<const QueueBatchContext> &prev,
const VkPresentInfoKHR &present_info, const PresentedImages &presented_images,
SignaledSemaphores &signaled) {
ConstBatchSet batches_resolved;
for (VkSemaphore sem : vvl::make_span(present_info.pWaitSemaphores, present_info.waitSemaphoreCount)) {
std::shared_ptr<QueueBatchContext> resolved = ResolveOneWaitSemaphore(sem, presented_images, signaled);
if (resolved) {
batches_resolved.emplace(std::move(resolved));
}
}
CommonSetupAccessContext(prev, batches_resolved);
}
bool QueueBatchContext::DoQueuePresentValidate(const Location &loc, const PresentedImages &presented_images) {
bool skip = false;
HazardDetector detector(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL);
// Tag the presented images so record doesn't have to know the tagging scheme
for (size_t index = 0; index < presented_images.size(); ++index) {
const PresentedImage &presented = presented_images[index];
// Need a copy that can be used as the pseudo-iterator...
HazardResult hazard = access_context_.DetectHazard(detector, presented.range_gen, AccessContext::DetectOptions::kDetectAll);
if (hazard.IsHazard()) {
const auto queue_handle = queue_state_->Handle();
const auto swap_handle = BASE_NODE::Handle(presented.swapchain_state.lock());
const auto image_handle = BASE_NODE::Handle(presented.image);
skip = sync_state_->LogError(
string_SyncHazardVUID(hazard.Hazard()), queue_handle, loc,
"Hazard %s for present pSwapchains[%" PRIu32 "] , swapchain %s, image index %" PRIu32 " %s, Access info %s.",
string_SyncHazard(hazard.Hazard()), presented.present_index, sync_state_->FormatHandle(swap_handle).c_str(),
presented.image_index, sync_state_->FormatHandle(image_handle).c_str(), FormatHazard(hazard).c_str());
if (skip) break;
}
}
return skip;
}
void QueueBatchContext::DoPresentOperations(const PresentedImages &presented_images) {
// For present, tagging is internal to the presented image record.
for (const auto &presented : presented_images) {
// Update memory state
presented.UpdateMemoryAccess(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL, presented.tag, access_context_);
}
}
void QueueBatchContext::LogPresentOperations(const PresentedImages &presented_images) {
if (tag_range_.size()) {
auto access_log = std::make_shared<AccessLog>();
batch_log_.Insert(batch_, tag_range_, access_log);
access_log->reserve(tag_range_.size());
assert(tag_range_.size() == presented_images.size());
for (const auto &presented : presented_images) {
access_log->emplace_back(PresentResourceRecord(static_cast<const PresentedImageRecord>(presented)));
}
}
}
void QueueBatchContext::DoAcquireOperation(const PresentedImage &presented) {
// Only one tag for acquire. The tag in presented is the present tag
presented.UpdateMemoryAccess(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_ACQUIRE_READ_SYNCVAL, tag_range_.begin, access_context_);
}
void QueueBatchContext::LogAcquireOperation(const PresentedImage &presented, vvl::Func command) {
auto access_log = std::make_shared<AccessLog>();
batch_log_.Insert(batch_, tag_range_, access_log);
access_log->emplace_back(AcquireResourceRecord(presented, tag_range_.begin, command));
}
void QueueBatchContext::SetupAccessContext(const std::shared_ptr<const QueueBatchContext> &prev, const VkSubmitInfo2 &submit_info,
SignaledSemaphores &signaled) {
// Import (resolve) the batches that are waited on, with the semaphore's effective barriers applied
ConstBatchSet batches_resolved;
const uint32_t wait_count = submit_info.waitSemaphoreInfoCount;
const VkSemaphoreSubmitInfo *wait_infos = submit_info.pWaitSemaphoreInfos;
for (const auto &wait_info : vvl::make_span(wait_infos, wait_count)) {
std::shared_ptr<QueueBatchContext> resolved = ResolveOneWaitSemaphore(wait_info.semaphore, wait_info.stageMask, signaled);
if (resolved) {
batches_resolved.emplace(std::move(resolved));
}
}
CommonSetupAccessContext(prev, batches_resolved);
}
void QueueBatchContext::SetupAccessContext(const PresentedImage &presented) {
if (presented.batch) {
access_context_.ResolveFromContext(NoopBarrierAction(), presented.batch->access_context_);
batch_log_.Import(presented.batch->batch_log_);
ImportSyncTags(*presented.batch);
}
}
void QueueBatchContext::CommonSetupAccessContext(const std::shared_ptr<const QueueBatchContext> &prev,
QueueBatchContext::ConstBatchSet &batches_resolved) {
// Import the previous batch information
if (prev) {
// Copy in the event state from the previous batch (on this queue)
events_context_.DeepCopy(prev->events_context_);
if (!vvl::Contains(batches_resolved, prev)) {
// If there are no semaphores to the previous batch, make sure a "submit order" non-barriered import is done
access_context_.ResolveFromContext(NoopBarrierAction(), prev->access_context_);
batches_resolved.emplace(prev);
}
}
// Get all the log and tag sync information for the resolved contexts
for (const auto &batch : batches_resolved) {
batch_log_.Import(batch->batch_log_);
ImportSyncTags(*batch);
}
// Gather async context information for hazard checks and conserve the QBC's for the async batches
async_batches_ =
sync_state_->GetQueueLastBatchSnapshot([&batches_resolved](const std::shared_ptr<const QueueBatchContext> &batch) {
return !vvl::Contains(batches_resolved, batch);
});
for (const auto &async_batch : async_batches_) {
const QueueId async_queue = async_batch->GetQueueId();
ResourceUsageTag sync_tag;
if (async_queue < queue_sync_tag_.size()) {
sync_tag = queue_sync_tag_[async_queue];
} else {
// If this isn't from a tracked queue, just check the batch itself
sync_tag = async_batch->GetTagRange().begin;
}
// The start of the asynchronous access range for a given queue is one more than the highest tagged reference
access_context_.AddAsyncContext(async_batch->GetCurrentAccessContext(), sync_tag);
// We need to snapshot the async log information for async hazard reporting
batch_log_.Import(async_batch->batch_log_);
}
}
void QueueBatchContext::SetupCommandBufferInfo(const VkSubmitInfo2 &submit_info) {
// Create the list of command buffers to submit
const uint32_t cb_count = submit_info.commandBufferInfoCount;
const VkCommandBufferSubmitInfo *const cb_infos = submit_info.pCommandBufferInfos;
command_buffers_.reserve(cb_count);
for (const auto &cb_info : vvl::make_span(cb_infos, cb_count)) {
auto cb_state = sync_state_->Get<syncval_state::CommandBuffer>(cb_info.commandBuffer);
if (cb_state) {
tag_range_.end += cb_state->access_context.GetTagLimit();
command_buffers_.emplace_back(static_cast<uint32_t>(&cb_info - cb_infos), std::move(cb_state));
}
}
}
// Look up the usage informaiton from the local or global logger
std::string QueueBatchContext::FormatUsage(ResourceUsageTag tag) const {
std::stringstream out;
BatchAccessLog::AccessRecord access = batch_log_[tag];
if (access.IsValid()) {
const BatchAccessLog::BatchRecord &batch = *access.batch;
const ResourceUsageRecord &record = *access.record;
if (batch.queue) {
// Queue and Batch information (for enqueued operations)
out << SyncNodeFormatter(*sync_state_, batch.queue->GetQueueState());
out << ", submit: " << batch.submit_index << ", batch: " << batch.batch_index;
}
out << ", batch_tag: " << batch.bias;
// Commandbuffer Usages Information
out << ", " << record.Formatter(*sync_state_, nullptr);
}
return out.str();
}
VkQueueFlags QueueBatchContext::GetQueueFlags() const { return queue_state_->GetQueueFlags(); }
QueueId QueueBatchContext::GetQueueId() const {
QueueId id = queue_state_ ? queue_state_->GetQueueId() : QueueSyncState::kQueueIdInvalid;
return id;
}
// For QueuePresent, the tag range is defined externally and must be passed in
void QueueBatchContext::SetupBatchTags(const ResourceUsageRange &tag_range) {
tag_range_ = tag_range;
SetupBatchTags();
}
// For QueueSubmit, the tag range is defined by the CommandBuffer setup.
// For QueuePresent, this is called when the tag_range is specified
void QueueBatchContext::SetupBatchTags() {
// Need new global tags for all accesses... the Reserve updates a mutable atomic
ResourceUsageRange global_tags = sync_state_->ReserveGlobalTagRange(GetTagRange().size());
SetTagBias(global_tags.begin);
}
void QueueBatchContext::InsertRecordedAccessLogEntries(const CommandBufferAccessContext &submitted_cb) {
const ResourceUsageTag end_tag = batch_log_.Import(batch_, submitted_cb);
batch_.bias = end_tag;
batch_.cb_index++;
}
void QueueBatchContext::SetTagBias(ResourceUsageTag bias) {
const auto size = tag_range_.size();
tag_range_.begin = bias;
tag_range_.end = bias + size;
access_context_.SetStartTag(bias);
batch_.bias = bias;
// Needed for ImportSyncTags to pick up the "from" own sync tag.
const QueueId this_q = GetQueueId();
if (this_q < queue_sync_tag_.size()) {
// If this is a non-queued operation we'll get a "special" value like invalid
queue_sync_tag_[this_q] = tag_range_.end;
}
}
// Since we're updating the QueueSync state, this is Record phase and the access log needs to point to the global one
// Batch Contexts saved during signalling have their AccessLog reset when the pending signals are signalled.
// NOTE: By design, QueueBatchContexts that are neither last, nor referenced by a signal are abandoned as unowned, since
// the contexts Resolve all history from previous all contexts when created
void QueueSyncState::UpdateLastBatch(std::shared_ptr<QueueBatchContext> &&new_last) {
// Update the queue to point to the last batch from the submit
if (new_last) {
// Clean up the events data in the previous last batch on queue, as only the subsequent batches have valid use for them
// and the QueueBatchContext::Setup calls have be copying them along from batch to batch during submit.
if (last_batch_) {
last_batch_->ResetEventsContext();
}
new_last->Trim();
last_batch_ = std::move(new_last);
}
}
// Note that function is const, but updates mutable submit_index to allow Validate to create correct tagging for command invocation
// scope state.
// Given that queue submits are supposed to be externally synchronized for the same queue, this should safe without being
// atomic... but as the ops are per submit, the performance cost is negible for the peace of mind.
uint64_t QueueSyncState::ReserveSubmitId() const { return submit_index_.fetch_add(1); }
// This is a const method, force the returned value to be const
std::shared_ptr<const SignaledSemaphores::Signal> SignaledSemaphores::GetPrev(VkSemaphore sem) const {
std::shared_ptr<Signal> prev_state;
if (prev_) {
prev_state = GetMapped(prev_->signaled_, sem, [&prev_state]() { return prev_state; });
}
return prev_state;
}
SignaledSemaphores::Signal::Signal(const std::shared_ptr<const SEMAPHORE_STATE> &sem_state_,
const std::shared_ptr<QueueBatchContext> &batch_, const SyncExecScope &exec_scope_)
: sem_state(sem_state_), batch(batch_), first_scope({batch->GetQueueId(), exec_scope_}) {
// Illegal to create a signal from no batch or an invalid semaphore... caller must assure validity
assert(batch);
assert(sem_state);
}
SignaledSemaphores::Signal::Signal(const std::shared_ptr<const SEMAPHORE_STATE> &sem_state_, const PresentedImage &presented,
ResourceUsageTag acq_tag)
: sem_state(sem_state_), batch(presented.batch), first_scope(), acquired(presented, acq_tag) {
// Illegal to create a signal from no batch or an invalid semaphore... caller must assure validity
assert(batch);
assert(sem_state);
}
FenceSyncState::FenceSyncState() : fence(), tag(kInvalidTag), queue_id(QueueSyncState::kQueueIdInvalid) {}
VkSemaphoreSubmitInfo SubmitInfoConverter::BatchStore::WaitSemaphore(const VkSubmitInfo &info, uint32_t index) {
VkSemaphoreSubmitInfo semaphore_info = vku::InitStructHelper();
semaphore_info.semaphore = info.pWaitSemaphores[index];
semaphore_info.stageMask = info.pWaitDstStageMask[index];
return semaphore_info;
}
VkCommandBufferSubmitInfo SubmitInfoConverter::BatchStore::CommandBuffer(const VkSubmitInfo &info, uint32_t index) {
VkCommandBufferSubmitInfo cb_info = vku::InitStructHelper();
cb_info.commandBuffer = info.pCommandBuffers[index];
return cb_info;
}
VkSemaphoreSubmitInfo SubmitInfoConverter::BatchStore::SignalSemaphore(const VkSubmitInfo &info, uint32_t index,
VkQueueFlags queue_flags) {
VkSemaphoreSubmitInfo semaphore_info = vku::InitStructHelper();
semaphore_info.semaphore = info.pSignalSemaphores[index];
// Can't just use BOTTOM, because of how access expansion is done
semaphore_info.stageMask =
sync_utils::ExpandPipelineStages(VK_PIPELINE_STAGE_ALL_COMMANDS_BIT, queue_flags, VK_PIPELINE_STAGE_2_HOST_BIT);
return semaphore_info;
}
SubmitInfoConverter::BatchStore::BatchStore(const VkSubmitInfo &info, VkQueueFlags queue_flags) {
info2 = vku::InitStructHelper();
info2.waitSemaphoreInfoCount = info.waitSemaphoreCount;
waits.reserve(info2.waitSemaphoreInfoCount);
for (uint32_t i = 0; i < info2.waitSemaphoreInfoCount; ++i) {
waits.emplace_back(WaitSemaphore(info, i));
}
info2.pWaitSemaphoreInfos = waits.data();
info2.commandBufferInfoCount = info.commandBufferCount;
cbs.reserve(info2.commandBufferInfoCount);
for (uint32_t i = 0; i < info2.commandBufferInfoCount; ++i) {
cbs.emplace_back(CommandBuffer(info, i));
}
info2.pCommandBufferInfos = cbs.data();
info2.signalSemaphoreInfoCount = info.signalSemaphoreCount;
signals.reserve(info2.signalSemaphoreInfoCount);
for (uint32_t i = 0; i < info2.signalSemaphoreInfoCount; ++i) {
signals.emplace_back(SignalSemaphore(info, i, queue_flags));
}
info2.pSignalSemaphoreInfos = signals.data();
}
SubmitInfoConverter::SubmitInfoConverter(uint32_t count, const VkSubmitInfo *infos, VkQueueFlags queue_flags) {
info_store.reserve(count);
info2s.reserve(count);
for (uint32_t batch = 0; batch < count; ++batch) {
info_store.emplace_back(infos[batch], queue_flags);
info2s.emplace_back(info_store.back().info2);
}
}
ResourceUsageTag BatchAccessLog::Import(const BatchRecord &batch, const CommandBufferAccessContext &cb_access) {
ResourceUsageTag bias = batch.bias;
ResourceUsageTag tag_limit = bias + cb_access.GetTagLimit();
ResourceUsageRange import_range = {bias, tag_limit};
log_map_.insert(std::make_pair(import_range, CBSubmitLog(batch, cb_access)));
return tag_limit;
}
void BatchAccessLog::Import(const BatchAccessLog &other) {
for (const auto &entry : other.log_map_) {
log_map_.insert(entry);
}
}
void BatchAccessLog::Insert(const BatchRecord &batch, const ResourceUsageRange &range,
std::shared_ptr<const CommandExecutionContext::AccessLog> log) {
log_map_.insert(std::make_pair(range, CBSubmitLog(batch, nullptr, std::move(log))));
}
// Trim: Remove any unreferenced AccessLog ranges from a BatchAccessLog
//
// In order to contain memory growth in the AccessLog information regarding prior submitted command buffers,
// the Trim call removes any AccessLog references that do not correspond to any tags in use. The set of referenced tag, used_tags,
// is generated by scanning the AccessContext and EventContext of the containing QueueBatchContext.
//
// Upon return the BatchAccessLog should only contain references to the AccessLog information needed by the
// containing parent QueueBatchContext.
//
// The algorithm used is another example of the "parallel iteration" pattern common within SyncVal. In this case we are
// traversing the ordered range_map containing the AccessLog references and the ordered set of tags in use.
//
// To efficiently perform the parallel iteration, optimizations within this function include:
// * when ranges are detected that have no tags referenced, all ranges between the last tag and the current tag are erased
// * when used tags prior to the current range are found, all tags up to the current range are skipped
// * when a tag is found within the current range, that range is skipped (and thus kept in the map), and further used tags
// within the range are skipped.
//
// Note that for each subcase, any "next steps" logic is designed to be handled within the subsequent iteration -- meaning that
// each subcase simply handles the specifics of the current update/skip/erase action needed, and leaves the iterators in a sensible
// state for the top of loop... intentionally eliding special case handling.
void BatchAccessLog::Trim(const ResourceUsageTagSet &used_tags) {
auto current_tag = used_tags.cbegin();
const auto end_tag = used_tags.cend();
auto current_map_range = log_map_.begin();
const auto end_map = log_map_.end();
while (current_map_range != end_map) {
if (current_tag == end_tag) {
// We're out of tags, the rest of the map isn't referenced, so erase it
current_map_range = log_map_.erase(current_map_range, end_map);
} else {
auto &range = current_map_range->first;
const ResourceUsageTag tag = *current_tag;
if (tag < range.begin) {
// Skip to the next tag potentially in range
// if this is end_tag, we'll handle that next iteration
current_tag = used_tags.lower_bound(range.begin);
} else if (tag >= range.end) {
// This tag is beyond the current range, delete all ranges between current_map_range,
// and the next that includes the tag. Next is not erased.
auto next_used = log_map_.lower_bound(ResourceUsageRange(tag, tag + 1));
current_map_range = log_map_.erase(current_map_range, next_used);
} else {
// Skip the rest of the tags in this range
// If this is end, the next iteration will handle
current_tag = used_tags.lower_bound(range.end);
// This is a range we will keep, advance to the next. Next iteration handles end condition
++current_map_range;
}
}
}
}
BatchAccessLog::AccessRecord BatchAccessLog::operator[](ResourceUsageTag tag) const {
auto found_log = log_map_.find(tag);
if (found_log != log_map_.cend()) {
return found_log->second[tag];
}
// tag not found
assert(false);
return AccessRecord();
}
BatchAccessLog::AccessRecord BatchAccessLog::CBSubmitLog::operator[](ResourceUsageTag tag) const {
assert(tag >= batch_.bias);
const size_t index = tag - batch_.bias;
assert(log_);
assert(index < log_->size());
return AccessRecord{&batch_, &(*log_)[index]};
}
PresentedImage::PresentedImage(const SyncValidator &sync_state, const std::shared_ptr<QueueBatchContext> batch_,
VkSwapchainKHR swapchain, uint32_t image_index_, uint32_t present_index_, ResourceUsageTag tag_)
: PresentedImageRecord{tag_, image_index_, present_index_, sync_state.Get<syncval_state::Swapchain>(swapchain), {}},
batch(std::move(batch_)) {
SetImage(image_index_);
}
PresentedImage::PresentedImage(std::shared_ptr<const syncval_state::Swapchain> swapchain, uint32_t at_index) : PresentedImage() {
swapchain_state = std::move(swapchain);
tag = kInvalidTag;
SetImage(at_index);
}
// Export uses move semantics...
void PresentedImage::ExportToSwapchain(SyncValidator &) { // Include this argument to prove the const cast is safe
// If the swapchain is dead just ignore the present
auto swap_lock = swapchain_state.lock();
if (BASE_NODE::Invalid(swap_lock)) return;
auto swap = std::const_pointer_cast<syncval_state::Swapchain>(swap_lock);
swap->RecordPresentedImage(std::move(*this));
}
void PresentedImage::SetImage(uint32_t at_index) {
image_index = at_index;
auto swap_lock = swapchain_state.lock();
if (BASE_NODE::Invalid(swap_lock)) return;
image = std::static_pointer_cast<const syncval_state::ImageState>(swap_lock->GetSwapChainImageShared(image_index));
if (Invalid()) {
range_gen = ImageRangeGen();
} else {
// For valid images create the type/range_gen to used to scope the semaphore operations
range_gen = image->MakeImageRangeGen(image->full_range, false);
}
}
void PresentedImage::UpdateMemoryAccess(SyncStageAccessIndex usage, ResourceUsageTag tag, AccessContext &access_context) const {
// Intentional copy. The range_gen argument is not copied by the Update... call below
subresource_adapter::ImageRangeGenerator generator = range_gen;
UpdateMemoryAccessStateFunctor action(access_context, usage, SyncOrdering::kNonAttachment, tag);
UpdateMemoryAccessState(access_context.GetAccessStateMap(), action, generator);
}
QueueBatchContext::PresentResourceRecord::Base_::Record QueueBatchContext::PresentResourceRecord::MakeRecord() const {
return std::make_unique<PresentResourceRecord>(presented_);
}
std::ostream &QueueBatchContext::PresentResourceRecord::Format(std::ostream &out, const SyncValidator &sync_state) const {
out << "vkQueuePresentKHR ";
out << "present_tag:" << presented_.tag;
out << ", pSwapchains[" << presented_.present_index << "]";
out << ": " << SyncNodeFormatter(sync_state, presented_.swapchain_state.lock().get());
out << ", image_index: " << presented_.image_index;
out << SyncNodeFormatter(sync_state, presented_.image.get());
return out;
}
QueueBatchContext::AcquireResourceRecord::Base_::Record QueueBatchContext::AcquireResourceRecord::MakeRecord() const {
return std::make_unique<AcquireResourceRecord>(presented_, acquire_tag_, command_);
}
std::ostream &QueueBatchContext::AcquireResourceRecord::Format(std::ostream &out, const SyncValidator &sync_state) const {
out << vvl::String(command_) << " ";
out << "aquire_tag:" << acquire_tag_;
out << ": " << SyncNodeFormatter(sync_state, presented_.swapchain_state.lock().get());
out << ", image_index: " << presented_.image_index;
out << SyncNodeFormatter(sync_state, presented_.image.get());
return out;
}
syncval_state::Swapchain::Swapchain(ValidationStateTracker *dev_data, const VkSwapchainCreateInfoKHR *pCreateInfo,
VkSwapchainKHR swapchain)
: SWAPCHAIN_NODE(dev_data, pCreateInfo, swapchain) {}
void syncval_state::Swapchain::RecordPresentedImage(PresentedImage &&presented_image) {
// All presented images are stored within the swapchain until the are reaquired.
const uint32_t image_index = presented_image.image_index;
if (image_index >= presented.size()) presented.resize(image_index + 1);
// Use move semantics to avoid atomic operations on the contained shared_ptrs
presented[image_index] = std::move(presented_image);
}
// We move from the presented images array 1) so we don't copy shared_ptr, and 2) to mark it acquired
PresentedImage syncval_state::Swapchain::MovePresentedImage(uint32_t image_index) {
if (presented.size() <= image_index) presented.resize(image_index + 1);
PresentedImage ret_val = std::move(presented[image_index]);
if (ret_val.Invalid()) {
// If this is the first time the image has been acquired, then it's valid to have no present record, so we create one
// Note: It's also possible this is an invalid acquire... but that's CoreChecks/Parameter validation's job to report
ret_val = PresentedImage(static_cast<const syncval_state::Swapchain *>(this)->shared_from_this(), image_index);
}
return ret_val;
}
AcquiredImage::AcquiredImage(const PresentedImage &presented, ResourceUsageTag acq_tag)
: image(presented.image),
generator(presented.range_gen),
present_tag(presented.tag),
acquire_tag(acq_tag) {}
FenceSyncState::FenceSyncState(const std::shared_ptr<const FENCE_STATE> &fence_, QueueId queue_id_, ResourceUsageTag tag_)
: fence(fence_), tag(tag_), queue_id(queue_id_) {}
FenceSyncState::FenceSyncState(const std::shared_ptr<const FENCE_STATE> &fence_, const PresentedImage &image, ResourceUsageTag tag_)
: fence(fence_), tag(tag_), queue_id(QueueSyncState::kQueueIdInvalid), acquired(image, tag) {}
// For RenderPass time validation this is "start tag", for QueueSubmit, this is the earliest
// unsynchronized tag for the Queue being tested against (max synchrononous + 1, perhaps)
ResourceUsageTag AccessContext::AsyncReference::StartTag() const { return (tag_ == kInvalidTag) ? context_->StartTag() : tag_; }
bool syncval_state::ImageState::IsSimplyBound() const {
bool simple = SimpleBinding(static_cast<const BINDABLE &>(*this)) || IsSwapchainImage() || bind_swapchain;
// If it's not simple we must have an encoder.
assert(!simple || fragment_encoder.get());
return simple;
}
void syncval_state::ImageState::SetOpaqueBaseAddress(ValidationStateTracker &dev_data) {
// This is safe to call if already called to simplify caller logic
// NOTE: Not asserting IsTiled, as there could in future be other reasons for opaque representations
if (opaque_base_address_) return;
VkDeviceSize opaque_base = 0U; // Fakespace Allocator starts > 0
auto get_opaque_base = [&opaque_base](const IMAGE_STATE &other) {
const ImageState &other_sync = static_cast<const ImageState &>(other);
opaque_base = other_sync.opaque_base_address_;
return true;
};
if (IsSwapchainImage()) {
AnyAliasBindingOf(bind_swapchain->ObjectBindings(), get_opaque_base);
} else {
AnyImageAliasOf(get_opaque_base);
}
if (!opaque_base) {
// The size of the opaque range is based on the SyncVal *internal* representation of the tiled resource, unrelated
// to the acutal size of the the resource in device memory. If differing representations become possible, the allocated
// size would need to be changed to those representation's size requirements.
opaque_base = dev_data.AllocFakeMemory(fragment_encoder->TotalSize());
}
opaque_base_address_ = opaque_base;
}
VkDeviceSize syncval_state::ImageState::GetResourceBaseAddress() const {
if (HasOpaqueMapping()) {
return GetOpaqueBaseAddress();
}
return GetFakeBaseAddress();
}
ImageRangeGen syncval_state::ImageState::MakeImageRangeGen(const VkImageSubresourceRange &subresource_range,
bool is_depth_sliced) const {
if (!fragment_encoder || !IsSimplyBound()) {
return ImageRangeGen(); // default range generators have an empty position (generator "end")
}
const auto base_address = GetResourceBaseAddress();
ImageRangeGen range_gen(*fragment_encoder.get(), subresource_range, base_address, is_depth_sliced);
return range_gen;
}
ImageRangeGen syncval_state::ImageState::MakeImageRangeGen(const VkImageSubresourceRange &subresource_range,
const VkOffset3D &offset, const VkExtent3D &extent,
bool is_depth_sliced) const {
if (!fragment_encoder || !IsSimplyBound()) {
return ImageRangeGen(); // default range generators have an empty position (generator "end")
}
const auto base_address = GetResourceBaseAddress();
subresource_adapter::ImageRangeGenerator range_gen(*fragment_encoder.get(), subresource_range, offset, extent, base_address,
is_depth_sliced);
return range_gen;
}
ReplayState::ReplayState(CommandExecutionContext &exec_context, const CommandBufferAccessContext &recorded_context,
const ErrorObject &error_obj, uint32_t index)
: exec_context_(exec_context),
recorded_context_(recorded_context),
error_obj_(error_obj),
index_(index),
base_tag_(exec_context.GetTagLimit()) {}
const AccessContext *ReplayState::GetRecordedAccessContext() const {
if (rp_replay_) {
return rp_replay_.replay_context;
}
return recorded_context_.GetCurrentAccessContext();
}
// intiially zero, but accumulating the dstStages of barriers if they chain.
ResourceAccessWriteState::ResourceAccessWriteState(const SyncStageAccessInfoType &usage_info, ResourceUsageTag tag)
: access_(&usage_info),
barriers_(),
tag_(tag),
queue_(QueueSyncState::kQueueIdInvalid),
dependency_chain_(VK_PIPELINE_STAGE_2_NONE_KHR),
pending_layout_ordering_(),
pending_dep_chain_(VK_PIPELINE_STAGE_2_NONE),
pending_barriers_() {}
bool ResourceAccessWriteState::IsWriteHazard(const SyncStageAccessInfoType &usage_info) const {
return !barriers_[usage_info.stage_access_index];
}
bool ResourceAccessWriteState::IsOrdered(const OrderingBarrier &ordering, QueueId queue_id) const {
assert(access_);
return (queue_ == queue_id) && ordering.access_scope[access_->stage_access_index];
}
bool ResourceAccessWriteState::IsOrderedWriteHazard(VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
// Must be neither in the access scope, nor in the chained access scope
return !WriteInScope(src_access_scope) && !WriteInChainedScope(src_exec_scope, src_access_scope);
}
bool ResourceAccessWriteState::IsWriteBarrierHazard(QueueId queue_id, VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
// Special rules for sequential ILT's
if (IsIndex(SYNC_IMAGE_LAYOUT_TRANSITION)) {
if (queue_id == queue_) {
// In queue, they are implicitly ordered
return false;
} else {
// In dep chain means that the ILT is *available*
return !WriteInChain(src_exec_scope);
}
}
// Otherwise treat as an ordinary write hazard check with ordering rules.
return IsOrderedWriteHazard(src_exec_scope, src_access_scope);
}
void ResourceAccessWriteState::Set(const SyncStageAccessInfoType &usage_info, ResourceUsageTag tag) {
access_ = &usage_info;
barriers_.reset();
dependency_chain_ = VK_PIPELINE_STAGE_2_NONE;
tag_ = tag;
queue_ = QueueSyncState::kQueueIdInvalid;
}
void ResourceAccessWriteState::MergeBarriers(const ResourceAccessWriteState &other) {
barriers_ |= other.barriers_;
dependency_chain_ |= other.dependency_chain_;
pending_barriers_ |= other.pending_barriers_;
pending_dep_chain_ |= other.pending_dep_chain_;
pending_layout_ordering_ |= other.pending_layout_ordering_;
}
void ResourceAccessWriteState::UpdatePendingBarriers(const SyncBarrier &barrier) {
pending_barriers_ |= barrier.dst_access_scope;
pending_dep_chain_ |= barrier.dst_exec_scope.exec_scope;
}
void ResourceAccessWriteState::ApplyPendingBarriers() {
dependency_chain_ |= pending_dep_chain_;
barriers_ |= pending_barriers_;
}
void ResourceAccessWriteState::UpdatePendingLayoutOrdering(const SyncBarrier &barrier) {
pending_layout_ordering_ |= OrderingBarrier(barrier.src_exec_scope.exec_scope, barrier.src_access_scope);
}
void ResourceAccessWriteState::ClearPending() {
pending_dep_chain_ = VK_PIPELINE_STAGE_2_NONE;
pending_barriers_.reset();
pending_layout_ordering_ = OrderingBarrier();
}
void ResourceAccessWriteState::SetQueueId(QueueId id) {
if (queue_ == QueueSyncState::kQueueIdInvalid) {
queue_ = id;
}
}
bool ResourceAccessWriteState::WriteInChain(VkPipelineStageFlags2KHR src_exec_scope) const {
return 0 != (dependency_chain_ & src_exec_scope);
}
bool ResourceAccessWriteState::WriteInScope(const SyncStageAccessFlags &src_access_scope) const {
assert(access_);
return src_access_scope[access_->stage_access_index];
}
bool ResourceAccessWriteState::WriteBarrierInScope(const SyncStageAccessFlags &src_access_scope) const {
return (barriers_ & src_access_scope).any();
}
bool ResourceAccessWriteState::WriteInSourceScopeOrChain(VkPipelineStageFlags2KHR src_exec_scope,
SyncStageAccessFlags src_access_scope) const {
assert(access_);
return WriteInChain(src_exec_scope) || WriteInScope(src_access_scope);
}
bool ResourceAccessWriteState::WriteInQueueSourceScopeOrChain(QueueId queue, VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
assert(access_);
return WriteInChain(src_exec_scope) || ((queue == queue_) && WriteInScope(src_access_scope));
}
bool ResourceAccessWriteState::WriteInEventScope(VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope, QueueId scope_queue,
ResourceUsageTag scope_tag) const {
// The scope logic for events is, if we're asking, the resource usage was flagged as "in the first execution scope" at
// the time of the SetEvent, thus all we need check is whether the access is the same one (i.e. before the scope tag
// in order to know if it's in the excecution scope
assert(access_);
return (tag_ < scope_tag) && WriteInQueueSourceScopeOrChain(scope_queue, src_exec_scope, src_access_scope);
}
bool ResourceAccessWriteState::WriteInChainedScope(VkPipelineStageFlags2KHR src_exec_scope,
const SyncStageAccessFlags &src_access_scope) const {
assert(access_);
return WriteInChain(src_exec_scope) && WriteBarrierInScope(src_access_scope);
}
HazardResult::HazardState::HazardState(const ResourceAccessState *access_state_, const SyncStageAccessInfoType &usage_info_,
SyncHazard hazard_, const SyncStageAccessFlags &prior_, ResourceUsageTag tag_)
: access_state(std::make_unique<const ResourceAccessState>(*access_state_)),
recorded_access(),
usage_index(usage_info_.stage_access_index),
prior_access(prior_),
tag(tag_),
hazard(hazard_) {
// Touchup the hazard to reflect "present as release" semantics
// NOTE: For implementing QFO release/acquire semantics... touch up here as well
if (access_state->IsLastWriteOp(SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL)) {
if (hazard == SyncHazard::READ_AFTER_WRITE) {
hazard = SyncHazard::READ_AFTER_PRESENT;
} else if (hazard == SyncHazard::WRITE_AFTER_WRITE) {
hazard = SyncHazard::WRITE_AFTER_PRESENT;
}
} else if (usage_index == SYNC_PRESENT_ENGINE_SYNCVAL_PRESENT_PRESENTED_SYNCVAL) {
if (hazard == SyncHazard::WRITE_AFTER_READ) {
hazard = SyncHazard::PRESENT_AFTER_READ;
} else if (hazard == SyncHazard::WRITE_AFTER_WRITE) {
hazard = SyncHazard::PRESENT_AFTER_WRITE;
}
}
}
syncval_state::ImageViewState::ImageViewState(const std::shared_ptr<IMAGE_STATE> &image_state, VkImageView iv,
const VkImageViewCreateInfo *ci, VkFormatFeatureFlags2KHR ff,
const VkFilterCubicImageViewImageFormatPropertiesEXT &cubic_props)
: IMAGE_VIEW_STATE(image_state, iv, ci, ff, cubic_props), view_range_gen(MakeImageRangeGen()) {}
ImageRangeGen syncval_state::ImageViewState::MakeImageRangeGen() const {
return GetImageState()->MakeImageRangeGen(normalized_subresource_range, IsDepthSliced());
}
ImageRangeGen syncval_state::ImageViewState::MakeImageRangeGen(const VkOffset3D &offset, const VkExtent3D &extent,
const VkImageAspectFlags aspect_mask) const {
if (Invalid()) ImageRangeGen();
// Intentional copy
VkImageSubresourceRange subresource_range = normalized_subresource_range;
if (aspect_mask) {
subresource_range.aspectMask &= aspect_mask;
assert(subresource_range.aspectMask);
}
return GetImageState()->MakeImageRangeGen(subresource_range, offset, extent, IsDepthSliced());
}