github-mirror/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-07-28 18:21:24 +02:00
Code Issues Actions Packages Projects Releases Wiki Activity Graph

drm/xe/svm: Consult madvise preferred location in prefetch

Browse Source 
When prefetch region is DRM_XE_CONSULT_MEM_ADVISE_PREF_LOC, prefetch svm
ranges to preferred location provided by madvise.

v2 (Matthew Brost)
- Fix region, devmem_fd usages
- consult madvise is applicable for other vma's too.

v3
- Fix atomic handling

v4
- Fix xe_svm_range_validate to check for
  DRM_XE_CONSULT_MEM_ADVISE_PREF_LOC too.

Cc: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250821173104.3030148-14-himal.prasad.ghimiray@intel.com
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
This commit is contained in:
Himal Prasad Ghimiray 2025-08-21 23:00:57 +05:30
parent fa1a82c985
commit c1bb69a2e8
2 changed files with 35 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
drivers/gpu/drm/xe/xe_vm.c
View File

@ -38,6 +38,7 @@
#include "xe_res_cursor.h" #include "xe_res_cursor.h"
#include "xe_svm.h" #include "xe_svm.h"
#include "xe_sync.h" #include "xe_sync.h"
#include "xe_tile.h"
#include "xe_trace_bo.h" #include "xe_trace_bo.h"
#include "xe_wa.h" #include "xe_wa.h"
#include "xe_hmm.h" #include "xe_hmm.h"
@ -2400,9 +2401,10 @@ vm_bind_ioctl_ops_create(struct xe_vm *vm, struct xe_vma_ops *vops,
__xe_vm_needs_clear_scratch_pages(vm, flags); __xe_vm_needs_clear_scratch_pages(vm, flags);
} else if (__op->op == DRM_GPUVA_OP_PREFETCH) { } else if (__op->op == DRM_GPUVA_OP_PREFETCH) {
struct xe_vma *vma = gpuva_to_vma(op->base.prefetch.va); struct xe_vma *vma = gpuva_to_vma(op->base.prefetch.va);
struct xe_tile *tile;
struct xe_svm_range *svm_range; struct xe_svm_range *svm_range;
struct drm_gpusvm_ctx ctx = {}; struct drm_gpusvm_ctx ctx = {};
struct xe_tile *tile; struct drm_pagemap *dpagemap;
u8 id, tile_mask = 0; u8 id, tile_mask = 0;
u32 i; u32 i;
@ -2419,8 +2421,24 @@ vm_bind_ioctl_ops_create(struct xe_vm *vm, struct xe_vma_ops *vops,
tile_mask |= 0x1 << id; tile_mask |= 0x1 << id;
xa_init_flags(&op->prefetch_range.range, XA_FLAGS_ALLOC); xa_init_flags(&op->prefetch_range.range, XA_FLAGS_ALLOC);
op->prefetch_range.region = prefetch_region;
op->prefetch_range.ranges_count = 0; op->prefetch_range.ranges_count = 0;
tile = NULL;
if (prefetch_region == DRM_XE_CONSULT_MEM_ADVISE_PREF_LOC) {
dpagemap = xe_vma_resolve_pagemap(vma,
xe_device_get_root_tile(vm->xe));
/*
* TODO: Once multigpu support is enabled will need
* something to dereference tile from dpagemap.
*/
if (dpagemap)
tile = xe_device_get_root_tile(vm->xe);
} else if (prefetch_region) {
tile = &vm->xe->tiles[region_to_mem_type[prefetch_region] -
XE_PL_VRAM0];
}
op->prefetch_range.tile = tile;
alloc_next_range: alloc_next_range:
svm_range = xe_svm_range_find_or_insert(vm, addr, vma, &ctx); svm_range = xe_svm_range_find_or_insert(vm, addr, vma, &ctx);
@ -2439,7 +2457,7 @@ vm_bind_ioctl_ops_create(struct xe_vm *vm, struct xe_vma_ops *vops,
goto unwind_prefetch_ops; goto unwind_prefetch_ops;
} }
if (xe_svm_range_validate(vm, svm_range, tile_mask, !!prefetch_region)) { if (xe_svm_range_validate(vm, svm_range, tile_mask, !!tile)) {
xe_svm_range_debug(svm_range, "PREFETCH - RANGE IS VALID"); xe_svm_range_debug(svm_range, "PREFETCH - RANGE IS VALID");
goto check_next_range; goto check_next_range;
} }
@ -2935,30 +2953,26 @@ static int prefetch_ranges(struct xe_vm *vm, struct xe_vma_op *op)
{ {
bool devmem_possible = IS_DGFX(vm->xe) && IS_ENABLED(CONFIG_DRM_XE_PAGEMAP); bool devmem_possible = IS_DGFX(vm->xe) && IS_ENABLED(CONFIG_DRM_XE_PAGEMAP);
struct xe_vma *vma = gpuva_to_vma(op->base.prefetch.va); struct xe_vma *vma = gpuva_to_vma(op->base.prefetch.va);
struct xe_tile *tile = op->prefetch_range.tile;
int err = 0; int err = 0;
struct xe_svm_range *svm_range; struct xe_svm_range *svm_range;
struct drm_gpusvm_ctx ctx = {}; struct drm_gpusvm_ctx ctx = {};
struct xe_tile *tile;
unsigned long i; unsigned long i;
u32 region;
if (!xe_vma_is_cpu_addr_mirror(vma)) if (!xe_vma_is_cpu_addr_mirror(vma))
return 0; return 0;
region = op->prefetch_range.region;
ctx.read_only = xe_vma_read_only(vma); ctx.read_only = xe_vma_read_only(vma);
ctx.devmem_possible = devmem_possible; ctx.devmem_possible = devmem_possible;
ctx.check_pages_threshold = devmem_possible ? SZ_64K : 0; ctx.check_pages_threshold = devmem_possible ? SZ_64K : 0;
/* TODO: Threading the migration */ /* TODO: Threading the migration */
xa_for_each(&op->prefetch_range.range, i, svm_range) { xa_for_each(&op->prefetch_range.range, i, svm_range) {
if (!region) if (!tile)
xe_svm_range_migrate_to_smem(vm, svm_range); xe_svm_range_migrate_to_smem(vm, svm_range);
if (xe_svm_range_needs_migrate_to_vram(svm_range, vma, region)) { if (xe_svm_range_needs_migrate_to_vram(svm_range, vma, !!tile)) {
tile = &vm->xe->tiles[region_to_mem_type[region] - XE_PL_VRAM0];
err = xe_svm_alloc_vram(tile, svm_range, &ctx); err = xe_svm_alloc_vram(tile, svm_range, &ctx);
if (err) { if (err) {
drm_dbg(&vm->xe->drm, "VRAM allocation failed, retry from userspace, asid=%u, gpusvm=%p, errno=%pe\n", drm_dbg(&vm->xe->drm, "VRAM allocation failed, retry from userspace, asid=%u, gpusvm=%p, errno=%pe\n",
@ -3021,12 +3035,11 @@ static int op_lock_and_prep(struct drm_exec *exec, struct xe_vm *vm,
struct xe_vma *vma = gpuva_to_vma(op->base.prefetch.va); struct xe_vma *vma = gpuva_to_vma(op->base.prefetch.va);
u32 region; u32 region;
if (xe_vma_is_cpu_addr_mirror(vma)) if (!xe_vma_is_cpu_addr_mirror(vma)) {
region = op->prefetch_range.region;
else
region = op->prefetch.region; region = op->prefetch.region;
xe_assert(vm->xe, region == DRM_XE_CONSULT_MEM_ADVISE_PREF_LOC ||
xe_assert(vm->xe, region <= ARRAY_SIZE(region_to_mem_type)); region <= ARRAY_SIZE(region_to_mem_type));
}
err = vma_lock_and_validate(exec, err = vma_lock_and_validate(exec,
gpuva_to_vma(op->base.prefetch.va), gpuva_to_vma(op->base.prefetch.va),
@ -3444,8 +3457,8 @@ static int vm_bind_ioctl_check_args(struct xe_device *xe, struct xe_vm *vm,
op == DRM_XE_VM_BIND_OP_PREFETCH) || op == DRM_XE_VM_BIND_OP_PREFETCH) ||
XE_IOCTL_DBG(xe, prefetch_region && XE_IOCTL_DBG(xe, prefetch_region &&
op != DRM_XE_VM_BIND_OP_PREFETCH) || op != DRM_XE_VM_BIND_OP_PREFETCH) ||
XE_IOCTL_DBG(xe, !(BIT(prefetch_region) & XE_IOCTL_DBG(xe, (prefetch_region != DRM_XE_CONSULT_MEM_ADVISE_PREF_LOC &&
xe->info.mem_region_mask)) || !(BIT(prefetch_region) & xe->info.mem_region_mask))) ||
XE_IOCTL_DBG(xe, obj && XE_IOCTL_DBG(xe, obj &&
op == DRM_XE_VM_BIND_OP_UNMAP)) { op == DRM_XE_VM_BIND_OP_UNMAP)) {
err = -EINVAL; err = -EINVAL;

7
drivers/gpu/drm/xe/xe_vm_types.h
View File

@ -428,8 +428,11 @@ struct xe_vma_op_prefetch_range {
struct xarray range; struct xarray range;
/** @ranges_count: number of svm ranges to map */ /** @ranges_count: number of svm ranges to map */
u32 ranges_count; u32 ranges_count;
/** @region: memory region to prefetch to */ /**
u32 region; * @tile: Pointer to the tile structure containing memory to prefetch.
* NULL if prefetch requested region is smem
*/
struct xe_tile *tile;
}; };
/** enum xe_vma_op_flags - flags for VMA operation */ /** enum xe_vma_op_flags - flags for VMA operation */