/************************************************************************** * * Copyright © 2007 Red Hat Inc. * Copyright © 2007-2021 Intel Corporation * Copyright 2006 Tungsten Graphics, Inc., Bismarck, ND., USA * All Rights Reserved. * * 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, sub license, 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 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 NON-INFRINGEMENT. IN NO EVENT SHALL * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS 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. * * The above copyright notice and this permission notice (including the * next paragraph) shall be included in all copies or substantial portions * of the Software. * * **************************************************************************/ /* * Authors: Thomas Hellström * Keith Whitwell * Eric Anholt * Dave Airlie */ #ifdef HAVE_CONFIG_H #include "config.h" #endif #include "xf86drm.h" #include "xf86atomic.h" #include #include #include #include #include #include #include #include #include #include #include #include "errno.h" #ifndef ETIME #define ETIME ETIMEDOUT #endif #include "libdrm_macros.h" #include "libdrm_lists.h" #include "mos_bufmgr.h" #include "mos_bufmgr_priv.h" #include "string.h" #include "i915_drm.h" #include "mos_vma.h" #ifdef HAVE_VALGRIND #include #include #define VG(x) x #else #define VG(x) #endif #define memclear(s) memset(&s, 0, sizeof(s)) #define MOS_DBG(...) do { \ if (bufmgr_gem->bufmgr.debug) \ fprintf(stderr, __VA_ARGS__); \ } while (0) #define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0])) #define MAX2(A, B) ((A) > (B) ? (A) : (B)) /** * upper_32_bits - return bits 32-63 of a number * @n: the number we're accessing * * A basic shift-right of a 64- or 32-bit quantity. Use this to suppress * the "right shift count >= width of type" warning when that quantity is * 32-bits. */ #define upper_32_bits(n) ((__u32)(((n) >> 16) >> 16)) /** * lower_32_bits - return bits 0-31 of a number * @n: the number we're accessing */ #define lower_32_bits(n) ((__u32)(n)) #define PCI_CHIP_I915_G 0x2582 #define PCI_CHIP_E7221_G 0x258A #define PCI_CHIP_I915_GM 0x2592 #define IS_915(devid) ((devid) == PCI_CHIP_I915_G || \ (devid) == PCI_CHIP_E7221_G || \ (devid) == PCI_CHIP_I915_GM) struct mos_gem_bo_bucket { drmMMListHead head; unsigned long size; }; struct mos_bufmgr_gem { struct mos_bufmgr bufmgr; atomic_t refcount; int fd; int max_relocs; pthread_mutex_t lock; struct drm_i915_gem_exec_object *exec_objects; struct drm_i915_gem_exec_object2 *exec2_objects; struct mos_linux_bo **exec_bos; int exec_size; int exec_count; /** Array of lists of cached gem objects of power-of-two sizes */ struct mos_gem_bo_bucket cache_bucket[14 * 4]; int num_buckets; time_t time; drmMMListHead managers; drmMMListHead named; uint64_t gtt_size; int available_fences; int pci_device; unsigned int has_bsd : 1; unsigned int has_blt : 1; unsigned int has_relaxed_fencing : 1; unsigned int has_llc : 1; unsigned int has_wait_timeout : 1; unsigned int bo_reuse : 1; unsigned int no_exec : 1; unsigned int has_vebox : 1; unsigned int has_ext_mmap : 1; bool fenced_relocs; struct { void *ptr; uint32_t handle; } userptr_active; // manage address for softpin buffer object mos_vma_heap vma_heap[MEMZONE_COUNT]; bool use_softpin; int device_type; } mos_bufmgr_gem; #define DRM_INTEL_RELOC_FENCE (1<<0) struct mos_reloc_target { struct mos_linux_bo *bo; int flags; }; struct mos_softpin_target { struct mos_linux_bo *bo; int flags; }; struct mos_bo_gem { struct mos_linux_bo bo; atomic_t refcount; uint32_t gem_handle; const char *name; /** * Kenel-assigned global name for this object * * List contains both flink named and prime fd'd objects */ unsigned int global_name; drmMMListHead name_list; /** * Index of the buffer within the validation list while preparing a * batchbuffer execution. */ int validate_index; /** * Current tiling mode */ uint32_t tiling_mode; uint32_t swizzle_mode; unsigned long stride; time_t free_time; /** Array passed to the DRM containing relocation information. */ struct drm_i915_gem_relocation_entry *relocs; /** * Array of info structs corresponding to relocs[i].target_handle etc */ struct mos_reloc_target *reloc_target_info; /** Number of entries in relocs */ int reloc_count; /** Array of BOs that are referenced by this buffer and will be softpinned */ struct mos_softpin_target *softpin_target; /** Number softpinned BOs that are referenced by this buffer */ int softpin_target_count; /** Maximum amount of softpinned BOs that are referenced by this buffer */ int softpin_target_size; /** Mapped address for the buffer, saved across map/unmap cycles */ void *mem_virtual; /** Uncached Mapped address for the buffer, saved across map/unmap cycles */ void *mem_wc_virtual; /** GTT virtual address for the buffer, saved across map/unmap cycles */ void *gtt_virtual; /** * Virtual address of the buffer allocated by user, used for userptr * objects only. */ void *user_virtual; int map_count; /** BO cache list */ drmMMListHead head; /** * Boolean of whether this BO and its children have been included in * the current drm_intel_bufmgr_check_aperture_space() total. */ bool included_in_check_aperture; /** * Boolean of whether this buffer has been used as a relocation * target and had its size accounted for, and thus can't have any * further relocations added to it. */ bool used_as_reloc_target; /** * Boolean of whether we have encountered an error whilst building the relocation tree. */ bool has_error; /** * Boolean of whether this buffer can be re-used */ bool reusable; /** * Boolean of whether the GPU is definitely not accessing the buffer. * * This is only valid when reusable, since non-reusable * buffers are those that have been shared wth other * processes, so we don't know their state. */ bool idle; /** * Boolean of whether this buffer was allocated with userptr */ bool is_userptr; /** * Boolean of whether this buffer can be placed in the full 48-bit * address range on gen8+. * * By default, buffers will be keep in a 32-bit range, unless this * flag is explicitly set. */ bool use_48b_address_range; /** * Whether this buffer is softpinned at offset specified by the user */ bool is_softpin; /* * Whether to remove the dependency of this bo in exebuf. */ bool exec_async; /** * Size in bytes of this buffer and its relocation descendents. * * Used to avoid costly tree walking in * drm_intel_bufmgr_check_aperture in the common case. */ int reloc_tree_size; /** * Number of potential fence registers required by this buffer and its * relocations. */ int reloc_tree_fences; /** Flags that we may need to do the SW_FINSIH ioctl on unmap. */ bool mapped_cpu_write; /** * Size to pad the object to. * */ uint64_t pad_to_size; }; static unsigned int mos_gem_estimate_batch_space(struct mos_linux_bo ** bo_array, int count); static unsigned int mos_gem_compute_batch_space(struct mos_linux_bo ** bo_array, int count); static int mos_gem_bo_get_tiling(struct mos_linux_bo *bo, uint32_t * tiling_mode, uint32_t * swizzle_mode); static int mos_gem_bo_set_tiling_internal(struct mos_linux_bo *bo, uint32_t tiling_mode, uint32_t stride); static void mos_gem_bo_unreference_locked_timed(struct mos_linux_bo *bo, time_t time); static void mos_gem_bo_unreference(struct mos_linux_bo *bo); static bool mos_gem_bo_is_softpin(struct mos_linux_bo *bo); static void mos_gem_bo_start_gtt_access(struct mos_linux_bo *bo, int write_enable); static void mos_gem_bo_free(struct mos_linux_bo *bo); static inline struct mos_bo_gem *to_bo_gem(struct mos_linux_bo *bo) { return (struct mos_bo_gem *)bo; } static int GetDrmMode() { return 1;//We always use SW Mode in libdrm mock. /* char *penvset_drm = getenv("DRMMode"); int DrmMode=0; if(penvset_drm != nullptr) { DrmMode= atoi(penvset_drm); //printf("DrmMode =%d\n", DrmMode); } return DrmMode; */ } static unsigned long mos_gem_bo_tile_size(struct mos_bufmgr_gem *bufmgr_gem, unsigned long size, uint32_t *tiling_mode) { unsigned long min_size, max_size; unsigned long i; if (*tiling_mode == I915_TILING_NONE) return size; /* 965+ just need multiples of page size for tiling */ return ROUND_UP_TO(size, 4096); } /* * Round a given pitch up to the minimum required for X tiling on a * given chip. We use 512 as the minimum to allow for a later tiling * change. */ static unsigned long mos_gem_bo_tile_pitch(struct mos_bufmgr_gem *bufmgr_gem, unsigned long pitch, uint32_t *tiling_mode) { unsigned long tile_width; unsigned long i; /* If untiled, then just align it so that we can do rendering * to it with the 3D engine. */ if (*tiling_mode == I915_TILING_NONE) return ALIGN(pitch, 64); if (*tiling_mode == I915_TILING_X || (IS_915(bufmgr_gem->pci_device) && *tiling_mode == I915_TILING_Y)) tile_width = 512; else tile_width = 128; /* 965 is flexible */ return ROUND_UP_TO(pitch, tile_width); } static struct mos_gem_bo_bucket * mos_gem_bo_bucket_for_size(struct mos_bufmgr_gem *bufmgr_gem, unsigned long size) { int i; for (i = 0; i < bufmgr_gem->num_buckets; i++) { struct mos_gem_bo_bucket *bucket = &bufmgr_gem->cache_bucket[i]; if (bucket->size >= size) { return bucket; } } return nullptr; } static void mos_gem_dump_validation_list(struct mos_bufmgr_gem *bufmgr_gem) { int i, j; for (i = 0; i < bufmgr_gem->exec_count; i++) { struct mos_linux_bo *bo = bufmgr_gem->exec_bos[i]; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; if (bo_gem->relocs == nullptr || bo_gem->softpin_target == nullptr) { MOS_DBG("%2d: %d %s(%s)\n", i, bo_gem->gem_handle, bo_gem->is_softpin ? "*" : "", bo_gem->name); continue; } for (j = 0; j < bo_gem->reloc_count; j++) { struct mos_linux_bo *target_bo = bo_gem->reloc_target_info[j].bo; struct mos_bo_gem *target_gem = (struct mos_bo_gem *) target_bo; MOS_DBG("%2d: %d %s(%s)@0x%08x %08x -> " "%d (%s)@0x%08x %08x + 0x%08x\n", i, bo_gem->gem_handle, bo_gem->is_softpin ? "*" : "", bo_gem->name, upper_32_bits(bo_gem->relocs[j].offset), lower_32_bits(bo_gem->relocs[j].offset), target_gem->gem_handle, target_gem->name, upper_32_bits(target_bo->offset64), lower_32_bits(target_bo->offset64), bo_gem->relocs[j].delta); } for (j = 0; j < bo_gem->softpin_target_count; j++) { struct mos_linux_bo *target_bo = bo_gem->softpin_target[j].bo; struct mos_bo_gem *target_gem = (struct mos_bo_gem *) target_bo; MOS_DBG("%2d: %d %s(%s) -> " "%d *(%s)@0x%08x %08x\n", i, bo_gem->gem_handle, bo_gem->is_softpin ? "*" : "", bo_gem->name, target_gem->gem_handle, target_gem->name, upper_32_bits(target_bo->offset64), lower_32_bits(target_bo->offset64)); } } } static inline void mos_gem_bo_reference(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; atomic_inc(&bo_gem->refcount); } /** * Adds the given buffer to the list of buffers to be validated (moved into the * appropriate memory type) with the next batch submission. * * If a buffer is validated multiple times in a batch submission, it ends up * with the intersection of the memory type flags and the union of the * access flags. */ static void mos_add_validate_buffer(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int index; struct drm_i915_gem_exec_object *exec_objects; struct mos_linux_bo **exec_bos; if (bo_gem->validate_index != -1) return; /* Extend the array of validation entries as necessary. */ if (bufmgr_gem->exec_count == bufmgr_gem->exec_size) { int new_size = bufmgr_gem->exec_size * 2; if (new_size == 0) new_size = 5; exec_objects = (struct drm_i915_gem_exec_object *)realloc(bufmgr_gem->exec_objects, sizeof(*bufmgr_gem->exec_objects) * new_size); if (!exec_objects) return; bufmgr_gem->exec_objects = exec_objects; exec_bos = (struct mos_linux_bo **)realloc(bufmgr_gem->exec_bos, sizeof(*bufmgr_gem->exec_bos) * new_size); if (!exec_bos) return; bufmgr_gem->exec_bos = exec_bos; bufmgr_gem->exec_size = new_size; } index = bufmgr_gem->exec_count; bo_gem->validate_index = index; /* Fill in array entry */ bufmgr_gem->exec_objects[index].handle = bo_gem->gem_handle; bufmgr_gem->exec_objects[index].relocation_count = bo_gem->reloc_count; bufmgr_gem->exec_objects[index].relocs_ptr = (uintptr_t) bo_gem->relocs; bufmgr_gem->exec_objects[index].alignment = bo->align; bufmgr_gem->exec_objects[index].offset = 0; bufmgr_gem->exec_bos[index] = bo; bufmgr_gem->exec_count++; } static void mos_add_validate_buffer2(struct mos_linux_bo *bo, int need_fence) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *)bo; int index; struct drm_i915_gem_exec_object2 *exec2_objects; struct mos_linux_bo **exec_bos; int flags = 0; if (need_fence) flags |= EXEC_OBJECT_NEEDS_FENCE; if (bo_gem->pad_to_size) flags |= EXEC_OBJECT_PAD_TO_SIZE; if (bo_gem->use_48b_address_range) flags |= EXEC_OBJECT_SUPPORTS_48B_ADDRESS; if (bo_gem->is_softpin) flags |= EXEC_OBJECT_PINNED; if (bo_gem->exec_async) flags |= EXEC_OBJECT_ASYNC; if (bo_gem->validate_index != -1) { bufmgr_gem->exec2_objects[bo_gem->validate_index].flags |= flags; return; } /* Extend the array of validation entries as necessary. */ if (bufmgr_gem->exec_count == bufmgr_gem->exec_size) { int new_size = bufmgr_gem->exec_size * 2; if (new_size == 0) new_size = 5; exec2_objects = (struct drm_i915_gem_exec_object2 *) realloc(bufmgr_gem->exec2_objects, sizeof(*bufmgr_gem->exec2_objects) * new_size); if (!exec2_objects) return; bufmgr_gem->exec2_objects = exec2_objects; exec_bos = (struct mos_linux_bo **)realloc(bufmgr_gem->exec_bos, sizeof(*bufmgr_gem->exec_bos) * new_size); if (!exec_bos) return; bufmgr_gem->exec_bos = exec_bos; bufmgr_gem->exec_size = new_size; } index = bufmgr_gem->exec_count; bo_gem->validate_index = index; /* Fill in array entry */ bufmgr_gem->exec2_objects[index].handle = bo_gem->gem_handle; bufmgr_gem->exec2_objects[index].relocation_count = bo_gem->reloc_count; bufmgr_gem->exec2_objects[index].relocs_ptr = (uintptr_t)bo_gem->relocs; bufmgr_gem->exec2_objects[index].alignment = bo->align; bufmgr_gem->exec2_objects[index].offset = bo_gem->is_softpin ? bo->offset64 : 0; bufmgr_gem->exec_bos[index] = bo; bufmgr_gem->exec2_objects[index].flags = flags; bufmgr_gem->exec2_objects[index].rsvd1 = 0; bufmgr_gem->exec2_objects[index].pad_to_size = bo_gem->pad_to_size; bufmgr_gem->exec2_objects[index].rsvd2 = 0; bufmgr_gem->exec_count++; } static void mos_add_reloc_objects(struct mos_reloc_target reloc_target) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)reloc_target.bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *)reloc_target.bo; int index; struct drm_i915_gem_exec_object2 *exec2_objects; struct mos_linux_bo **exec_bos; if (bo_gem->validate_index != -1) { bufmgr_gem->exec2_objects[bo_gem->validate_index].flags |= reloc_target.flags; return; } /* Extend the array of validation entries as necessary. */ if (bufmgr_gem->exec_count == bufmgr_gem->exec_size) { int new_size = bufmgr_gem->exec_size * 2; if (new_size == 0) new_size = 5; exec2_objects = (struct drm_i915_gem_exec_object2 *) realloc(bufmgr_gem->exec2_objects, sizeof(*bufmgr_gem->exec2_objects) * new_size); if (!exec2_objects) return; bufmgr_gem->exec2_objects = exec2_objects; exec_bos = (struct mos_linux_bo **)realloc(bufmgr_gem->exec_bos, sizeof(*bufmgr_gem->exec_bos) * new_size); if (!exec_bos) return; bufmgr_gem->exec_bos = exec_bos; bufmgr_gem->exec_size = new_size; } index = bufmgr_gem->exec_count; bo_gem->validate_index = index; /* Fill in array entry */ bufmgr_gem->exec2_objects[index].handle = bo_gem->gem_handle; bufmgr_gem->exec2_objects[index].relocation_count = bo_gem->reloc_count; bufmgr_gem->exec2_objects[index].relocs_ptr = (uintptr_t)bo_gem->relocs; bufmgr_gem->exec2_objects[index].alignment = reloc_target.bo->align; bufmgr_gem->exec2_objects[index].offset = 0; bufmgr_gem->exec_bos[index] = reloc_target.bo; bufmgr_gem->exec2_objects[index].flags = reloc_target.flags; bufmgr_gem->exec2_objects[index].rsvd1 = 0; bufmgr_gem->exec2_objects[index].pad_to_size = bo_gem->pad_to_size; bufmgr_gem->exec2_objects[index].rsvd2 = 0; bufmgr_gem->exec_count++; } static void mos_add_softpin_objects(struct mos_softpin_target softpin_target) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)softpin_target.bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *)softpin_target.bo; int index; struct drm_i915_gem_exec_object2 *exec2_objects; struct mos_linux_bo **exec_bos; if (bo_gem->validate_index != -1) { bufmgr_gem->exec2_objects[bo_gem->validate_index].flags |= softpin_target.flags; return; } /* Extend the array of validation entries as necessary. */ if (bufmgr_gem->exec_count == bufmgr_gem->exec_size) { int new_size = bufmgr_gem->exec_size * 2; if (new_size == 0) new_size = 5; exec2_objects = (struct drm_i915_gem_exec_object2 *) realloc(bufmgr_gem->exec2_objects, sizeof(*bufmgr_gem->exec2_objects) * new_size); if (!exec2_objects) return; bufmgr_gem->exec2_objects = exec2_objects; exec_bos = (struct mos_linux_bo **)realloc(bufmgr_gem->exec_bos, sizeof(*bufmgr_gem->exec_bos) * new_size); if (!exec_bos) return; bufmgr_gem->exec_bos = exec_bos; bufmgr_gem->exec_size = new_size; } index = bufmgr_gem->exec_count; bo_gem->validate_index = index; /* Fill in array entry */ bufmgr_gem->exec2_objects[index].handle = bo_gem->gem_handle; bufmgr_gem->exec2_objects[index].relocation_count = bo_gem->reloc_count; bufmgr_gem->exec2_objects[index].relocs_ptr = (uintptr_t)bo_gem->relocs; bufmgr_gem->exec2_objects[index].alignment = softpin_target.bo->align; bufmgr_gem->exec2_objects[index].offset = softpin_target.bo->offset64; bufmgr_gem->exec2_objects[index].flags = softpin_target.flags; bufmgr_gem->exec2_objects[index].pad_to_size = bo_gem->pad_to_size; bufmgr_gem->exec2_objects[index].rsvd1 = 0; bufmgr_gem->exec2_objects[index].rsvd2 = 0; bufmgr_gem->exec_bos[index] = softpin_target.bo; bufmgr_gem->exec_count++; } #define RELOC_BUF_SIZE(x) ((I915_RELOC_HEADER + x * I915_RELOC0_STRIDE) * \ sizeof(uint32_t)) static void mos_bo_gem_set_in_aperture_size(struct mos_bufmgr_gem *bufmgr_gem, struct mos_bo_gem *bo_gem, unsigned int alignment) { unsigned int size; assert(!bo_gem->used_as_reloc_target); /* The older chipsets are far-less flexible in terms of tiling, * and require tiled buffer to be size aligned in the aperture. * This means that in the worst possible case we will need a hole * twice as large as the object in order for it to fit into the * aperture. Optimal packing is for wimps. */ size = bo_gem->bo.size; bo_gem->reloc_tree_size = size + alignment; } static int mos_setup_reloc_list(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; unsigned int max_relocs = bufmgr_gem->max_relocs; if (bo->size / 4 < max_relocs) max_relocs = bo->size / 4; bo_gem->relocs = (struct drm_i915_gem_relocation_entry *)malloc(max_relocs * sizeof(struct drm_i915_gem_relocation_entry)); bo_gem->reloc_target_info = (struct mos_reloc_target *)malloc(max_relocs * sizeof(struct mos_reloc_target)); if (bo_gem->relocs == nullptr || bo_gem->reloc_target_info == nullptr) { bo_gem->has_error = true; free (bo_gem->relocs); bo_gem->relocs = nullptr; free (bo_gem->reloc_target_info); bo_gem->reloc_target_info = nullptr; return 1; } return 0; } static int mos_gem_bo_busy(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_busy busy; int ret; if (bo_gem->reusable && bo_gem->idle) return false; memclear(busy); busy.handle = bo_gem->gem_handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_BUSY, &busy); if (ret == 0) { bo_gem->idle = !busy.busy; return busy.busy; } else { return false; } return (ret == 0 && busy.busy); } static int mos_gem_bo_madvise_internal(struct mos_bufmgr_gem *bufmgr_gem, struct mos_bo_gem *bo_gem, int state) { struct drm_i915_gem_madvise madv; memclear(madv); madv.handle = bo_gem->gem_handle; madv.madv = state; madv.retained = 1; drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_MADVISE, &madv); return madv.retained; } static int mos_gem_bo_madvise(struct mos_linux_bo *bo, int madv) { return mos_gem_bo_madvise_internal ((struct mos_bufmgr_gem *) bo->bufmgr, (struct mos_bo_gem *) bo, madv); } /* drop the oldest entries that have been purged by the kernel */ static void mos_gem_bo_cache_purge_bucket(struct mos_bufmgr_gem *bufmgr_gem, struct mos_gem_bo_bucket *bucket) { while (!DRMLISTEMPTY(&bucket->head)) { struct mos_bo_gem *bo_gem; bo_gem = DRMLISTENTRY(struct mos_bo_gem, bucket->head.next, head); if (mos_gem_bo_madvise_internal (bufmgr_gem, bo_gem, I915_MADV_DONTNEED)) break; DRMLISTDEL(&bo_gem->head); mos_gem_bo_free(&bo_gem->bo); } } static enum mos_memory_zone mos_gem_bo_memzone_for_address(uint64_t address) { if (address >= MEMZONE_DEVICE_START) return MEMZONE_DEVICE; else return MEMZONE_SYS; } /** * Allocate a section of virtual memory for a buffer, assigning an address. */ static uint64_t mos_gem_bo_vma_alloc(struct mos_bufmgr *bufmgr, enum mos_memory_zone memzone, uint64_t size, uint64_t alignment) { CHK_CONDITION(bufmgr == nullptr, "nullptr bufmgr.\n", 0); struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; /* Force alignment to be some number of pages */ alignment = ALIGN(alignment, PAGE_SIZE); uint64_t addr = mos_vma_heap_alloc(&bufmgr_gem->vma_heap[memzone], size, alignment); // currently only support 48bit range address CHK_CONDITION((addr >> 48ull) != 0, "invalid address, over 48bit range.\n", 0); CHK_CONDITION((addr >> (memzone == MEMZONE_SYS ? 40ull : 41ull)) != 0, "invalid address, over memory zone range.\n", 0); CHK_CONDITION((addr % alignment) != 0, "invalid address, not meet aligment requirement.\n", 0); return addr; } static void mos_gem_bo_vma_free(struct mos_bufmgr *bufmgr, uint64_t address, uint64_t size) { CHK_CONDITION(bufmgr == nullptr, "nullptr bufmgr.\n", ); struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; CHK_CONDITION(address == 0ull, "invalid address.\n", ); enum mos_memory_zone memzone = mos_gem_bo_memzone_for_address(address); mos_vma_heap_free(&bufmgr_gem->vma_heap[memzone], address, size); } drm_export struct mos_linux_bo * mos_gem_bo_alloc_internal(struct mos_bufmgr *bufmgr, struct mos_drm_bo_alloc *alloc) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; struct mos_bo_gem *bo_gem; unsigned int page_size = getpagesize(); int ret; struct mos_gem_bo_bucket *bucket; bool alloc_from_cache; unsigned long bo_size; bool for_render = false; if (alloc->ext.flags & BO_ALLOC_FOR_RENDER) for_render = true; /* Round the allocated size up to a power of two number of pages. */ bucket = mos_gem_bo_bucket_for_size(bufmgr_gem, alloc->size); /* If we don't have caching at this size, don't actually round the * allocation up. */ if (bucket == nullptr) { bo_size = alloc->size; if (bo_size < page_size) bo_size = page_size; } else { bo_size = bucket->size; } if(GetDrmMode())//libdrm_mock { pthread_mutex_lock(&bufmgr_gem->lock); bo_gem = (struct mos_bo_gem *)calloc(1, sizeof(*bo_gem)); if (!bo_gem) return nullptr; bo_gem->bo.size = bo_size; bo_gem->bo.handle = -1; bo_gem->bo.bufmgr = bufmgr; bo_gem->bo.align = alloc->alignment; #ifdef __cplusplus bo_gem->bo.virt = malloc(bo_size); bo_gem->mem_virtual = bo_gem->bo.virt; #else bo_gem->bo.virtual = malloc(bo_size); bo_gem->mem_virtual = bo_gem->bo.virtual; #endif atomic_set(&bo_gem->refcount, 1); pthread_mutex_unlock(&bufmgr_gem->lock); return &bo_gem->bo; } pthread_mutex_lock(&bufmgr_gem->lock); /* Get a buffer out of the cache if available */ retry: alloc_from_cache = false; if (bucket != nullptr && !DRMLISTEMPTY(&bucket->head)) { if (for_render) { /* Allocate new render-target BOs from the tail (MRU) * of the list, as it will likely be hot in the GPU * cache and in the aperture for us. */ bo_gem = DRMLISTENTRY(struct mos_bo_gem, bucket->head.prev, head); DRMLISTDEL(&bo_gem->head); alloc_from_cache = true; bo_gem->bo.align = alloc->alignment; } else { assert(alloc->alignment == 0); /* For non-render-target BOs (where we're probably * going to map it first thing in order to fill it * with data), check if the last BO in the cache is * unbusy, and only reuse in that case. Otherwise, * allocating a new buffer is probably faster than * waiting for the GPU to finish. */ bo_gem = DRMLISTENTRY(struct mos_bo_gem, bucket->head.next, head); if (!mos_gem_bo_busy(&bo_gem->bo)) { alloc_from_cache = true; DRMLISTDEL(&bo_gem->head); } } if (alloc_from_cache) { if (!mos_gem_bo_madvise_internal (bufmgr_gem, bo_gem, I915_MADV_WILLNEED)) { mos_gem_bo_free(&bo_gem->bo); mos_gem_bo_cache_purge_bucket(bufmgr_gem, bucket); goto retry; } if (mos_gem_bo_set_tiling_internal(&bo_gem->bo, alloc->ext.tiling_mode, alloc->stride)) { mos_gem_bo_free(&bo_gem->bo); goto retry; } } } pthread_mutex_unlock(&bufmgr_gem->lock); if (!alloc_from_cache) { struct drm_i915_gem_create create; bo_gem = (struct mos_bo_gem *)calloc(1, sizeof(*bo_gem)); if (!bo_gem) return nullptr; bo_gem->bo.size = bo_size; memclear(create); create.size = bo_size; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CREATE, &create); bo_gem->gem_handle = create.handle; bo_gem->bo.handle = bo_gem->gem_handle; if (ret != 0) { free(bo_gem); return nullptr; } bo_gem->bo.bufmgr = bufmgr; bo_gem->bo.align = alloc->alignment; bo_gem->tiling_mode = I915_TILING_NONE; bo_gem->swizzle_mode = I915_BIT_6_SWIZZLE_NONE; bo_gem->stride = 0; /* drm_intel_gem_bo_free calls DRMLISTDEL() for an uninitialized list (vma_list), so better set the list head here */ DRMINITLISTHEAD(&bo_gem->name_list); if (mos_gem_bo_set_tiling_internal(&bo_gem->bo, alloc->ext.tiling_mode, alloc->stride)) { mos_gem_bo_free(&bo_gem->bo); return nullptr; } } bo_gem->name = alloc->name; atomic_set(&bo_gem->refcount, 1); bo_gem->validate_index = -1; bo_gem->reloc_tree_fences = 0; bo_gem->used_as_reloc_target = false; bo_gem->has_error = false; bo_gem->reusable = true; bo_gem->use_48b_address_range = bufmgr_gem->bufmgr.bo_use_48b_address_range ? true : false; mos_bo_gem_set_in_aperture_size(bufmgr_gem, bo_gem, alloc->alignment); if (bufmgr_gem->use_softpin) { mos_bo_set_softpin(&bo_gem->bo); } MOS_DBG("bo_create: buf %d (%s) %ldb\n", bo_gem->gem_handle, bo_gem->name, alloc->size); return &bo_gem->bo; } static struct mos_linux_bo * mos_gem_bo_alloc(struct mos_bufmgr *bufmgr, struct mos_drm_bo_alloc *alloc) { alloc->ext.flags = 0; alloc->alignment = 0; alloc->stride = 0; return mos_gem_bo_alloc_internal(bufmgr, alloc); } static struct mos_linux_bo * mos_gem_bo_alloc_tiled(struct mos_bufmgr *bufmgr, struct mos_drm_bo_alloc_tiled *alloc_tiled) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; unsigned long size, stride; uint32_t tiling; do { unsigned long aligned_y, height_alignment; tiling = alloc_tiled->ext.tiling_mode; /* If we're tiled, our allocations are in 8 or 32-row blocks, * so failure to align our height means that we won't allocate * enough pages. * * If we're untiled, we still have to align to 2 rows high * because the data port accesses 2x2 blocks even if the * bottom row isn't to be rendered, so failure to align means * we could walk off the end of the GTT and fault. This is * documented on 965, and may be the case on older chipsets * too so we try to be careful. */ aligned_y = alloc_tiled->y; height_alignment = 2; if (tiling == I915_TILING_X || (IS_915(bufmgr_gem->pci_device) && tiling == I915_TILING_Y)) height_alignment = 8; else if (tiling == I915_TILING_Y) height_alignment = 32; aligned_y = ALIGN(alloc_tiled->y, height_alignment); stride = alloc_tiled->x * alloc_tiled->cpp; stride = mos_gem_bo_tile_pitch(bufmgr_gem, stride, &alloc_tiled->ext.tiling_mode); size = stride * aligned_y; size = mos_gem_bo_tile_size(bufmgr_gem, size, &alloc_tiled->ext.tiling_mode); } while (alloc_tiled->ext.tiling_mode != tiling); alloc_tiled->pitch = stride; if (tiling == I915_TILING_NONE) stride = 0; struct mos_drm_bo_alloc alloc; alloc.name = alloc_tiled->name; alloc.size = size; alloc.stride = stride; alloc.ext = alloc_tiled->ext; return mos_gem_bo_alloc_internal(bufmgr, &alloc); } static struct mos_linux_bo * mos_gem_bo_alloc_userptr(struct mos_bufmgr *bufmgr, struct mos_drm_bo_alloc_userptr *alloc_uptr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; struct mos_bo_gem *bo_gem; int ret; struct drm_i915_gem_userptr userptr; /* Tiling with userptr surfaces is not supported * on all hardware so refuse it for time being. */ if (alloc_uptr->tiling_mode != I915_TILING_NONE) return nullptr; bo_gem = (struct mos_bo_gem *)calloc(1, sizeof(*bo_gem)); if (!bo_gem) return nullptr; bo_gem->bo.size = alloc_uptr->size; memclear(userptr); userptr.user_ptr = (__u64)((unsigned long)alloc_uptr->addr); userptr.user_size = alloc_uptr->size; userptr.flags = alloc_uptr->flags; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_USERPTR, &userptr); if (ret != 0) { MOS_DBG("bo_create_userptr: " "ioctl failed with user ptr %p size 0x%lx, " "user flags 0x%lx\n", alloc_uptr->addr, alloc_uptr->size, alloc_uptr->flags); free(bo_gem); return nullptr; } bo_gem->gem_handle = userptr.handle; bo_gem->bo.handle = bo_gem->gem_handle; bo_gem->bo.bufmgr = bufmgr; bo_gem->is_userptr = true; #ifdef __cplusplus bo_gem->bo.virt = alloc_uptr->addr; #else bo_gem->bo.virtual = alloc_uptr->addr; #endif /* Save the address provided by user */ bo_gem->user_virtual = alloc_uptr->addr; bo_gem->tiling_mode = I915_TILING_NONE; bo_gem->swizzle_mode = I915_BIT_6_SWIZZLE_NONE; bo_gem->stride = 0; DRMINITLISTHEAD(&bo_gem->name_list); bo_gem->name = alloc_uptr->name; atomic_set(&bo_gem->refcount, 1); bo_gem->validate_index = -1; bo_gem->reloc_tree_fences = 0; bo_gem->used_as_reloc_target = false; bo_gem->has_error = false; bo_gem->reusable = false; bo_gem->use_48b_address_range = bufmgr_gem->bufmgr.bo_use_48b_address_range ? true : false; mos_bo_gem_set_in_aperture_size(bufmgr_gem, bo_gem, 0); if (bufmgr_gem->use_softpin) { mos_bo_set_softpin(&bo_gem->bo); } MOS_DBG("bo_create_userptr: " "ptr %p buf %d (%s) size %ldb, stride 0x%x, tile mode %d\n", alloc_uptr->addr, bo_gem->gem_handle, bo_gem->name, alloc_uptr->size, alloc_uptr->stride, alloc_uptr->tiling_mode); return &bo_gem->bo; } static bool has_userptr(struct mos_bufmgr_gem *bufmgr_gem) { int ret; void *ptr; long pgsz; struct drm_i915_gem_userptr userptr; pgsz = sysconf(_SC_PAGESIZE); assert(pgsz > 0); ret = posix_memalign(&ptr, pgsz, pgsz); if (ret) { MOS_DBG("Failed to get a page (%ld) for userptr detection!\n", pgsz); return false; } memclear(userptr); userptr.user_ptr = (__u64)(unsigned long)ptr; userptr.user_size = pgsz; retry: ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_USERPTR, &userptr); if (ret) { if (errno == ENODEV && userptr.flags == 0) { userptr.flags = I915_USERPTR_UNSYNCHRONIZED; goto retry; } free(ptr); return false; } /* We don't release the userptr bo here as we want to keep the * kernel mm tracking alive for our lifetime. The first time we * create a userptr object the kernel has to install a mmu_notifer * which is a heavyweight operation (e.g. it requires taking all * mm_locks and stop_machine()). */ bufmgr_gem->userptr_active.ptr = ptr; bufmgr_gem->userptr_active.handle = userptr.handle; return true; } static struct mos_linux_bo * check_bo_alloc_userptr(struct mos_bufmgr *bufmgr, struct mos_drm_bo_alloc_userptr *alloc_uptr) { if (has_userptr((struct mos_bufmgr_gem *)bufmgr)) bufmgr->bo_alloc_userptr = mos_gem_bo_alloc_userptr; else bufmgr->bo_alloc_userptr = nullptr; return mos_bo_alloc_userptr(bufmgr, alloc_uptr); } /** * Returns a drm_intel_bo wrapping the given buffer object handle. * * This can be used when one application needs to pass a buffer object * to another. */ static struct mos_linux_bo * mos_bufmgr_bo_gem_create_from_name(struct mos_bufmgr *bufmgr, const char *name, unsigned int handle) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; struct mos_bo_gem *bo_gem; int ret; struct drm_gem_open open_arg; struct drm_i915_gem_get_tiling get_tiling; drmMMListHead *list; /* At the moment most applications only have a few named bo. * For instance, in a DRI client only the render buffers passed * between X and the client are named. And since X returns the * alternating names for the front/back buffer a linear search * provides a sufficiently fast match. */ pthread_mutex_lock(&bufmgr_gem->lock); for (list = bufmgr_gem->named.next; list != &bufmgr_gem->named; list = list->next) { bo_gem = DRMLISTENTRY(struct mos_bo_gem, list, name_list); if (bo_gem->global_name == handle) { mos_gem_bo_reference(&bo_gem->bo); pthread_mutex_unlock(&bufmgr_gem->lock); return &bo_gem->bo; } } memclear(open_arg); open_arg.name = handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_GEM_OPEN, &open_arg); if (ret != 0) { MOS_DBG("Couldn't reference %s handle 0x%08x: %s\n", name, handle, strerror(errno)); pthread_mutex_unlock(&bufmgr_gem->lock); return nullptr; } /* Now see if someone has used a prime handle to get this * object from the kernel before by looking through the list * again for a matching gem_handle */ for (list = bufmgr_gem->named.next; list != &bufmgr_gem->named; list = list->next) { bo_gem = DRMLISTENTRY(struct mos_bo_gem, list, name_list); if (bo_gem->gem_handle == open_arg.handle) { mos_gem_bo_reference(&bo_gem->bo); pthread_mutex_unlock(&bufmgr_gem->lock); return &bo_gem->bo; } } bo_gem = (struct mos_bo_gem *)calloc(1, sizeof(*bo_gem)); if (!bo_gem) { pthread_mutex_unlock(&bufmgr_gem->lock); return nullptr; } bo_gem->bo.size = open_arg.size; bo_gem->bo.offset = 0; bo_gem->bo.offset64 = 0; #if defined(__cplusplus) bo_gem->bo.virt = nullptr; #else bo_gem->bo.virtual = nullptr; #endif bo_gem->bo.bufmgr = bufmgr; bo_gem->name = name; atomic_set(&bo_gem->refcount, 1); bo_gem->validate_index = -1; bo_gem->gem_handle = open_arg.handle; bo_gem->bo.handle = open_arg.handle; bo_gem->global_name = handle; bo_gem->reusable = false; bo_gem->use_48b_address_range = bufmgr_gem->bufmgr.bo_use_48b_address_range ? true : false; memclear(get_tiling); get_tiling.handle = bo_gem->gem_handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_GET_TILING, &get_tiling); if (ret != 0) { mos_gem_bo_unreference(&bo_gem->bo); pthread_mutex_unlock(&bufmgr_gem->lock); return nullptr; } bo_gem->tiling_mode = get_tiling.tiling_mode; bo_gem->swizzle_mode = get_tiling.swizzle_mode; /* XXX stride is unknown */ mos_bo_gem_set_in_aperture_size(bufmgr_gem, bo_gem, 0); if (bufmgr_gem->use_softpin) { mos_bo_set_softpin(&bo_gem->bo); } DRMLISTADDTAIL(&bo_gem->name_list, &bufmgr_gem->named); pthread_mutex_unlock(&bufmgr_gem->lock); MOS_DBG("bo_create_from_handle: %d (%s)\n", handle, bo_gem->name); return &bo_gem->bo; } static void mos_gem_bo_free(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_gem_close close; int ret; if(GetDrmMode())//libdrm_mock { free(bo_gem->mem_virtual); free(bo); return; } if (bo_gem->mem_virtual) { VG(VALGRIND_MAKE_MEM_NOACCESS(bo_gem->mem_virtual, 0)); drm_munmap(bo_gem->mem_virtual, bo_gem->bo.size); bo_gem->mem_virtual = nullptr; } if (bo_gem->gtt_virtual) { VG(VALGRIND_MAKE_MEM_NOACCESS(bo_gem->gtt_virtual, 0)); drm_munmap(bo_gem->gtt_virtual, bo_gem->bo.size); bo_gem->gtt_virtual = nullptr; } if (bo_gem->mem_wc_virtual) { VG(VALGRIND_MAKE_MEM_NOACCESS(bo_gem->mem_wc_virtual, 0)); drm_munmap(bo_gem->mem_wc_virtual, bo_gem->bo.size); bo_gem->mem_wc_virtual = nullptr; } /* Close this object */ memclear(close); close.handle = bo_gem->gem_handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_GEM_CLOSE, &close); if (ret != 0) { MOS_DBG("DRM_IOCTL_GEM_CLOSE %d failed (%s): %s\n", bo_gem->gem_handle, bo_gem->name, strerror(errno)); } if (bufmgr_gem->use_softpin) { /* Return the VMA for reuse */ mos_gem_bo_vma_free(bo->bufmgr, bo->offset64, bo->size); } free(bo); } static void mos_gem_bo_mark_mmaps_incoherent(struct mos_linux_bo *bo) { #if HAVE_VALGRIND struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; if (bo_gem->mem_virtual) VALGRIND_MAKE_MEM_NOACCESS(bo_gem->mem_virtual, bo->size); if (bo_gem->gtt_virtual) VALGRIND_MAKE_MEM_NOACCESS(bo_gem->gtt_virtual, bo->size); if (bo_gem->mem_wc_virtual) VALGRIND_MAKE_MEM_NOACCESS(bo_gem->mem_wc_virtual, bo->size); #endif } /** Frees all cached buffers significantly older than @time. */ static void mos_gem_cleanup_bo_cache(struct mos_bufmgr_gem *bufmgr_gem, time_t time) { int i; if (bufmgr_gem->time == time) return; for (i = 0; i < bufmgr_gem->num_buckets; i++) { struct mos_gem_bo_bucket *bucket = &bufmgr_gem->cache_bucket[i]; while (!DRMLISTEMPTY(&bucket->head)) { struct mos_bo_gem *bo_gem; bo_gem = DRMLISTENTRY(struct mos_bo_gem, bucket->head.next, head); if (time - bo_gem->free_time <= 1) break; DRMLISTDEL(&bo_gem->head); mos_gem_bo_free(&bo_gem->bo); } } bufmgr_gem->time = time; } drm_export void mos_gem_bo_unreference_final(struct mos_linux_bo *bo, time_t time) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct mos_gem_bo_bucket *bucket; int i; /* Unreference all the target buffers */ for (i = 0; i < bo_gem->reloc_count; i++) { if (bo_gem->reloc_target_info[i].bo != bo) { mos_gem_bo_unreference_locked_timed(bo_gem-> reloc_target_info[i].bo, time); } } for (i = 0; i < bo_gem->softpin_target_count; i++) mos_gem_bo_unreference_locked_timed(bo_gem->softpin_target[i].bo, time); bo_gem->reloc_count = 0; bo_gem->used_as_reloc_target = false; bo_gem->softpin_target_count = 0; bo_gem->exec_async = false; MOS_DBG("bo_unreference final: %d (%s)\n", bo_gem->gem_handle, bo_gem->name); bo_gem->pad_to_size = 0; /* release memory associated with this object */ if (bo_gem->reloc_target_info) { free(bo_gem->reloc_target_info); bo_gem->reloc_target_info = nullptr; } if (bo_gem->relocs) { free(bo_gem->relocs); bo_gem->relocs = nullptr; } if (bo_gem->softpin_target) { free(bo_gem->softpin_target); bo_gem->softpin_target = nullptr; bo_gem->softpin_target_size = 0; } if(GetDrmMode()){ mos_gem_bo_free(bo); return; } /* Clear any left-over mappings */ if (bo_gem->map_count) { MOS_DBG("bo freed with non-zero map-count %d\n", bo_gem->map_count); bo_gem->map_count = 0; mos_gem_bo_mark_mmaps_incoherent(bo); } DRMLISTDEL(&bo_gem->name_list); bucket = mos_gem_bo_bucket_for_size(bufmgr_gem, bo->size); /* Put the buffer into our internal cache for reuse if we can. */ if (bufmgr_gem->bo_reuse && bo_gem->reusable && bucket != nullptr && mos_gem_bo_madvise_internal(bufmgr_gem, bo_gem, I915_MADV_DONTNEED)) { bo_gem->free_time = time; bo_gem->name = nullptr; bo_gem->validate_index = -1; DRMLISTADDTAIL(&bo_gem->head, &bucket->head); } else { mos_gem_bo_free(bo); } } static void mos_gem_bo_unreference_locked_timed(struct mos_linux_bo *bo, time_t time) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; assert(atomic_read(&bo_gem->refcount) > 0); if (atomic_dec_and_test(&bo_gem->refcount)) mos_gem_bo_unreference_final(bo, time); } static void mos_gem_bo_unreference(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; assert(atomic_read(&bo_gem->refcount) > 0); if (atomic_add_unless(&bo_gem->refcount, -1, 1)) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct timespec time; clock_gettime(CLOCK_MONOTONIC, &time); pthread_mutex_lock(&bufmgr_gem->lock); if (atomic_dec_and_test(&bo_gem->refcount)) { mos_gem_bo_unreference_final(bo, time.tv_sec); mos_gem_cleanup_bo_cache(bufmgr_gem, time.tv_sec); } pthread_mutex_unlock(&bufmgr_gem->lock); } } static int map_wc(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int ret; if (bo_gem->is_userptr) return -EINVAL; if (!bufmgr_gem->has_ext_mmap) return -EINVAL; /* Get a mapping of the buffer if we haven't before. */ if (bo_gem->mem_wc_virtual == nullptr) { struct drm_i915_gem_mmap mmap_arg; MOS_DBG("bo_map_wc: mmap %d (%s), map_count=%d\n", bo_gem->gem_handle, bo_gem->name, bo_gem->map_count); memclear(mmap_arg); mmap_arg.handle = bo_gem->gem_handle; /* To indicate the uncached virtual mapping to KMD */ mmap_arg.flags = I915_MMAP_WC; mmap_arg.size = bo->size; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_MMAP, &mmap_arg); if (ret != 0) { ret = -errno; MOS_DBG("%s:%d: Error mapping buffer %d (%s): %s .\n", __FILE__, __LINE__, bo_gem->gem_handle, bo_gem->name, strerror(errno)); return ret; } VG(VALGRIND_MALLOCLIKE_BLOCK(mmap_arg.addr_ptr, mmap_arg.size, 0, 1)); bo_gem->mem_wc_virtual = (void *)(uintptr_t) mmap_arg.addr_ptr; } #ifdef __cplusplus bo->virt = bo_gem->mem_wc_virtual; #else bo->virtual = bo_gem->mem_wc_virtual; #endif MOS_DBG("bo_map_wc: %d (%s) -> %p\n", bo_gem->gem_handle, bo_gem->name, bo_gem->mem_wc_virtual); return 0; } /* To be used in a similar way to mmap_gtt */ drm_export int mos_gem_bo_map_wc(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_set_domain set_domain; int ret; if(GetDrmMode())//libdrm_mock { #ifdef __cplusplus bo->virt = bo_gem->mem_virtual; #else bo->virtual = bo_gem->mem_virtual; #endif bo_gem->map_count++; return 0; } pthread_mutex_lock(&bufmgr_gem->lock); ret = map_wc(bo); if (ret) { pthread_mutex_unlock(&bufmgr_gem->lock); return ret; } /* Now move it to the GTT domain so that the GPU and CPU * caches are flushed and the GPU isn't actively using the * buffer. * * The domain change is done even for the objects which * are not bounded. For them first the pages are acquired, * before the domain change. */ memclear(set_domain); set_domain.handle = bo_gem->gem_handle; set_domain.read_domains = I915_GEM_DOMAIN_GTT; set_domain.write_domain = I915_GEM_DOMAIN_GTT; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_SET_DOMAIN, &set_domain); if (ret != 0) { MOS_DBG("%s:%d: Error setting domain %d: %s\n", __FILE__, __LINE__, bo_gem->gem_handle, strerror(errno)); } mos_gem_bo_mark_mmaps_incoherent(bo); VG(VALGRIND_MAKE_MEM_DEFINED(bo_gem->mem_wc_virtual, bo->size)); pthread_mutex_unlock(&bufmgr_gem->lock); return 0; } drm_export int mos_gem_bo_map(struct mos_linux_bo *bo, int write_enable) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_set_domain set_domain; int ret; if(GetDrmMode())//libdrm_mock { #ifdef __cplusplus bo->virt = bo_gem->mem_virtual; #else bo->virtual = bo_gem->mem_virtual; #endif bo_gem->map_count++; return 0; } if (bo_gem->is_userptr) { /* Return the same user ptr */ #ifdef __cplusplus bo->virt = bo_gem->user_virtual; #else bo->virtual = bo_gem->user_virtual; #endif return 0; } pthread_mutex_lock(&bufmgr_gem->lock); if (!bo_gem->mem_virtual) { struct drm_i915_gem_mmap mmap_arg; MOS_DBG("bo_map: %d (%s), map_count=%d\n", bo_gem->gem_handle, bo_gem->name, bo_gem->map_count); memclear(mmap_arg); mmap_arg.handle = bo_gem->gem_handle; mmap_arg.size = bo->size; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_MMAP, &mmap_arg); if (ret != 0) { ret = -errno; MOS_DBG("%s:%d: Error mapping buffer %d (%s): %s .\n", __FILE__, __LINE__, bo_gem->gem_handle, bo_gem->name, strerror(errno)); pthread_mutex_unlock(&bufmgr_gem->lock); return ret; } VG(VALGRIND_MALLOCLIKE_BLOCK(mmap_arg.addr_ptr, mmap_arg.size, 0, 1)); bo_gem->mem_virtual = (void *)(uintptr_t) mmap_arg.addr_ptr; } MOS_DBG("bo_map: %d (%s) -> %p\n", bo_gem->gem_handle, bo_gem->name, bo_gem->mem_virtual); #ifdef __cplusplus bo->virt = bo_gem->mem_virtual; #else bo->virtual = bo_gem->mem_virtual; #endif memclear(set_domain); set_domain.handle = bo_gem->gem_handle; set_domain.read_domains = I915_GEM_DOMAIN_CPU; if (write_enable) set_domain.write_domain = I915_GEM_DOMAIN_CPU; else set_domain.write_domain = 0; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_SET_DOMAIN, &set_domain); if (ret != 0) { MOS_DBG("%s:%d: Error setting to CPU domain %d: %s\n", __FILE__, __LINE__, bo_gem->gem_handle, strerror(errno)); } if (write_enable) bo_gem->mapped_cpu_write = true; mos_gem_bo_mark_mmaps_incoherent(bo); VG(VALGRIND_MAKE_MEM_DEFINED(bo_gem->mem_virtual, bo->size)); pthread_mutex_unlock(&bufmgr_gem->lock); return 0; } drm_export int map_gtt(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int ret; if(GetDrmMode())//libdrm_mock { #ifdef __cplusplus bo->virt = bo_gem->mem_virtual; #else bo->virtual = bo_gem->mem_virtual; #endif bo_gem->map_count++; return 0; } if (bo_gem->is_userptr) return -EINVAL; /* Get a mapping of the buffer if we haven't before. */ if (bo_gem->gtt_virtual == nullptr) { struct drm_i915_gem_mmap_gtt mmap_arg; MOS_DBG("bo_map_gtt: mmap %d (%s), map_count=%d\n", bo_gem->gem_handle, bo_gem->name, bo_gem->map_count); memclear(mmap_arg); mmap_arg.handle = bo_gem->gem_handle; /* Get the fake offset back... */ ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_MMAP_GTT, &mmap_arg); if (ret != 0) { ret = -errno; MOS_DBG("%s:%d: Error preparing buffer map %d (%s): %s .\n", __FILE__, __LINE__, bo_gem->gem_handle, bo_gem->name, strerror(errno)); return ret; } /* and mmap it */ bo_gem->gtt_virtual = drm_mmap(0, bo->size, PROT_READ | PROT_WRITE, MAP_SHARED, bufmgr_gem->fd, mmap_arg.offset); if (bo_gem->gtt_virtual == MAP_FAILED) { bo_gem->gtt_virtual = nullptr; ret = -errno; MOS_DBG("%s:%d: Error mapping buffer %d (%s): %s .\n", __FILE__, __LINE__, bo_gem->gem_handle, bo_gem->name, strerror(errno)); return ret; } } #ifdef __cplusplus bo->virt = bo_gem->gtt_virtual; #else bo->virtual = bo_gem->gtt_virtual; #endif MOS_DBG("bo_map_gtt: %d (%s) -> %p\n", bo_gem->gem_handle, bo_gem->name, bo_gem->gtt_virtual); return 0; } static int mos_gem_bo_map_gtt(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_set_domain set_domain; int ret; pthread_mutex_lock(&bufmgr_gem->lock); ret = map_gtt(bo); if (ret) { pthread_mutex_unlock(&bufmgr_gem->lock); return ret; } /* Now move it to the GTT domain so that the GPU and CPU * caches are flushed and the GPU isn't actively using the * buffer. * * The pagefault handler does this domain change for us when * it has unbound the BO from the GTT, but it's up to us to * tell it when we're about to use things if we had done * rendering and it still happens to be bound to the GTT. */ memclear(set_domain); set_domain.handle = bo_gem->gem_handle; set_domain.read_domains = I915_GEM_DOMAIN_GTT; set_domain.write_domain = I915_GEM_DOMAIN_GTT; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_SET_DOMAIN, &set_domain); if (ret != 0) { MOS_DBG("%s:%d: Error setting domain %d: %s\n", __FILE__, __LINE__, bo_gem->gem_handle, strerror(errno)); } mos_gem_bo_mark_mmaps_incoherent(bo); VG(VALGRIND_MAKE_MEM_DEFINED(bo_gem->gtt_virtual, bo->size)); pthread_mutex_unlock(&bufmgr_gem->lock); return 0; } /** * Performs a mapping of the buffer object like the normal GTT * mapping, but avoids waiting for the GPU to be done reading from or * rendering to the buffer. * * This is used in the implementation of GL_ARB_map_buffer_range: The * user asks to create a buffer, then does a mapping, fills some * space, runs a drawing command, then asks to map it again without * synchronizing because it guarantees that it won't write over the * data that the GPU is busy using (or, more specifically, that if it * does write over the data, it acknowledges that rendering is * undefined). */ static int mos_gem_bo_map_unsynchronized(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; #ifdef HAVE_VALGRIND struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; #endif int ret; /* If the CPU cache isn't coherent with the GTT, then use a * regular synchronized mapping. The problem is that we don't * track where the buffer was last used on the CPU side in * terms of drm_intel_bo_map vs drm_intel_gem_bo_map_gtt, so * we would potentially corrupt the buffer even when the user * does reasonable things. */ if (!bufmgr_gem->has_llc) return mos_gem_bo_map_gtt(bo); pthread_mutex_lock(&bufmgr_gem->lock); ret = map_gtt(bo); if (ret == 0) { mos_gem_bo_mark_mmaps_incoherent(bo); VG(VALGRIND_MAKE_MEM_DEFINED(bo_gem->gtt_virtual, bo->size)); } pthread_mutex_unlock(&bufmgr_gem->lock); return ret; } drm_export int mos_gem_bo_unmap(struct mos_linux_bo *bo) { struct mos_bufmgr_gem *bufmgr_gem; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int ret = 0; if (bo == nullptr) return 0; if (bo_gem->is_userptr) return 0; bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; if(GetDrmMode())//libdrm_mock { --bo_gem->map_count ; return 0; } pthread_mutex_lock(&bufmgr_gem->lock); if (bo_gem->map_count <= 0) { MOS_DBG("attempted to unmap an unmapped bo\n"); pthread_mutex_unlock(&bufmgr_gem->lock); /* Preserve the old behaviour of just treating this as a * no-op rather than reporting the error. */ return 0; } if (bo_gem->mapped_cpu_write) { struct drm_i915_gem_sw_finish sw_finish; /* Cause a flush to happen if the buffer's pinned for * scanout, so the results show up in a timely manner. * Unlike GTT set domains, this only does work if the * buffer should be scanout-related. */ memclear(sw_finish); sw_finish.handle = bo_gem->gem_handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_SW_FINISH, &sw_finish); ret = ret == -1 ? -errno : 0; bo_gem->mapped_cpu_write = false; } /* We need to unmap after every innovation as we cannot track * an open vma for every bo as that will exhaasut the system * limits and cause later failures. */ if (--bo_gem->map_count == 0) { mos_gem_bo_mark_mmaps_incoherent(bo); #ifdef __cplusplus bo->virt = nullptr; #else bo->virtual = nullptr; #endif } pthread_mutex_unlock(&bufmgr_gem->lock); return ret; } static int mos_gem_bo_unmap_wc(struct mos_linux_bo *bo) { return mos_gem_bo_unmap(bo); } static int mos_gem_bo_unmap_gtt(struct mos_linux_bo *bo) { return mos_gem_bo_unmap(bo); } /** Waits for all GPU rendering with the object to have completed. */ static void mos_gem_bo_wait_rendering(struct mos_linux_bo *bo) { mos_gem_bo_start_gtt_access(bo, 1); } /** * Waits on a BO for the given amount of time. * * @bo: buffer object to wait for * @timeout_ns: amount of time to wait in nanoseconds. * If value is less than 0, an infinite wait will occur. * * Returns 0 if the wait was successful ie. the last batch referencing the * object has completed within the allotted time. Otherwise some negative return * value describes the error. Of particular interest is -ETIME when the wait has * failed to yield the desired result. * * Similar to drm_intel_gem_bo_wait_rendering except a timeout parameter allows * the operation to give up after a certain amount of time. Another subtle * difference is the internal locking semantics are different (this variant does * not hold the lock for the duration of the wait). This makes the wait subject * to a larger userspace race window. * * The implementation shall wait until the object is no longer actively * referenced within a batch buffer at the time of the call. The wait will * not guarantee that the buffer is re-issued via another thread, or an flinked * handle. Userspace must make sure this race does not occur if such precision * is important. * * Note that some kernels have broken the inifite wait for negative values * promise, upgrade to latest stable kernels if this is the case. */ static int mos_gem_bo_wait(struct mos_linux_bo *bo, int64_t timeout_ns) { if(GetDrmMode()) return 0; //libdrm_mock struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_wait wait; int ret; if (!bufmgr_gem->has_wait_timeout) { MOS_DBG("%s:%d: Timed wait is not supported. Falling back to " "infinite wait\n", __FILE__, __LINE__); if (timeout_ns) { mos_gem_bo_wait_rendering(bo); return 0; } else { return mos_gem_bo_busy(bo) ? -ETIME : 0; } } memclear(wait); wait.bo_handle = bo_gem->gem_handle; wait.timeout_ns = timeout_ns; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_WAIT, &wait); if (ret == -1) return -errno; return ret; } /** * Sets the object to the GTT read and possibly write domain, used by the X * 2D driver in the absence of kernel support to do drm_intel_gem_bo_map_gtt(). * * In combination with drm_intel_gem_bo_pin() and manual fence management, we * can do tiled pixmaps this way. */ static void mos_gem_bo_start_gtt_access(struct mos_linux_bo *bo, int write_enable) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_set_domain set_domain; int ret; memclear(set_domain); set_domain.handle = bo_gem->gem_handle; set_domain.read_domains = I915_GEM_DOMAIN_GTT; set_domain.write_domain = write_enable ? I915_GEM_DOMAIN_GTT : 0; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_SET_DOMAIN, &set_domain); if (ret != 0) { MOS_DBG("%s:%d: Error setting memory domains %d (%08x %08x): %s .\n", __FILE__, __LINE__, bo_gem->gem_handle, set_domain.read_domains, set_domain.write_domain, strerror(errno)); } } static void mos_bufmgr_gem_destroy(struct mos_bufmgr *bufmgr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; struct drm_gem_close close_bo; int i, ret; free(bufmgr_gem->exec2_objects); free(bufmgr_gem->exec_objects); free(bufmgr_gem->exec_bos); pthread_mutex_destroy(&bufmgr_gem->lock); /* Free any cached buffer objects we were going to reuse */ for (i = 0; i < bufmgr_gem->num_buckets; i++) { struct mos_gem_bo_bucket *bucket = &bufmgr_gem->cache_bucket[i]; struct mos_bo_gem *bo_gem; while (!DRMLISTEMPTY(&bucket->head)) { bo_gem = DRMLISTENTRY(struct mos_bo_gem, bucket->head.next, head); DRMLISTDEL(&bo_gem->head); mos_gem_bo_free(&bo_gem->bo); } } /* Release userptr bo kept hanging around for optimisation. */ if (bufmgr_gem->userptr_active.ptr) { memclear(close_bo); close_bo.handle = bufmgr_gem->userptr_active.handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_GEM_CLOSE, &close_bo); free(bufmgr_gem->userptr_active.ptr); if (ret) fprintf(stderr, "Failed to release test userptr object! (%d) " "i915 kernel driver may not be sane!\n", errno); } mos_vma_heap_finish(&bufmgr_gem->vma_heap[MEMZONE_SYS]); mos_vma_heap_finish(&bufmgr_gem->vma_heap[MEMZONE_DEVICE]); free(bufmgr); } static int do_bo_emit_reloc(struct mos_linux_bo *bo, uint32_t offset, struct mos_linux_bo *target_bo, uint32_t target_offset, uint32_t read_domains, uint32_t write_domain, bool need_fence, uint64_t presumed_offset) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct mos_bo_gem *target_bo_gem = (struct mos_bo_gem *) target_bo; if (bo_gem->has_error) return -ENOMEM; if (target_bo_gem->has_error) { bo_gem->has_error = true; return -ENOMEM; } /* Create a new relocation list if needed */ if (bo_gem->relocs == nullptr && mos_setup_reloc_list(bo)) return -ENOMEM; /* Check overflow */ assert(bo_gem->reloc_count < bufmgr_gem->max_relocs); /* Check args */ assert(offset <= bo->size - 4); assert((write_domain & (write_domain - 1)) == 0); /* An object needing a fence is a tiled buffer, so it won't have * relocs to other buffers. */ if (need_fence) { assert(target_bo_gem->reloc_count == 0); target_bo_gem->reloc_tree_fences = 1; } /* Make sure that we're not adding a reloc to something whose size has * already been accounted for. */ assert(!bo_gem->used_as_reloc_target); if (target_bo_gem != bo_gem) { target_bo_gem->used_as_reloc_target = true; bo_gem->reloc_tree_size += target_bo_gem->reloc_tree_size; bo_gem->reloc_tree_fences += target_bo_gem->reloc_tree_fences; } int flags = 0; if (target_bo_gem->pad_to_size) flags |= EXEC_OBJECT_PAD_TO_SIZE; if (target_bo_gem->use_48b_address_range) flags |= EXEC_OBJECT_SUPPORTS_48B_ADDRESS; if (target_bo_gem->exec_async) flags |= EXEC_OBJECT_ASYNC; if (target_bo != bo) mos_gem_bo_reference(target_bo); bo_gem->reloc_target_info[bo_gem->reloc_count].bo = target_bo; bo_gem->reloc_target_info[bo_gem->reloc_count].flags = flags; bo_gem->relocs[bo_gem->reloc_count].offset = offset; bo_gem->relocs[bo_gem->reloc_count].delta = target_offset; bo_gem->relocs[bo_gem->reloc_count].target_handle = target_bo_gem->gem_handle; bo_gem->relocs[bo_gem->reloc_count].read_domains = read_domains; bo_gem->relocs[bo_gem->reloc_count].write_domain = write_domain; bo_gem->relocs[bo_gem->reloc_count].presumed_offset = presumed_offset; bo_gem->reloc_count++; return 0; } static void mos_gem_bo_use_48b_address_range(struct mos_linux_bo *bo, uint32_t enable) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; bo_gem->use_48b_address_range = enable; } static void mos_gem_bo_set_object_async(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *)bo; bo_gem->exec_async = true; } static void mos_gem_bo_set_exec_object_async(struct mos_linux_bo *bo, struct mos_linux_bo *target_bo) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct mos_bo_gem *target_bo_gem = (struct mos_bo_gem *) target_bo; int i; for (i = 0; i < bo_gem->softpin_target_count; i++) { if (bo_gem->softpin_target[i].bo == target_bo) { bo_gem->softpin_target[i].flags |= EXEC_OBJECT_ASYNC; break; } } } static int mos_gem_bo_add_softpin_target(struct mos_linux_bo *bo, struct mos_linux_bo *target_bo, bool write_flag) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct mos_bo_gem *target_bo_gem = (struct mos_bo_gem *) target_bo; if (bo_gem->has_error) return -ENOMEM; if (target_bo_gem->has_error) { bo_gem->has_error = true; return -ENOMEM; } if (!target_bo_gem->is_softpin) return -EINVAL; if (target_bo_gem == bo_gem) return -EINVAL; if (bo_gem->softpin_target_count == bo_gem->softpin_target_size) { int new_size = bo_gem->softpin_target_size * 2; if (new_size == 0) new_size = bufmgr_gem->max_relocs; bo_gem->softpin_target = (struct mos_softpin_target *)realloc(bo_gem->softpin_target, new_size * sizeof(struct mos_softpin_target)); if (!bo_gem->softpin_target) return -ENOMEM; bo_gem->softpin_target_size = new_size; } int flags = EXEC_OBJECT_PINNED; if (target_bo_gem->pad_to_size) flags |= EXEC_OBJECT_PAD_TO_SIZE; if (target_bo_gem->use_48b_address_range) flags |= EXEC_OBJECT_SUPPORTS_48B_ADDRESS; if (target_bo_gem->exec_async) flags |= EXEC_OBJECT_ASYNC; if (write_flag) flags |= EXEC_OBJECT_WRITE; bo_gem->softpin_target[bo_gem->softpin_target_count].bo = target_bo; bo_gem->softpin_target[bo_gem->softpin_target_count].flags = flags; mos_gem_bo_reference(target_bo); bo_gem->softpin_target_count++; return 0; } static int mos_gem_bo_pad_to_size(struct mos_linux_bo *bo, uint64_t pad_to_size) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; if (pad_to_size && pad_to_size < bo->size) return -EINVAL; bo_gem->pad_to_size = pad_to_size; return 0; } static int mos_gem_bo_emit_reloc(struct mos_linux_bo *bo, uint32_t offset, struct mos_linux_bo *target_bo, uint32_t target_offset, uint32_t read_domains, uint32_t write_domain, uint64_t presumed_offset) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bo->bufmgr; return do_bo_emit_reloc(bo, offset, target_bo, target_offset, read_domains, write_domain, false, presumed_offset); } /** * Removes existing relocation entries in the BO after "start". * * This allows a user to avoid a two-step process for state setup with * counting up all the buffer objects and doing a * drm_intel_bufmgr_check_aperture_space() before emitting any of the * relocations for the state setup. Instead, save the state of the * batchbuffer including drm_intel_gem_get_reloc_count(), emit all the * state, and then check if it still fits in the aperture. * * Any further drm_intel_bufmgr_check_aperture_space() queries * involving this buffer in the tree are undefined after this call. * * This also removes all softpinned targets being referenced by the BO. */ static void mos_gem_bo_clear_relocs(struct mos_linux_bo *bo, int start) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int i; struct timespec time; if(GetDrmMode()) return; //libdrm_mock clock_gettime(CLOCK_MONOTONIC, &time); assert(bo_gem->reloc_count >= start); /* Unreference the cleared target buffers */ pthread_mutex_lock(&bufmgr_gem->lock); for (i = start; i < bo_gem->reloc_count; i++) { struct mos_bo_gem *target_bo_gem = (struct mos_bo_gem *) bo_gem->reloc_target_info[i].bo; if (&target_bo_gem->bo != bo) { bo_gem->reloc_tree_fences -= target_bo_gem->reloc_tree_fences; mos_gem_bo_unreference_locked_timed(&target_bo_gem->bo, time.tv_sec); } } bo_gem->reloc_count = start; for (i = 0; i < bo_gem->softpin_target_count; i++) { struct mos_bo_gem *target_bo_gem = (struct mos_bo_gem *) bo_gem->softpin_target[i].bo; mos_gem_bo_unreference_locked_timed(&target_bo_gem->bo, time.tv_sec); } bo_gem->softpin_target_count = 0; pthread_mutex_unlock(&bufmgr_gem->lock); } /** * Walk the tree of relocations rooted at BO and accumulate the list of * validations to be performed and update the relocation buffers with * index values into the validation list. */ static void mos_gem_bo_process_reloc(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int i; if (bo_gem->relocs == nullptr) return; for (i = 0; i < bo_gem->reloc_count; i++) { struct mos_linux_bo *target_bo = bo_gem->reloc_target_info[i].bo; if (target_bo == bo) continue; mos_gem_bo_mark_mmaps_incoherent(bo); /* Continue walking the tree depth-first. */ mos_gem_bo_process_reloc(target_bo); /* Add the target to the validate list */ mos_add_validate_buffer(target_bo); } } static void mos_gem_bo_process_reloc2(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *)bo; int i; if (bo_gem->relocs == nullptr && bo_gem->softpin_target == nullptr) return; for (i = 0; i < bo_gem->reloc_count; i++) { struct mos_linux_bo *target_bo = bo_gem->reloc_target_info[i].bo; int need_fence; if (target_bo == bo) continue; mos_gem_bo_mark_mmaps_incoherent(bo); /* Continue walking the tree depth-first. */ mos_gem_bo_process_reloc2(target_bo); need_fence = (bo_gem->reloc_target_info[i].flags & DRM_INTEL_RELOC_FENCE); /* Add the target to the validate list */ mos_add_validate_buffer2(target_bo, need_fence); } for (i = 0; i < bo_gem->softpin_target_count; i++) { struct mos_linux_bo *target_bo = bo_gem->softpin_target[i].bo; if (target_bo == bo) continue; mos_gem_bo_mark_mmaps_incoherent(bo); mos_gem_bo_process_reloc2(target_bo); mos_add_validate_buffer2(target_bo, false); } } static void mos_update_buffer_offsets(struct mos_bufmgr_gem *bufmgr_gem) { int i; for (i = 0; i < bufmgr_gem->exec_count; i++) { struct mos_linux_bo *bo = bufmgr_gem->exec_bos[i]; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; /* Update the buffer offset */ if (bufmgr_gem->exec_objects[i].offset != bo->offset64) { MOS_DBG("BO %d (%s) migrated: 0x%08x %08x -> 0x%08x %08x\n", bo_gem->gem_handle, bo_gem->name, upper_32_bits(bo->offset64), lower_32_bits(bo->offset64), upper_32_bits(bufmgr_gem->exec_objects[i].offset), lower_32_bits(bufmgr_gem->exec_objects[i].offset)); bo->offset64 = bufmgr_gem->exec_objects[i].offset; bo->offset = bufmgr_gem->exec_objects[i].offset; } } } static void mos_update_buffer_offsets2 (struct mos_bufmgr_gem *bufmgr_gem, mos_linux_context *ctx, mos_linux_bo *cmd_bo) { int i; for (i = 0; i < bufmgr_gem->exec_count; i++) { struct mos_linux_bo *bo = bufmgr_gem->exec_bos[i]; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *)bo; /* Update the buffer offset */ if (bufmgr_gem->exec2_objects[i].offset != bo->offset64) { /* If we're seeing softpinned object here it means that the kernel * has relocated our object... Indicating a programming error */ assert(!bo_gem->is_softpin); MOS_DBG("BO %d (%s) migrated: 0x%08x %08x -> 0x%08x %08x\n", bo_gem->gem_handle, bo_gem->name, upper_32_bits(bo->offset64), lower_32_bits(bo->offset64), upper_32_bits(bufmgr_gem->exec2_objects[i].offset), lower_32_bits(bufmgr_gem->exec2_objects[i].offset)); bo->offset64 = bufmgr_gem->exec2_objects[i].offset; bo->offset = bufmgr_gem->exec2_objects[i].offset; } #if 0//ndef ANDROID if (cmd_bo != bo) { auto item_ctx = ctx->pOsContext->contextOffsetList.begin(); for (; item_ctx != ctx->pOsContext->contextOffsetList.end(); item_ctx++) { if (item_ctx->intel_context == ctx && item_ctx->target_bo == bo) { item_ctx->offset64 = bo->offset64; break; } } if ( item_ctx == ctx->pOsContext->contextOffsetList.end()) { struct MOS_CONTEXT_OFFSET newContext = {ctx, bo, bo->offset64}; ctx->pOsContext->contextOffsetList.push_back(newContext); } } #endif } } drm_export int do_exec2(struct mos_linux_bo *bo, int used, struct mos_linux_context *ctx, drm_clip_rect_t *cliprects, int num_cliprects, int DR4, unsigned int flags, int *fence) { if(GetDrmMode()) return 0; //libdrm_mock struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bo->bufmgr; struct drm_i915_gem_execbuffer2 execbuf; int ret = 0; int i; if (to_bo_gem(bo)->has_error) return -ENOMEM; switch (flags & 0x7) { default: return -EINVAL; case I915_EXEC_BLT: if (!bufmgr_gem->has_blt) return -EINVAL; break; case I915_EXEC_BSD: if (!bufmgr_gem->has_bsd) return -EINVAL; break; case I915_EXEC_VEBOX: if (!bufmgr_gem->has_vebox) return -EINVAL; break; case I915_EXEC_RENDER: case I915_EXEC_DEFAULT: break; } pthread_mutex_lock(&bufmgr_gem->lock); /* Update indices and set up the validate list. */ mos_gem_bo_process_reloc2(bo); /* Add the batch buffer to the validation list. There are no relocations * pointing to it. */ mos_add_validate_buffer2(bo, 0); memclear(execbuf); execbuf.buffers_ptr = (uintptr_t)bufmgr_gem->exec2_objects; execbuf.buffer_count = bufmgr_gem->exec_count; execbuf.batch_start_offset = 0; execbuf.batch_len = used; execbuf.cliprects_ptr = (uintptr_t)cliprects; execbuf.num_cliprects = num_cliprects; execbuf.DR1 = 0; execbuf.DR4 = DR4; execbuf.flags = flags; if (ctx == nullptr) i915_execbuffer2_set_context_id(execbuf, 0); else i915_execbuffer2_set_context_id(execbuf, ctx->ctx_id); execbuf.rsvd2 = 0; if(flags & I915_EXEC_FENCE_SUBMIT) { execbuf.rsvd2 = *fence; } if(flags & I915_EXEC_FENCE_OUT) { execbuf.rsvd2 = -1; } if (bufmgr_gem->no_exec) goto skip_execution; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_EXECBUFFER2_WR, &execbuf); if (ret != 0) { ret = -errno; if (ret == -ENOSPC) { MOS_DBG("Execbuffer fails to pin. " "Estimate: %u. Actual: %u. Available: %u\n", mos_gem_estimate_batch_space(bufmgr_gem->exec_bos, bufmgr_gem->exec_count), mos_gem_compute_batch_space(bufmgr_gem->exec_bos, bufmgr_gem->exec_count), (unsigned int) bufmgr_gem->gtt_size); } } if (ctx != nullptr) { mos_update_buffer_offsets2(bufmgr_gem, ctx, bo); } if(flags & I915_EXEC_FENCE_OUT) { *fence = execbuf.rsvd2 >> 32; } skip_execution: if (bufmgr_gem->bufmgr.debug) mos_gem_dump_validation_list(bufmgr_gem); for (i = 0; i < bufmgr_gem->exec_count; i++) { struct mos_bo_gem *bo_gem = to_bo_gem(bufmgr_gem->exec_bos[i]); bo_gem->idle = false; /* Disconnect the buffer from the validate list */ bo_gem->validate_index = -1; bufmgr_gem->exec_bos[i] = nullptr; } bufmgr_gem->exec_count = 0; pthread_mutex_unlock(&bufmgr_gem->lock); return ret; } static int mos_gem_bo_exec2(struct mos_linux_bo *bo, int used, drm_clip_rect_t *cliprects, int num_cliprects, int DR4) { return do_exec2(bo, used, nullptr, cliprects, num_cliprects, DR4, I915_EXEC_RENDER, nullptr); } static int mos_gem_bo_mrb_exec2(struct mos_linux_bo *bo, int used, drm_clip_rect_t *cliprects, int num_cliprects, int DR4, unsigned int flags) { return do_exec2(bo, used, nullptr, cliprects, num_cliprects, DR4, flags, nullptr); } static int mos_gem_bo_context_exec2(struct mos_linux_bo *bo, int used, struct mos_linux_context *ctx, drm_clip_rect_t *cliprects, int num_cliprects, int DR4, unsigned int flags,int *fence) { return do_exec2(bo, used, ctx, cliprects, num_cliprects, DR4, flags,fence); } static int mos_gem_bo_set_tiling_internal(struct mos_linux_bo *bo, uint32_t tiling_mode, uint32_t stride) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; struct drm_i915_gem_set_tiling set_tiling; int ret; if (bo_gem->global_name == 0 && tiling_mode == bo_gem->tiling_mode && stride == bo_gem->stride) return 0; memset(&set_tiling, 0, sizeof(set_tiling)); do { /* set_tiling is slightly broken and overwrites the * input on the error path, so we have to open code * rmIoctl. */ set_tiling.handle = bo_gem->gem_handle; set_tiling.tiling_mode = tiling_mode; set_tiling.stride = stride; ret = ioctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_SET_TILING, &set_tiling); } while (ret == -1 && (errno == EINTR || errno == EAGAIN)); if (ret == -1) return -errno; bo_gem->tiling_mode = set_tiling.tiling_mode; bo_gem->swizzle_mode = set_tiling.swizzle_mode; bo_gem->stride = set_tiling.stride; return 0; } static int mos_gem_bo_set_tiling(struct mos_linux_bo *bo, uint32_t * tiling_mode, uint32_t stride) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int ret; /* Tiling with userptr surfaces is not supported * on all hardware so refuse it for time being. */ if (bo_gem->is_userptr) return -EINVAL; /* Linear buffers have no stride. By ensuring that we only ever use * stride 0 with linear buffers, we simplify our code. */ if (*tiling_mode == I915_TILING_NONE) stride = 0; ret = mos_gem_bo_set_tiling_internal(bo, *tiling_mode, stride); if (ret == 0) mos_bo_gem_set_in_aperture_size(bufmgr_gem, bo_gem, 0); *tiling_mode = bo_gem->tiling_mode; return ret; } static int mos_gem_bo_get_tiling(struct mos_linux_bo *bo, uint32_t * tiling_mode, uint32_t * swizzle_mode) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; *tiling_mode = bo_gem->tiling_mode; *swizzle_mode = bo_gem->swizzle_mode; return 0; } static int mos_gem_bo_set_softpin_offset(struct mos_linux_bo *bo, uint64_t offset) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; bo_gem->is_softpin = true; bo->offset64 = offset; bo->offset = offset; return 0; } static int mos_gem_bo_set_softpin(MOS_LINUX_BO *bo) { int ret = 0; struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; if (!mos_gem_bo_is_softpin(bo)) { uint64_t offset = mos_gem_bo_vma_alloc(bo->bufmgr, MEMZONE_SYS, bo->size, PAGE_SIZE_64K); ret = mos_gem_bo_set_softpin_offset(bo, offset); } if (ret == 0) { ret = mos_bo_use_48b_address_range(bo, 1); } return ret; } static struct mos_linux_bo * mos_gem_bo_create_from_prime(struct mos_bufmgr *bufmgr, struct mos_drm_bo_alloc_prime *alloc_prime) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; int ret; uint32_t handle; struct mos_bo_gem *bo_gem; struct drm_i915_gem_get_tiling get_tiling; int prime_fd = alloc_prime->prime_fd; int size = alloc_prime->size; drmMMListHead *list; pthread_mutex_lock(&bufmgr_gem->lock); ret = drmPrimeFDToHandle(bufmgr_gem->fd, prime_fd, &handle); if (ret) { MOS_DBG("create_from_prime: failed to obtain handle from fd: %s\n", strerror(errno)); pthread_mutex_unlock(&bufmgr_gem->lock); return nullptr; } /* * See if the kernel has already returned this buffer to us. Just as * for named buffers, we must not create two bo's pointing at the same * kernel object */ for (list = bufmgr_gem->named.next; list != &bufmgr_gem->named; list = list->next) { bo_gem = DRMLISTENTRY(struct mos_bo_gem, list, name_list); if (bo_gem->gem_handle == handle) { mos_gem_bo_reference(&bo_gem->bo); pthread_mutex_unlock(&bufmgr_gem->lock); return &bo_gem->bo; } } bo_gem = (struct mos_bo_gem *)calloc(1, sizeof(*bo_gem)); if (!bo_gem) { pthread_mutex_unlock(&bufmgr_gem->lock); return nullptr; } /* Determine size of bo. The fd-to-handle ioctl really should * return the size, but it doesn't. If we have kernel 3.12 or * later, we can lseek on the prime fd to get the size. Older * kernels will just fail, in which case we fall back to the * provided (estimated or guess size). */ ret = lseek(prime_fd, 0, SEEK_END); if (ret != -1) bo_gem->bo.size = ret; else bo_gem->bo.size = size; bo_gem->bo.handle = handle; bo_gem->bo.bufmgr = bufmgr; bo_gem->gem_handle = handle; atomic_set(&bo_gem->refcount, 1); bo_gem->name = alloc_prime->name; bo_gem->validate_index = -1; bo_gem->reloc_tree_fences = 0; bo_gem->used_as_reloc_target = false; bo_gem->has_error = false; bo_gem->reusable = false; bo_gem->use_48b_address_range = bufmgr_gem->bufmgr.bo_use_48b_address_range ? true : false; DRMLISTADDTAIL(&bo_gem->name_list, &bufmgr_gem->named); pthread_mutex_unlock(&bufmgr_gem->lock); memclear(get_tiling); get_tiling.handle = bo_gem->gem_handle; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_GET_TILING, &get_tiling); if (ret != 0) { MOS_DBG("create_from_prime: failed to get tiling: %s\n", strerror(errno)); mos_gem_bo_unreference(&bo_gem->bo); return nullptr; } bo_gem->tiling_mode = get_tiling.tiling_mode; bo_gem->swizzle_mode = get_tiling.swizzle_mode; /* XXX stride is unknown */ mos_bo_gem_set_in_aperture_size(bufmgr_gem, bo_gem, 0); return &bo_gem->bo; } static int mos_gem_bo_export_to_prime(struct mos_linux_bo *bo, int *prime_fd) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; pthread_mutex_lock(&bufmgr_gem->lock); if (DRMLISTEMPTY(&bo_gem->name_list)) DRMLISTADDTAIL(&bo_gem->name_list, &bufmgr_gem->named); pthread_mutex_unlock(&bufmgr_gem->lock); if (drmPrimeHandleToFD(bufmgr_gem->fd, bo_gem->gem_handle, DRM_CLOEXEC, prime_fd) != 0) return -errno; bo_gem->reusable = false; return 0; } static int mos_gem_bo_flink(struct mos_linux_bo *bo, uint32_t * name) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo->bufmgr; struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int ret; if (!bo_gem->global_name) { struct drm_gem_flink flink; memclear(flink); flink.handle = bo_gem->gem_handle; pthread_mutex_lock(&bufmgr_gem->lock); ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_GEM_FLINK, &flink); if (ret != 0) { pthread_mutex_unlock(&bufmgr_gem->lock); return -errno; } bo_gem->global_name = flink.name; bo_gem->reusable = false; if (DRMLISTEMPTY(&bo_gem->name_list)) DRMLISTADDTAIL(&bo_gem->name_list, &bufmgr_gem->named); pthread_mutex_unlock(&bufmgr_gem->lock); } *name = bo_gem->global_name; return 0; } /** * Enables unlimited caching of buffer objects for reuse. * * This is potentially very memory expensive, as the cache at each bucket * size is only bounded by how many buffers of that size we've managed to have * in flight at once. */ static void mos_gem_enable_reuse(struct mos_bufmgr *bufmgr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bufmgr; bufmgr_gem->bo_reuse = true; } /** * Return the additional aperture space required by the tree of buffer objects * rooted at bo. */ static int mos_gem_bo_get_aperture_space(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int i; int total = 0; if (bo == nullptr || bo_gem->included_in_check_aperture) return 0; total += bo->size; bo_gem->included_in_check_aperture = true; for (i = 0; i < bo_gem->reloc_count; i++) total += mos_gem_bo_get_aperture_space(bo_gem-> reloc_target_info[i].bo); return total; } /** * Count the number of buffers in this list that need a fence reg * * If the count is greater than the number of available regs, we'll have * to ask the caller to resubmit a batch with fewer tiled buffers. * * This function over-counts if the same buffer is used multiple times. */ static unsigned int mos_gem_total_fences(struct mos_linux_bo ** bo_array, int count) { int i; unsigned int total = 0; for (i = 0; i < count; i++) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo_array[i]; if (bo_gem == nullptr) continue; total += bo_gem->reloc_tree_fences; } return total; } /** * Clear the flag set by drm_intel_gem_bo_get_aperture_space() so we're ready * for the next drm_intel_bufmgr_check_aperture_space() call. */ static void mos_gem_bo_clear_aperture_space_flag(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int i; if (bo == nullptr || !bo_gem->included_in_check_aperture) return; bo_gem->included_in_check_aperture = false; for (i = 0; i < bo_gem->reloc_count; i++) mos_gem_bo_clear_aperture_space_flag(bo_gem-> reloc_target_info[i].bo); } /** * Return a conservative estimate for the amount of aperture required * for a collection of buffers. This may double-count some buffers. */ static unsigned int mos_gem_estimate_batch_space(struct mos_linux_bo **bo_array, int count) { int i; unsigned int total = 0; for (i = 0; i < count; i++) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo_array[i]; if (bo_gem != nullptr) total += bo_gem->reloc_tree_size; } return total; } /** * Return the amount of aperture needed for a collection of buffers. * This avoids double counting any buffers, at the cost of looking * at every buffer in the set. */ static unsigned int mos_gem_compute_batch_space(struct mos_linux_bo **bo_array, int count) { int i; unsigned int total = 0; for (i = 0; i < count; i++) { total += mos_gem_bo_get_aperture_space(bo_array[i]); /* For the first buffer object in the array, we get an * accurate count back for its reloc_tree size (since nothing * had been flagged as being counted yet). We can save that * value out as a more conservative reloc_tree_size that * avoids double-counting target buffers. Since the first * buffer happens to usually be the batch buffer in our * callers, this can pull us back from doing the tree * walk on every new batch emit. */ if (i == 0) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo_array[i]; bo_gem->reloc_tree_size = total; } } for (i = 0; i < count; i++) mos_gem_bo_clear_aperture_space_flag(bo_array[i]); return total; } /** * Return -1 if the batchbuffer should be flushed before attempting to * emit rendering referencing the buffers pointed to by bo_array. * * This is required because if we try to emit a batchbuffer with relocations * to a tree of buffers that won't simultaneously fit in the aperture, * the rendering will return an error at a point where the software is not * prepared to recover from it. * * However, we also want to emit the batchbuffer significantly before we reach * the limit, as a series of batchbuffers each of which references buffers * covering almost all of the aperture means that at each emit we end up * waiting to evict a buffer from the last rendering, and we get synchronous * performance. By emitting smaller batchbuffers, we eat some CPU overhead to * get better parallelism. */ static int mos_gem_check_aperture_space(struct mos_linux_bo **bo_array, int count) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *) bo_array[0]->bufmgr; unsigned int total = 0; unsigned int threshold = bufmgr_gem->gtt_size * 3 / 4; int total_fences; /* Check for fence reg constraints if necessary */ if (bufmgr_gem->available_fences) { total_fences = mos_gem_total_fences(bo_array, count); if (total_fences > bufmgr_gem->available_fences) return -ENOSPC; } total = mos_gem_estimate_batch_space(bo_array, count); if (total > threshold) total = mos_gem_compute_batch_space(bo_array, count); if (total > threshold) { MOS_DBG("check_space: overflowed available aperture, " "%dkb vs %dkb\n", total / 1024, (int)bufmgr_gem->gtt_size / 1024); return -ENOSPC; } else { MOS_DBG("drm_check_space: total %dkb vs bufgr %dkb\n", total / 1024, (int)bufmgr_gem->gtt_size / 1024); return 0; } } /* * Disable buffer reuse for objects which are shared with the kernel * as scanout buffers */ static int mos_gem_bo_disable_reuse(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; bo_gem->reusable = false; return 0; } static int mos_gem_bo_is_reusable(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; return bo_gem->reusable; } static int _mos_gem_bo_references(struct mos_linux_bo *bo, struct mos_linux_bo *target_bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; int i; for (i = 0; i < bo_gem->reloc_count; i++) { if (bo_gem->reloc_target_info[i].bo == target_bo) return 1; if (bo == bo_gem->reloc_target_info[i].bo) continue; if (_mos_gem_bo_references(bo_gem->reloc_target_info[i].bo, target_bo)) return 1; } for (i = 0; i< bo_gem->softpin_target_count; i++) { if (bo_gem->softpin_target[i].bo == target_bo) return 1; if (_mos_gem_bo_references(bo_gem->softpin_target[i].bo, target_bo)) return 1; } return 0; } /** Return true if target_bo is referenced by bo's relocation tree. */ static int mos_gem_bo_references(struct mos_linux_bo *bo, struct mos_linux_bo *target_bo) { struct mos_bo_gem *target_bo_gem = (struct mos_bo_gem *) target_bo; if (bo == nullptr || target_bo == nullptr) return 0; if (target_bo_gem->used_as_reloc_target) return _mos_gem_bo_references(bo, target_bo); return 0; } static void add_bucket(struct mos_bufmgr_gem *bufmgr_gem, int size) { unsigned int i = bufmgr_gem->num_buckets; assert(i < ARRAY_SIZE(bufmgr_gem->cache_bucket)); DRMINITLISTHEAD(&bufmgr_gem->cache_bucket[i].head); bufmgr_gem->cache_bucket[i].size = size; bufmgr_gem->num_buckets++; } static void init_cache_buckets(struct mos_bufmgr_gem *bufmgr_gem) { unsigned long size, cache_max_size = 64 * 1024 * 1024; /* OK, so power of two buckets was too wasteful of memory. * Give 3 other sizes between each power of two, to hopefully * cover things accurately enough. (The alternative is * probably to just go for exact matching of sizes, and assume * that for things like composited window resize the tiled * width/height alignment and rounding of sizes to pages will * get us useful cache hit rates anyway) */ add_bucket(bufmgr_gem, 4096); add_bucket(bufmgr_gem, 4096 * 2); add_bucket(bufmgr_gem, 4096 * 3); /* Initialize the linked lists for BO reuse cache. */ for (size = 4 * 4096; size <= cache_max_size; size *= 2) { add_bucket(bufmgr_gem, size); add_bucket(bufmgr_gem, size + size * 1 / 4); add_bucket(bufmgr_gem, size + size * 2 / 4); add_bucket(bufmgr_gem, size + size * 3 / 4); } } /** * Get the PCI ID for the device. This can be overridden by setting the * INTEL_DEVID_OVERRIDE environment variable to the desired ID. */ static int get_pci_device_id(struct mos_bufmgr_gem *bufmgr_gem) { char *devid_override; int devid = 0; int ret; drm_i915_getparam_t gp; if (geteuid() == getuid()) { devid_override = getenv("INTEL_DEVID_OVERRIDE"); if (devid_override) { bufmgr_gem->no_exec = true; return strtod(devid_override, nullptr); } } memclear(gp); gp.param = I915_PARAM_CHIPSET_ID; gp.value = &devid; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); if (ret) { fprintf(stderr, "get chip id failed: %d [%d]\n", ret, errno); fprintf(stderr, "param: %d, val: %d\n", gp.param, *gp.value); } return devid; } static int mos_gem_get_devid(struct mos_bufmgr *bufmgr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; return bufmgr_gem->pci_device; } static struct mos_linux_context * mos_gem_context_create(struct mos_bufmgr *bufmgr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; struct drm_i915_gem_context_create create; struct mos_linux_context *context = nullptr; int ret; context = (struct mos_linux_context *)calloc(1, sizeof(*context)); if (!context) return nullptr; memclear(create); ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE, &create); if (ret != 0) { MOS_DBG("DRM_IOCTL_I915_GEM_CONTEXT_CREATE failed: %s\n", strerror(errno)); free(context); return nullptr; } context->ctx_id = create.ctx_id; context->bufmgr = bufmgr; return context; } static void mos_gem_context_destroy(struct mos_linux_context *ctx) { struct mos_bufmgr_gem *bufmgr_gem; struct drm_i915_gem_context_destroy destroy; int ret; if (ctx == nullptr) return; memclear(destroy); bufmgr_gem = (struct mos_bufmgr_gem *)ctx->bufmgr; destroy.ctx_id = ctx->ctx_id; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_DESTROY, &destroy); if (ret != 0) fprintf(stderr, "DRM_IOCTL_I915_GEM_CONTEXT_DESTROY failed: %s\n", strerror(errno)); free(ctx); } static int mos_bufmg_get_reset_stats(struct mos_linux_context *ctx, uint32_t *reset_count, uint32_t *active, uint32_t *pending) { struct mos_bufmgr_gem *bufmgr_gem; struct drm_i915_reset_stats stats; int ret; if (ctx == nullptr) return -EINVAL; memclear(stats); bufmgr_gem = (struct mos_bufmgr_gem *)ctx->bufmgr; stats.ctx_id = ctx->ctx_id; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GET_RESET_STATS, &stats); if (ret == 0) { if (reset_count != nullptr) *reset_count = stats.reset_count; if (active != nullptr) *active = stats.batch_active; if (pending != nullptr) *pending = stats.batch_pending; } return ret; } static unsigned int mos_bufmgr_hweight8(struct mos_linux_context *ctx, uint8_t w) { uint32_t i, weight = 0; for (i=0; i<8; i++) { weight += !!((w) & (1UL << i)); } return weight; } static uint8_t mos_bufmgr_switch_off_n_bits(struct mos_linux_context *ctx, uint8_t in_mask, int n) { int i,count; uint8_t bi,out_mask; assert (n>0 && n<=8); out_mask = in_mask; count = n; for(i=0; i<8; i++) { bi = 1UL<bufmgr; memset(&context_param, 0, sizeof(context_param)); context_param.ctx_id = ctx->ctx_id; context_param.param = I915_CONTEXT_PARAM_SSEU; context_param.value = (uint64_t) sseu; context_param.size = sizeof(struct drm_i915_gem_context_param_sseu); ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_GETPARAM, &context_param); return ret; } static int mos_bufmgr_set_context_param_sseu(struct mos_linux_context *ctx, struct drm_i915_gem_context_param_sseu sseu) { struct mos_bufmgr_gem *bufmgr_gem; struct drm_i915_gem_context_param context_param; int ret; if (ctx == nullptr) return -EINVAL; bufmgr_gem = (struct mos_bufmgr_gem *)ctx->bufmgr; memset(&context_param, 0, sizeof(context_param)); context_param.ctx_id = ctx->ctx_id; context_param.param = I915_CONTEXT_PARAM_SSEU; context_param.value = (uint64_t) &sseu; context_param.size = sizeof(struct drm_i915_gem_context_param_sseu); ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_SETPARAM, &context_param); return ret; } static int mos_gem_get_context_param(struct mos_linux_context *ctx, uint32_t size, uint64_t param, uint64_t *value) { struct mos_bufmgr_gem *bufmgr_gem; struct drm_i915_gem_context_param context_param; int ret; if (ctx == nullptr) return -EINVAL; bufmgr_gem = (struct mos_bufmgr_gem *)ctx->bufmgr; context_param.ctx_id = ctx->ctx_id; context_param.size = size; context_param.param = param; context_param.value = 0; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_GETPARAM, &context_param); *value = context_param.value; return ret; } static int mos_gem_set_context_param(struct mos_linux_context *ctx, uint32_t size, uint64_t param, uint64_t value) { struct mos_bufmgr_gem *bufmgr_gem; struct drm_i915_gem_context_param context_param; int ret; if (ctx == nullptr) return -EINVAL; bufmgr_gem = (struct mos_bufmgr_gem *)ctx->bufmgr; context_param.ctx_id = ctx->ctx_id; context_param.size = size; context_param.param = param; context_param.value = value; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_SETPARAM, &context_param); return ret; } static int mos_bufmg_reg_read(struct mos_bufmgr *bufmgr, uint32_t offset, uint64_t *result) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; struct drm_i915_reg_read reg_read; int ret; memclear(reg_read); reg_read.offset = offset; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_REG_READ, ®_read); *result = reg_read.val; return ret; } static pthread_mutex_t bufmgr_list_mutex = PTHREAD_MUTEX_INITIALIZER; static drmMMListHead bufmgr_list = { &bufmgr_list, &bufmgr_list }; static struct mos_bufmgr_gem * mos_bufmgr_gem_find(int fd) { struct mos_bufmgr_gem *bufmgr_gem; DRMLISTFOREACHENTRY(bufmgr_gem, &bufmgr_list, managers) { if (bufmgr_gem->fd == fd) { atomic_inc(&bufmgr_gem->refcount); return bufmgr_gem; } } return nullptr; } static void mos_bufmgr_gem_unref(struct mos_bufmgr *bufmgr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; if (atomic_add_unless(&bufmgr_gem->refcount, -1, 1)) { pthread_mutex_lock(&bufmgr_list_mutex); if (atomic_dec_and_test(&bufmgr_gem->refcount)) { DRMLISTDEL(&bufmgr_gem->managers); mos_bufmgr_gem_destroy(bufmgr); } pthread_mutex_unlock(&bufmgr_list_mutex); } } static struct mos_linux_context * mos_gem_context_create_ext(struct mos_bufmgr *bufmgr, __u32 flags, bool bContextProtected) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; struct drm_i915_gem_context_create_ext create; struct mos_linux_context *context = nullptr; struct drm_i915_gem_context_create_ext_setparam p_protected; struct drm_i915_gem_context_create_ext_setparam p_norecover; int ret; context = (struct mos_linux_context *)calloc(1, sizeof(*context)); if (!context) return nullptr; memclear(create); create.flags = flags; create.extensions = 0; if (bContextProtected) { memclear(p_protected); memclear(p_norecover); p_protected.base.next_extension = 0; p_protected.base.name = I915_CONTEXT_CREATE_EXT_SETPARAM; p_protected.param.param = I915_CONTEXT_PARAM_PROTECTED_CONTENT; p_protected.param.value = 1; p_norecover.base.next_extension = (uintptr_t)&p_protected; p_norecover.base.name = I915_CONTEXT_CREATE_EXT_SETPARAM; p_norecover.param.param = I915_CONTEXT_PARAM_RECOVERABLE; p_norecover.param.value = 0; create.flags = flags|I915_CONTEXT_CREATE_FLAGS_USE_EXTENSIONS; create.extensions = (uintptr_t)&p_norecover; } ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE_EXT, &create); if (ret != 0) { MOS_DBG("DRM_IOCTL_I915_GEM_CONTEXT_CREATE failed: %s\n", strerror(errno)); free(context); return nullptr; } context->ctx_id = create.ctx_id; context->bufmgr = bufmgr; return context; } static __u32 mos_gem_vm_create(struct mos_bufmgr *bufmgr) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; struct drm_i915_gem_vm_control *vm = nullptr; __u32 vm_id; int ret; vm = (struct drm_i915_gem_vm_control *)calloc(1, sizeof(struct drm_i915_gem_vm_control)); if (nullptr == vm) { MOS_DBG("vm calloc failed\n" ); return INVALID_VM; } memset(vm, 0, sizeof(*vm)); ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_VM_CREATE, vm); if (ret != 0) { MOS_DBG("DRM_IOCTL_I915_GEM_VM_CREATE failed: %s\n", strerror(errno)); free(vm); return INVALID_VM; } vm_id = vm->vm_id; free(vm); return vm_id; } static void mos_gem_vm_destroy(struct mos_bufmgr *bufmgr, __u32 vm_id) { struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; struct drm_i915_gem_vm_control *vm = nullptr; int ret; if (vm_id == INVALID_VM) { MOS_DBG("input invalid param\n" ); return; } vm = (struct drm_i915_gem_vm_control *)calloc(1, sizeof(struct drm_i915_gem_vm_control)); if (nullptr == vm) { MOS_DBG("vm calloc failed\n" ); return; } vm->vm_id = vm_id; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_VM_DESTROY, vm); if (ret != 0) { MOS_DBG("DRM_IOCTL_I915_GEM_VM_DESTROY failed: %s\n", strerror(errno)); } free(vm); } static struct mos_linux_context * mos_gem_context_create_shared(struct mos_bufmgr *bufmgr, mos_linux_context* ctx, __u32 flags, bool bContextProtected, void *engine_map, uint8_t ctx_width, uint8_t num_placements, uint32_t ctx_type) { MOS_UNUSED(engine_map); MOS_UNUSED(ctx_width); MOS_UNUSED(num_placements); MOS_UNUSED(ctx_type); struct mos_bufmgr_gem *bufmgr_gem = (struct mos_bufmgr_gem *)bufmgr; struct drm_i915_gem_context_create_ext create; struct mos_linux_context *context = nullptr; struct drm_i915_gem_context_create_ext_setparam p_protected; struct drm_i915_gem_context_create_ext_setparam p_norecover; int ret; if (ctx == nullptr || ctx->vm_id == INVALID_VM) return nullptr; context = (struct mos_linux_context *)calloc(1, sizeof(*context)); if (!context) return nullptr; memclear(create); create.flags = flags; create.extensions = 0; if (bContextProtected) { memclear(p_protected); memclear(p_norecover); p_protected.base.next_extension = 0; p_protected.base.name = I915_CONTEXT_CREATE_EXT_SETPARAM; p_protected.param.param = I915_CONTEXT_PARAM_PROTECTED_CONTENT; p_protected.param.value = 1; p_norecover.base.next_extension = (uintptr_t)&p_protected; p_norecover.base.name = I915_CONTEXT_CREATE_EXT_SETPARAM; p_norecover.param.param = I915_CONTEXT_PARAM_RECOVERABLE; p_norecover.param.value = 0; create.flags = flags|I915_CONTEXT_CREATE_FLAGS_USE_EXTENSIONS; create.extensions = (uintptr_t)&p_norecover; } ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE_EXT, &create); if (ret != 0) { MOS_DBG("DRM_IOCTL_I915_GEM_CONTEXT_CREATE failed: %s\n", strerror(errno)); free(context); return nullptr; } context->ctx_id = create.ctx_id; context->bufmgr = bufmgr; ret = mos_set_context_param(context, 0, I915_CONTEXT_PARAM_VM, ctx->vm_id); if(ret != 0) { MOS_DBG("I915_CONTEXT_PARAM_VM failed: %s\n", strerror(errno)); free(context); return nullptr; } return context; } static int mos_bufmgr_query_engines(struct mos_bufmgr *bufmgr, __u16 engine_class, __u64 caps, unsigned int *nengine, void *engine_map) { struct i915_engine_class_instance *ci = (struct i915_engine_class_instance *)engine_map; if((bufmgr == nullptr) || (nengine == nullptr) || (ci == nullptr)) { return -EINVAL; } struct drm_i915_query query; struct drm_i915_query_item query_item; struct drm_i915_query_engine_info *engines = nullptr; int ret, len; int fd = ((struct mos_bufmgr_gem*)bufmgr)->fd; memclear(query_item); query_item.query_id = DRM_I915_QUERY_ENGINE_INFO; query_item.length = 0; memclear(query); query.num_items = 1; query.items_ptr = (uintptr_t)&query_item; ret = drmIoctl(fd, DRM_IOCTL_I915_QUERY, &query); if (ret) { goto fini; } len = query_item.length; engines = (drm_i915_query_engine_info *)malloc(len); if (nullptr == engines) { ret = -ENOMEM; goto fini; } memset(engines,0,len); memclear(query_item); query_item.query_id = DRM_I915_QUERY_ENGINE_INFO; query_item.length = len; query_item.data_ptr = (uintptr_t)engines; memclear(query); query.num_items = 1; query.items_ptr = (uintptr_t)&query_item; ret = drmIoctl(fd, DRM_IOCTL_I915_QUERY, &query); if (ret) { goto fini; } int i, num; for (i = 0, num = 0; i < engines->num_engines; i++) { struct drm_i915_engine_info *engine = (struct drm_i915_engine_info *)&engines->engines[i]; if ( engine_class == engine->engine.engine_class && ((caps & engine->capabilities) == caps )) { ci->engine_class = engine_class; ci->engine_instance = engine->engine.engine_instance; ci++; num++; } if (num > *nengine) { fprintf(stderr,"%s: Number of engine instances out of range, %d,%d\n", __FUNCTION__, num, *nengine); goto fini; } } *nengine = num; fini: if (engines) free(engines); return ret; } static int mos_gem_set_context_param_load_balance(struct mos_linux_context *ctx, struct i915_engine_class_instance *ci, unsigned int count) { int ret; uint32_t size; struct i915_context_engines_load_balance* balancer = nullptr; struct i915_context_param_engines* set_engines = nullptr; assert(ci); /* I915_DEFINE_CONTEXT_ENGINES_LOAD_BALANCE */ size = sizeof(struct i915_context_engines_load_balance) + count * sizeof(*ci); balancer = (struct i915_context_engines_load_balance*)malloc(size); if (NULL == balancer) { ret = -ENOMEM; goto fini; } memset(balancer, 0, size); balancer->base.name = I915_CONTEXT_ENGINES_EXT_LOAD_BALANCE; balancer->num_siblings = count; memcpy(balancer->engines, ci, count * sizeof(*ci)); /* I915_DEFINE_CONTEXT_PARAM_ENGINES */ size = sizeof(uint64_t) + sizeof(*ci); set_engines = (struct i915_context_param_engines*) malloc(size); if (NULL == set_engines) { ret = -ENOMEM; goto fini; } set_engines->extensions = (uintptr_t)(balancer); set_engines->engines[0].engine_class = I915_ENGINE_CLASS_INVALID; set_engines->engines[0].engine_instance = I915_ENGINE_CLASS_INVALID_NONE; ret = mos_set_context_param(ctx, size, I915_CONTEXT_PARAM_ENGINES, (uintptr_t)set_engines); fini: if (set_engines) free(set_engines); if (balancer) free(balancer); return ret; } static bool mos_gem_bo_is_softpin(struct mos_linux_bo *bo) { struct mos_bo_gem *bo_gem = (struct mos_bo_gem *) bo; if (bo_gem == nullptr) { return false; } return bo_gem->is_softpin; } static uint64_t mos_bufmgr_get_platform_information(struct mos_bufmgr *bufmgr) { assert(bufmgr); return bufmgr->platform_information; } static void mos_bufmgr_set_platform_information(struct mos_bufmgr *bufmgr, uint64_t p) { assert(bufmgr); bufmgr->platform_information |= p; } /** * Initializes the GEM buffer manager, which uses the kernel to allocate, map, * and manage map buffer objections. * * \param fd File descriptor of the opened DRM device. */ struct mos_bufmgr * mos_bufmgr_gem_init(int fd, int batch_size, int *device_type) { struct mos_bufmgr_gem *bufmgr_gem; struct drm_i915_gem_get_aperture aperture; drm_i915_getparam_t gp; int ret, tmp; bool exec2 = false; if (device_type != nullptr) { *device_type = DEVICE_TYPE_I915; } pthread_mutex_lock(&bufmgr_list_mutex); bufmgr_gem = mos_bufmgr_gem_find(fd); if (bufmgr_gem) goto exit; bufmgr_gem = (struct mos_bufmgr_gem *)calloc(1, sizeof(*bufmgr_gem)); if (bufmgr_gem == nullptr) goto exit; bufmgr_gem->fd = fd; atomic_set(&bufmgr_gem->refcount, 1); if (pthread_mutex_init(&bufmgr_gem->lock, nullptr) != 0) { free(bufmgr_gem); bufmgr_gem = nullptr; goto exit; } bufmgr_gem->bufmgr.bo_alloc = mos_gem_bo_alloc; bufmgr_gem->bufmgr.bo_alloc_tiled = mos_gem_bo_alloc_tiled; bufmgr_gem->bufmgr.bo_reference = mos_gem_bo_reference; bufmgr_gem->bufmgr.bo_unreference = mos_gem_bo_unreference; bufmgr_gem->bufmgr.bo_map = mos_gem_bo_map; bufmgr_gem->bufmgr.bo_unmap = mos_gem_bo_unmap; bufmgr_gem->bufmgr.bo_wait_rendering = mos_gem_bo_wait_rendering; bufmgr_gem->bufmgr.bo_pad_to_size = mos_gem_bo_pad_to_size; bufmgr_gem->bufmgr.bo_emit_reloc = mos_gem_bo_emit_reloc; bufmgr_gem->bufmgr.bo_get_tiling = mos_gem_bo_get_tiling; bufmgr_gem->bufmgr.bo_set_tiling = mos_gem_bo_set_tiling; bufmgr_gem->bufmgr.bo_flink = mos_gem_bo_flink; bufmgr_gem->bufmgr.bo_exec = mos_gem_bo_exec2; bufmgr_gem->bufmgr.bo_mrb_exec = mos_gem_bo_mrb_exec2; bufmgr_gem->bufmgr.bo_busy = mos_gem_bo_busy; bufmgr_gem->bufmgr.bo_madvise = mos_gem_bo_madvise; bufmgr_gem->bufmgr.destroy = mos_bufmgr_gem_unref; bufmgr_gem->bufmgr.debug = 0; bufmgr_gem->bufmgr.check_aperture_space = mos_gem_check_aperture_space; bufmgr_gem->bufmgr.bo_disable_reuse = mos_gem_bo_disable_reuse; bufmgr_gem->bufmgr.bo_is_reusable = mos_gem_bo_is_reusable; bufmgr_gem->bufmgr.bo_references = mos_gem_bo_references; bufmgr_gem->bufmgr.bo_wait = mos_gem_bo_wait; bufmgr_gem->bufmgr.bo_clear_relocs = mos_gem_bo_clear_relocs; bufmgr_gem->bufmgr.context_create = mos_gem_context_create; bufmgr_gem->bufmgr.context_create_ext = mos_gem_context_create_ext; bufmgr_gem->bufmgr.context_create_shared = mos_gem_context_create_shared; bufmgr_gem->bufmgr.context_destroy = mos_gem_context_destroy; bufmgr_gem->bufmgr.vm_create = mos_gem_vm_create; bufmgr_gem->bufmgr.vm_destroy = mos_gem_vm_destroy; bufmgr_gem->bufmgr.bo_context_exec2 = mos_gem_bo_context_exec2; bufmgr_gem->bufmgr.bo_is_softpin = mos_gem_bo_is_softpin; bufmgr_gem->bufmgr.bo_map_gtt = mos_gem_bo_map_gtt; bufmgr_gem->bufmgr.bo_unmap_gtt = mos_gem_bo_unmap_gtt; bufmgr_gem->bufmgr.bo_map_wc = mos_gem_bo_map_wc; bufmgr_gem->bufmgr.bo_unmap_gtt = mos_gem_bo_unmap_wc; bufmgr_gem->bufmgr.bo_map_unsynchronized = mos_gem_bo_map_unsynchronized; bufmgr_gem->bufmgr.bo_start_gtt_access = mos_gem_bo_start_gtt_access; bufmgr_gem->bufmgr.enable_reuse = mos_gem_enable_reuse; bufmgr_gem->bufmgr.get_devid = mos_gem_get_devid; bufmgr_gem->bufmgr.set_context_param = mos_gem_set_context_param; bufmgr_gem->bufmgr.set_context_param_load_balance = mos_gem_set_context_param_load_balance; bufmgr_gem->bufmgr.get_context_param = mos_gem_get_context_param; bufmgr_gem->bufmgr.bo_create_from_prime = mos_gem_bo_create_from_prime; bufmgr_gem->bufmgr.bo_export_to_prime = mos_gem_bo_export_to_prime; bufmgr_gem->bufmgr.reg_read = mos_bufmg_reg_read; bufmgr_gem->bufmgr.get_reset_stats = mos_bufmg_get_reset_stats; bufmgr_gem->bufmgr.get_context_param_sseu = mos_bufmgr_get_context_param_sseu; bufmgr_gem->bufmgr.set_context_param_sseu = mos_bufmgr_set_context_param_sseu; bufmgr_gem->bufmgr.get_platform_information = mos_bufmgr_get_platform_information; bufmgr_gem->bufmgr.set_platform_information = mos_bufmgr_set_platform_information; bufmgr_gem->bufmgr.switch_off_n_bits = mos_bufmgr_switch_off_n_bits; bufmgr_gem->bufmgr.hweight8 = mos_bufmgr_hweight8; bufmgr_gem->bufmgr.query_engines = mos_bufmgr_query_engines; memclear(aperture); ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GEM_GET_APERTURE, &aperture); if (ret == 0) bufmgr_gem->gtt_size = aperture.aper_available_size; else { fprintf(stderr, "DRM_IOCTL_I915_GEM_APERTURE failed: %s\n", strerror(errno)); bufmgr_gem->gtt_size = 128 * 1024 * 1024; fprintf(stderr, "Assuming %dkB available aperture size.\n" "May lead to reduced performance or incorrect " "rendering.\n", (int)bufmgr_gem->gtt_size / 1024); } bufmgr_gem->pci_device = get_pci_device_id(bufmgr_gem); memclear(gp); gp.value = &tmp; gp.param = I915_PARAM_HAS_EXECBUF2; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); if (!ret) exec2 = true; gp.param = I915_PARAM_HAS_BSD; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_bsd = ret == 0; gp.param = I915_PARAM_HAS_BLT; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_blt = ret == 0; gp.param = I915_PARAM_HAS_RELAXED_FENCING; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_relaxed_fencing = ret == 0; bufmgr_gem->bufmgr.bo_alloc_userptr = check_bo_alloc_userptr; gp.param = I915_PARAM_HAS_WAIT_TIMEOUT; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_wait_timeout = ret == 0; gp.param = I915_PARAM_HAS_LLC; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_llc = *gp.value; gp.param = I915_PARAM_HAS_VEBOX; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_vebox = (ret == 0) & (*gp.value > 0); gp.param = I915_PARAM_MMAP_VERSION; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); bufmgr_gem->has_ext_mmap = (ret == 0) & (*gp.value > 0); gp.param = I915_PARAM_HAS_EXEC_SOFTPIN; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); if (ret == 0 && *gp.value > 0) { bufmgr_gem->bufmgr.bo_set_softpin = mos_gem_bo_set_softpin; bufmgr_gem->bufmgr.bo_add_softpin_target = mos_gem_bo_add_softpin_target; } gp.param = I915_PARAM_HAS_EXEC_ASYNC; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); if (ret == 0 && *gp.value > 0){ bufmgr_gem->bufmgr.set_object_async = mos_gem_bo_set_object_async; bufmgr_gem->bufmgr.set_exec_object_async = mos_gem_bo_set_exec_object_async; } gp.param = I915_PARAM_HAS_ALIASING_PPGTT; ret = drmIoctl(bufmgr_gem->fd, DRM_IOCTL_I915_GETPARAM, &gp); if (ret == 0 && *gp.value == 3) bufmgr_gem->bufmgr.bo_use_48b_address_range = mos_gem_bo_use_48b_address_range; /* Let's go with one relocation per every 2 dwords (but round down a bit * since a power of two will mean an extra page allocation for the reloc * buffer). * * Every 4 was too few for the blender benchmark. */ bufmgr_gem->max_relocs = batch_size / sizeof(uint32_t) / 2 - 2; bufmgr_gem->device_type = DEVICE_TYPE_I915; DRMINITLISTHEAD(&bufmgr_gem->named); init_cache_buckets(bufmgr_gem); DRMLISTADD(&bufmgr_gem->managers, &bufmgr_list); bufmgr_gem->use_softpin = false; mos_vma_heap_init(&bufmgr_gem->vma_heap[MEMZONE_SYS], MEMZONE_SYS_START, MEMZONE_SYS_SIZE); mos_vma_heap_init(&bufmgr_gem->vma_heap[MEMZONE_DEVICE], MEMZONE_DEVICE_START, MEMZONE_DEVICE_SIZE); exit: pthread_mutex_unlock(&bufmgr_list_mutex); return bufmgr_gem != nullptr ? &bufmgr_gem->bufmgr : nullptr; }