blob: b85fe16f31f478a9166de8c07c1e8d7fcdc88f51 [file] [log] [blame]
/*
* Copyright © 2022 Imagination Technologies Ltd.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#ifndef PVR_FORMATS_H
#define PVR_FORMATS_H
#include <stdbool.h>
#include <stdint.h>
#include <vulkan/vulkan.h>
#include "util/format/u_formats.h"
#include "vk_format.h"
/* This is based on VkClearColorValue which is an array of RGBA, and on the
* output register usage for the biggest 32 bit 4 component formats which use up
* all 4 output registers.
* So this can be used for both unpacked RGBA value and to represent values
* packed according to the hardware (the accum format).
*/
#define PVR_CLEAR_COLOR_ARRAY_SIZE 4
#define PVR_TEX_FORMAT_COUNT (ROGUE_TEXSTATE_IMAGE_WORD0_TEXFORMAT_MAX_SIZE + 1)
enum pvr_pbe_accum_format {
PVR_PBE_ACCUM_FORMAT_INVALID = 0, /* Explicitly treat 0 as invalid. */
PVR_PBE_ACCUM_FORMAT_U8,
PVR_PBE_ACCUM_FORMAT_S8,
PVR_PBE_ACCUM_FORMAT_U16,
PVR_PBE_ACCUM_FORMAT_S16,
PVR_PBE_ACCUM_FORMAT_F16,
PVR_PBE_ACCUM_FORMAT_F32,
PVR_PBE_ACCUM_FORMAT_UINT8,
PVR_PBE_ACCUM_FORMAT_UINT16,
PVR_PBE_ACCUM_FORMAT_UINT32,
PVR_PBE_ACCUM_FORMAT_SINT8,
PVR_PBE_ACCUM_FORMAT_SINT16,
PVR_PBE_ACCUM_FORMAT_SINT32,
/* Formats with medp shader output precision. */
PVR_PBE_ACCUM_FORMAT_UINT32_MEDP,
PVR_PBE_ACCUM_FORMAT_SINT32_MEDP,
PVR_PBE_ACCUM_FORMAT_U1010102,
PVR_PBE_ACCUM_FORMAT_U24,
};
/**
* Pixel related shader selector. The logic selecting the shader has to take
* into account the pixel related properties (controlling the conversion path in
* the shader) and the geometry related properties (controlling the sample
* position calcs). These two can be orthogonal.
*
* integer format conversions, bit depth : 8, 16, 32 per ch formats : signed,
* unsigned. Strategy: convert everything to U32 or S32 then USC pack. PBE just
* pass through.
*
* fixed point format conversions, bit depth 565, 1555, 555 etc. Strategy:
* fcnorm to 4 F32, then USC pack to F16F16. PBE converts to destination
*
* float/fixed format conversions
* strategy: fcnorm, then pack to f16 _when_ destination is not f32.
* fmt | unorm | flt |
* 8 | x | |
* 16 | x | x |
* 32 | x | x |
*
*
* non-merge type DS blit table
* **********************************************
* * * S8 D16 D24S8 D32 D32S8 *
* **********************************************
* * S8 * cpy i i i i *
* * D16 * i cpy i - i *
* * D24S8 * swiz - cpy (1) - *
* * D32 * i - i cpy i *
* * D32S8 * (2) - - cpy cpy *
* **********************************************
*
* merge with stencil pick type DS blit table
* **********************************************
* * * S8 D16 D24S8 D32 D32S8 *
* **********************************************
* * S8 * i i (1) i (2) *
* * D16 * i i i i i *
* * D24S8 * i i (3) i (4) *
* * D32 * i i i i i *
* * D32S8 * i i (5) i (6) *
* **********************************************
*
* merge with depth pick type DS blit table
* **********************************************
* * * S8 D16 D24S8 D32 D32S8 *
* **********************************************
* * S8 * i i i i i *
* * D16 * - - - - - *
* * D24S8 * - - (s) - - *
* * D32 * - - (1) - (2) *
* * D32S8 * - - - - (s) *
* **********************************************
*
* D formats are unpacked into a single register according to their format
* S formats are unpacked into a single register in U8
* D24S8 is in a single 32 bit register (as the PBE can't read it from
* unpacked.)
*
* Swizzles are applied on the TPU not the PBE because of potential
* accumulation i.e. a non-iterated shader doesn't know if it writes the output
* buffer for PBE emit or a second pass blend.
*/
enum pvr_transfer_pbe_pixel_src {
PVR_TRANSFER_PBE_PIXEL_SRC_UU8888 = 0,
PVR_TRANSFER_PBE_PIXEL_SRC_US8888 = 1,
PVR_TRANSFER_PBE_PIXEL_SRC_UU16U16 = 2,
PVR_TRANSFER_PBE_PIXEL_SRC_US16S16 = 3,
PVR_TRANSFER_PBE_PIXEL_SRC_SU8888 = 4,
PVR_TRANSFER_PBE_PIXEL_SRC_SS8888 = 5,
PVR_TRANSFER_PBE_PIXEL_SRC_SU16U16 = 6,
PVR_TRANSFER_PBE_PIXEL_SRC_SS16S16 = 7,
PVR_TRANSFER_PBE_PIXEL_SRC_UU1010102 = 8,
PVR_TRANSFER_PBE_PIXEL_SRC_SU1010102 = 9,
PVR_TRANSFER_PBE_PIXEL_SRC_RBSWAP_UU1010102 = 10,
PVR_TRANSFER_PBE_PIXEL_SRC_RBSWAP_SU1010102 = 11,
PVR_TRANSFER_PBE_PIXEL_SRC_SU32U32 = 12,
PVR_TRANSFER_PBE_PIXEL_SRC_S4XU32 = 13,
PVR_TRANSFER_PBE_PIXEL_SRC_US32S32 = 14,
PVR_TRANSFER_PBE_PIXEL_SRC_U4XS32 = 15,
PVR_TRANSFER_PBE_PIXEL_SRC_F16F16 = 16,
PVR_TRANSFER_PBE_PIXEL_SRC_U16NORM = 17,
PVR_TRANSFER_PBE_PIXEL_SRC_S16NORM = 18,
PVR_TRANSFER_PBE_PIXEL_SRC_F32X4 = 19,
PVR_TRANSFER_PBE_PIXEL_SRC_F32X2 = 20,
PVR_TRANSFER_PBE_PIXEL_SRC_F32 = 21,
PVR_TRANSFER_PBE_PIXEL_SRC_RAW32 = 22,
PVR_TRANSFER_PBE_PIXEL_SRC_RAW64 = 23,
PVR_TRANSFER_PBE_PIXEL_SRC_RAW128 = 24,
/* f16 to U8 conversion in shader. */
PVR_TRANSFER_PBE_PIXEL_SRC_F16_U8 = 25,
PVR_TRANSFER_PBE_PIXEL_SRC_SWAP_LMSB = 26,
PVR_TRANSFER_PBE_PIXEL_SRC_MOV_BY45 = 27,
PVR_TRANSFER_PBE_PIXEL_SRC_D24S8 = 28,
PVR_TRANSFER_PBE_PIXEL_SRC_S8D24 = 29,
PVR_TRANSFER_PBE_PIXEL_SRC_D32S8 = 30,
/* D: D32_S8 */
PVR_TRANSFER_PBE_PIXEL_SRC_SMRG_S8_D32S8 = 31,
PVR_TRANSFER_PBE_PIXEL_SRC_SMRG_D24S8_D32S8 = 32,
PVR_TRANSFER_PBE_PIXEL_SRC_SMRG_D32S8_D32S8 = 33,
PVR_TRANSFER_PBE_PIXEL_SRC_DMRG_D32S8_D32S8 = 34,
/* D: D32 */
PVR_TRANSFER_PBE_PIXEL_SRC_CONV_D24_D32 = 35,
PVR_TRANSFER_PBE_PIXEL_SRC_CONV_D32U_D32F = 36,
/* D : D24_S8 */
PVR_TRANSFER_PBE_PIXEL_SRC_SMRG_S8_D24S8 = 37,
PVR_TRANSFER_PBE_PIXEL_SRC_SMRG_D24S8_D24S8 = 38,
PVR_TRANSFER_PBE_PIXEL_SRC_DMRG_D24S8_D24S8 = 39,
PVR_TRANSFER_PBE_PIXEL_SRC_CONV_D32_D24S8 = 40,
PVR_TRANSFER_PBE_PIXEL_SRC_DMRG_D32_D24S8 = 41,
PVR_TRANSFER_PBE_PIXEL_SRC_DMRG_D32U_D24S8 = 42,
/* ob0 holds Y and ob0 holds U or V. */
PVR_TRANSFER_PBE_PIXEL_SRC_YUV_PACKED = 43,
/* ob0 holds Y, ob1 holds U, ob2 holds V. */
PVR_TRANSFER_PBE_PIXEL_SRC_Y_U_V = 44,
PVR_TRANSFER_PBE_PIXEL_SRC_MASK16 = 45,
PVR_TRANSFER_PBE_PIXEL_SRC_MASK32 = 46,
PVR_TRANSFER_PBE_PIXEL_SRC_MASK48 = 47,
PVR_TRANSFER_PBE_PIXEL_SRC_MASK64 = 48,
PVR_TRANSFER_PBE_PIXEL_SRC_MASK96 = 49,
PVR_TRANSFER_PBE_PIXEL_SRC_MASK128 = 50,
PVR_TRANSFER_PBE_PIXEL_SRC_CONV_S8D24_D24S8 = 51,
/* ob0 holds Y and ob0 holds V or U. */
PVR_TRANSFER_PBE_PIXEL_SRC_YVU_PACKED = 52,
/* ob0 holds Y, ob1 holds UV interleaved. */
PVR_TRANSFER_PBE_PIXEL_SRC_Y_UV_INTERLEAVED = 53,
/* FIXME: This changes for other BVNC's which may change the hashing logic
* in pvr_hash_shader.
*/
PVR_TRANSFER_PBE_PIXEL_SRC_NUM = 54,
};
/* FIXME: Replace all instances of uint32_t with ROGUE_TEXSTATE_FORMAT or
* ROGUE_TEXSTATE_FORMAT_COMPRESSED after the pvr_common cleanup is complete.
*/
struct pvr_tex_format_description {
uint32_t tex_format;
enum pipe_format pipe_format_int;
enum pipe_format pipe_format_float;
};
struct pvr_tex_format_compressed_description {
uint32_t tex_format;
enum pipe_format pipe_format;
uint32_t tex_format_simple;
};
bool pvr_tex_format_is_supported(uint32_t tex_format);
const struct pvr_tex_format_description *
pvr_get_tex_format_description(uint32_t tex_format);
bool pvr_tex_format_compressed_is_supported(uint32_t tex_format);
const struct pvr_tex_format_compressed_description *
pvr_get_tex_format_compressed_description(uint32_t tex_format);
const uint8_t *pvr_get_format_swizzle(VkFormat vk_format);
uint32_t pvr_get_tex_format(VkFormat vk_format);
uint32_t pvr_get_tex_format_aspect(VkFormat vk_format,
VkImageAspectFlags aspect_mask);
uint32_t pvr_get_pbe_packmode(VkFormat vk_format);
uint32_t pvr_get_pbe_accum_format(VkFormat vk_format);
uint32_t pvr_get_pbe_accum_format_size_in_bytes(VkFormat vk_format);
bool pvr_format_is_pbe_downscalable(VkFormat vk_format);
void pvr_get_hw_clear_color(VkFormat vk_format,
VkClearColorValue value,
uint32_t packed_out[static const 4]);
uint32_t pvr_pbe_pixel_num_loads(enum pvr_transfer_pbe_pixel_src pbe_format);
static inline bool pvr_vk_format_has_32bit_component(VkFormat vk_format)
{
const struct util_format_description *desc =
vk_format_description(vk_format);
for (uint32_t i = 0; i < desc->nr_channels; i++) {
if (desc->channel[i].size == 32U)
return true;
}
return false;
}
static inline bool pvr_vk_format_is_fully_normalized(VkFormat vk_format)
{
const struct util_format_description *desc =
vk_format_description(vk_format);
for (uint32_t i = 0; i < desc->nr_channels; i++) {
if (!desc->channel[i].normalized)
return false;
}
return true;
}
static inline uint32_t
pvr_vk_format_get_common_color_channel_count(VkFormat src_format,
VkFormat dst_format)
{
const struct util_format_description *dst_desc =
vk_format_description(dst_format);
const struct util_format_description *src_desc =
vk_format_description(src_format);
uint32_t count = 0;
/* Check if destination format is alpha only and source format has alpha
* channel.
*/
if (util_format_is_alpha(vk_format_to_pipe_format(dst_format))) {
count = 1;
} else if (dst_desc->nr_channels <= src_desc->nr_channels) {
for (uint32_t i = 0; i < dst_desc->nr_channels; i++) {
enum pipe_swizzle swizzle = dst_desc->swizzle[i];
if (swizzle > PIPE_SWIZZLE_W)
continue;
for (uint32_t j = 0; j < src_desc->nr_channels; j++) {
if (src_desc->swizzle[j] == swizzle) {
count++;
break;
}
}
}
} else {
count = dst_desc->nr_channels;
}
return count;
}
#endif /* PVR_FORMATS_H */