src/gallium/drivers/panfrost/pan_mempool.c

/*
 * © Copyright 2018 Alyssa Rosenzweig
 * Copyright (C) 2019 Collabora, 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.
 *
 */

#include <unistd.h>
#include <sys/mman.h>

#include "pan_device.h"
#include "pan_mempool.h"

/* Knockoff u_upload_mgr. Uploads wherever we left off, allocating new entries
 * when needed.
 *
 * In "owned" mode, a single parent owns the entire pool, and the pool owns all
 * created BOs. All BOs are tracked and addable as
 * panfrost_pool_get_bo_handles. Freeing occurs at the level of an entire pool.
 * This is useful for streaming uploads, where the batch owns the pool.
 *
 * In "unowned" mode, the pool is freestanding. It does not track created BOs
 * or hold references. Instead, the consumer must manage the created BOs. This
 * is more flexible, enabling non-transient CSO state or shader code to be
 * packed with conservative lifetime handling.
 */

static struct panfrost_bo *
panfrost_pool_alloc_backing(struct panfrost_pool *pool, size_t bo_sz)
{
        /* We don't know what the BO will be used for, so let's flag it
         * RW and attach it to both the fragment and vertex/tiler jobs.
         * TODO: if we want fine grained BO assignment we should pass
         * flags to this function and keep the read/write,
         * fragment/vertex+tiler pools separate.
         */
        struct panfrost_bo *bo = panfrost_bo_create(pool->base.dev, bo_sz,
                        pool->base.create_flags, pool->base.label);

        if (pool->owned)
                util_dynarray_append(&pool->bos, struct panfrost_bo *, bo);
        else
                panfrost_bo_unreference(pool->transient_bo);

        pool->transient_bo = bo;
        pool->transient_offset = 0;

        return bo;
}

void
panfrost_pool_init(struct panfrost_pool *pool, void *memctx,
                   struct panfrost_device *dev,
                   unsigned create_flags, size_t slab_size, const char *label,
                   bool prealloc, bool owned)
{
        memset(pool, 0, sizeof(*pool));
        pan_pool_init(&pool->base, dev, create_flags, slab_size, label);
        pool->owned = owned;

        if (owned)
                util_dynarray_init(&pool->bos, memctx);

        if (prealloc)
                panfrost_pool_alloc_backing(pool, pool->base.slab_size);
}

void
panfrost_pool_cleanup(struct panfrost_pool *pool)
{
        if (!pool->owned) {
                panfrost_bo_unreference(pool->transient_bo);
                return;
        }

        util_dynarray_foreach(&pool->bos, struct panfrost_bo *, bo)
                panfrost_bo_unreference(*bo);

        util_dynarray_fini(&pool->bos);
}

void
panfrost_pool_get_bo_handles(struct panfrost_pool *pool, uint32_t *handles)
{
        assert(pool->owned && "pool does not track BOs in unowned mode");

        unsigned idx = 0;
        util_dynarray_foreach(&pool->bos, struct panfrost_bo *, bo) {
                assert((*bo)->gem_handle > 0);
                handles[idx++] = (*bo)->gem_handle;

               /* Update the BO access flags so that panfrost_bo_wait() knows
                * about all pending accesses.
                * We only keep the READ/WRITE info since this is all the BO
                * wait logic cares about.
                * We also preserve existing flags as this batch might not
                * be the first one to access the BO.
                */
                (*bo)->gpu_access |= PAN_BO_ACCESS_RW;
        }
}

#define PAN_GUARD_SIZE 4096

static struct panfrost_ptr
panfrost_pool_alloc_aligned(struct panfrost_pool *pool, size_t sz, unsigned alignment)
{
        assert(alignment == util_next_power_of_two(alignment));

        /* Find or create a suitable BO */
        struct panfrost_bo *bo = pool->transient_bo;
        unsigned offset = ALIGN_POT(pool->transient_offset, alignment);

#ifdef PAN_DBG_OVERFLOW
        if (unlikely(pool->base.dev->debug & PAN_DBG_OVERFLOW) &&
            !(pool->base.create_flags & PAN_BO_INVISIBLE)) {
                unsigned aligned = ALIGN_POT(sz, sysconf(_SC_PAGESIZE));
                unsigned bo_size = aligned + PAN_GUARD_SIZE;

                bo = panfrost_pool_alloc_backing(pool, bo_size);
                memset(bo->ptr.cpu, 0xbb, bo_size);

                /* Place the object as close as possible to the protected
                 * region at the end of the buffer while keeping alignment. */
                offset = ROUND_DOWN_TO(aligned - sz, alignment);

                if (mprotect(bo->ptr.cpu + aligned,
                             PAN_GUARD_SIZE, PROT_NONE) == -1)
                        perror("mprotect");

                pool->transient_bo = NULL;
        }
#endif

        /* If we don't fit, allocate a new backing */
        if (unlikely(bo == NULL || (offset + sz) >= pool->base.slab_size)) {
                bo = panfrost_pool_alloc_backing(pool,
                                ALIGN_POT(MAX2(pool->base.slab_size, sz), 4096));
                offset = 0;
        }

        pool->transient_offset = offset + sz;

        struct panfrost_ptr ret = {
                .cpu = bo->ptr.cpu + offset,
                .gpu = bo->ptr.gpu + offset,
        };

        return ret;
}
PAN_POOL_ALLOCATOR(struct panfrost_pool, panfrost_pool_alloc_aligned)