drm/amdkfd: AIP mGPUs best prefetch location for xnack on
Subsystems (3)
PATCH 1/1 drm/amdkfd: AIP mGPUs best prefetch location for xnack on
For xnack on, if range ACCESS or ACCESS_IN_PLACE (AIP) by single GPU, or range is ACCESS_IN_PLACE by mGPUs and all mGPUs connection on xgmi same hive, the best prefetch location is prefetch_loc GPU. Otherwise, the best prefetch location is always CPU because GPU can not map vram of other GPUs through small bar PCIe.
--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 44 +++++++++++++++++----------- 1 file changed, 27 insertions(+), 17 deletions(-)
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +27 -17
@@ -2754,22 +2754,29 @@ svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size,return 0 ; } -/* svm_range_best_prefetch_location - decide the best prefetch location +/** +* svm_range_best_prefetch_location - decide the best prefetch location * @prange: svm range structure * * For xnack off: -* If range map to single GPU, the best acutal location is prefetch loc, which +* If range map to single GPU, the best prefetch location is prefetch_loc, which * can be CPU or GPU. * -* If range map to multiple GPUs, only if mGPU connection on xgmi same hive, -* the best actual location could be prefetch_loc GPU. If mGPU connection on -* PCIe, the best actual location is always CPU, because GPU cannot access vram -* of other GPUs, assuming PCIe small bar (large bar support is not upstream). +* If range is ACCESS or ACCESS_IN_PLACE by mGPUs, only if mGPU connection on +* xgmi same hive, the best prefetch location is prefetch_loc GPU. If mGPUs +* connection on PCIe, the best prefetch location is always CPU, because GPU +* cannot map vram of other GPUs, assuming PCIe small bar (large bar support is +* not upstream). * * For xnack on: -* The best actual location is prefetch location. If mGPU connection on xgmi -* same hive, range map to multiple GPUs. Otherwise, the range only map to -* actual location GPU. Other GPU access vm fault will trigger migration. +* If range map to single GPU or is not ACCESS_IN_PLACE by mGPUs, the best +* prefetch location is prefetch_loc, other GPU access vm fault will trigger +* migration. +* +* If range is ACCESS_IN_PLACE by mGPUs, only if mGPU connection on xgmi same +* hive, the best prefetch location is prefetch_loc GPU. If mGPUs connection on +* PCIe, the best prefetch location is always CPU, because GPU cannot map vram +* of other GPUs. * * Context: Process context *
@@ -2787,14 +2794,13 @@ svm_range_best_prefetch_location(struct svm_range *prange)struct kfd_process *p;uint32_t gpuidx; - p = container_of(prange->svms, struct kfd_process, svms); - -/* xnack on */ -if (p->xnack_enabled) +if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) goto out; -/* xnack off */ -if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) + p = container_of(prange->svms, struct kfd_process, svms); + +if (p->xnack_enabled && +bitmap_weight (prange->bitmap_aip,MAX_GPU_INSTANCE ) <=1 ) goto out; bo_adev =svm_range_get_adev_by_id (prange, best_loc);
@@ -2803,8 +2809,12 @@ svm_range_best_prefetch_location(struct svm_range *prange) best_loc =0 ; goto out; } -bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, -MAX_GPU_INSTANCE ); + +if (p->xnack_enabled) +bitmap_copy (bitmap, prange->bitmap_aip,MAX_GPU_INSTANCE ); +else +bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, +MAX_GPU_INSTANCE );for_each_set_bit (gpuidx, bitmap,MAX_GPU_INSTANCE ) { pdd =kfd_process_device_from_gpuidx (p, gpuidx);
Signature · 1 line
2.17.1
Felix Kuehling
maintainerAm 2021-08-02 um 12:28 p.m. schrieb Philip Yang:
Quoted 61 lines
For xnack on, if range ACCESS or
ACCESS_IN_PLACE(AIP) by single GPU, or range isACCESS_IN_PLACEby mGPUs and all mGPUs connection on xgmi same hive, the best prefetch location isprefetch_locGPU. Otherwise, the best prefetch location is always CPU because GPU can not map vram of other GPUs through small bar PCIe.Signed-off-by Philip Yang <[email protected]>--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 44 +++++++++++++++++----------- 1 file changed, 27 insertions(+), 17 deletions(-)drivers/gpu/drm/amd/amdkfd/kfd_svm.c +16
-12@@ -2754,22 +2754,29 @@ svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size,return 0 ; } -/* svm_range_best_prefetch_location - decide the best prefetch location +/** +* svm_range_best_prefetch_location - decide the best prefetch location * @prange: svm range structure * * For xnack off: -* If range map to single GPU, the best acutal location is prefetch loc, which +* If range map to single GPU, the best prefetch location is prefetch_loc, which * can be CPU or GPU. * -* If range map to multiple GPUs, only if mGPU connection on xgmi same hive, -* the best actual location could be prefetch_loc GPU. If mGPU connection on -* PCIe, the best actual location is always CPU, because GPU cannot access vram -* of other GPUs, assuming PCIe small bar (large bar support is not upstream). +* If range is ACCESS or ACCESS_IN_PLACE by mGPUs, only if mGPU connection on +* xgmi same hive, the best prefetch location is prefetch_loc GPU. If mGPUs +* connection on PCIe, the best prefetch location is always CPU, because GPU +* cannot map vram of other GPUs, assuming PCIe small bar (large bar support is +* not upstream). * * For xnack on: -* The best actual location is prefetch location. If mGPU connection on xgmi -* same hive, range map to multiple GPUs. Otherwise, the range only map to -* actual location GPU. Other GPU access vm fault will trigger migration. +* If range map to single GPU or is not ACCESS_IN_PLACE by mGPUs, the best +* prefetch location is prefetch_loc, other GPU access vm fault will trigger +* migration. +* +* If range is ACCESS_IN_PLACE by mGPUs, only if mGPU connection on xgmi same +* hive, the best prefetch location is prefetch_loc GPU. If mGPUs connection on +* PCIe, the best prefetch location is always CPU, because GPU cannot map vram +* of other GPUs. * * Context: Process context * @@ -2787,14 +2794,13 @@ svm_range_best_prefetch_location(struct svm_range *prange)struct kfd_process *p;uint32_t gpuidx; - p = container_of(prange->svms, struct kfd_process, svms); - -/* xnack on */
drivers/gpu/drm/amd/amdkfd/kfd_svm.c svm_range_best_prefetch_location(struct svm_range *prange) · line 2803
Quoted 10 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +5
-3-if (p->xnack_enabled) +if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) goto out; -/* xnack off */ -if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) + p = container_of(prange->svms, struct kfd_process, svms); + +if (p->xnack_enabled && +bitmap_weight (prange->bitmap_aip,MAX_GPU_INSTANCE ) <=1 )
I'm not sure why this condition is needed. I think there is an implicit assumption that the prefetch_location is the one GPU that's set in the bitmap_aip. But is that really a safe assumption? If you're prefetching to a different GPU than the one that needs to access the memory in-place, then the access is still going to trigger a page fault, so it's not "in-place".
I think you can just drop this condition. It's a questionable short-cut for something that's handled more correctly in the loop below.
Regards,
Felix
Quoted 18 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +6
-2goto out; bo_adev =svm_range_get_adev_by_id (prange, best_loc);@@ -2803,8 +2809,12 @@ svm_range_best_prefetch_location(struct svm_range *prange) best_loc =0 ; goto out; } -bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, -MAX_GPU_INSTANCE ); + +if (p->xnack_enabled) +bitmap_copy (bitmap, prange->bitmap_aip,MAX_GPU_INSTANCE ); +else +bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, +MAX_GPU_INSTANCE );for_each_set_bit (gpuidx, bitmap,MAX_GPU_INSTANCE ) { pdd =kfd_process_device_from_gpuidx (p, gpuidx);
PATCH v2 1/1 drm/amdkfd: AIP mGPUs best prefetch location for xnack on
In reply to Philip YangFor xnack on, if range ACCESS or ACCESS_IN_PLACE (AIP) by single GPU, or range is ACCESS_IN_PLACE by mGPUs and all mGPUs connection on xgmi same hive, the best prefetch location is prefetch_loc GPU. Otherwise, the best prefetch location is always CPU because GPU can not map vram of other GPUs through small bar PCIe.
--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 35 +++++++++++++++------------- 1 file changed, 19 insertions(+), 16 deletions(-)
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +19 -16
@@ -2719,22 +2719,26 @@ svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size,return 0 ; } -/* svm_range_best_prefetch_location - decide the best prefetch location +/** +* svm_range_best_prefetch_location - decide the best prefetch location * @prange: svm range structure * * For xnack off: -* If range map to single GPU, the best acutal location is prefetch loc, which +* If range map to single GPU, the best prefetch location is prefetch_loc, which * can be CPU or GPU. * -* If range map to multiple GPUs, only if mGPU connection on xgmi same hive, -* the best actual location could be prefetch_loc GPU. If mGPU connection on -* PCIe, the best actual location is always CPU, because GPU cannot access vram -* of other GPUs, assuming PCIe small bar (large bar support is not upstream). +* If range is ACCESS or ACCESS_IN_PLACE by mGPUs, only if mGPU connection on +* XGMI same hive, the best prefetch location is prefetch_loc GPU, othervise +* the best prefetch location is always CPU, because GPU can not map vram of +* other GPUs, assuming PCIe small bar (large bar support is not upstream). * * For xnack on: -* The best actual location is prefetch location. If mGPU connection on xgmi -* same hive, range map to multiple GPUs. Otherwise, the range only map to -* actual location GPU. Other GPU access vm fault will trigger migration. +* If range is not ACCESS_IN_PLACE by mGPUs, the best prefetch location is +* prefetch_loc, other GPU access will generate vm fault and trigger migration. +* +* If range is ACCESS_IN_PLACE by mGPUs, only if mGPU connection on XGMI same +* hive, the best prefetch location is prefetch_loc GPU, otherwise the best +* prefetch location is always CPU, because GPU cannot map vram of other GPUs. * * Context: Process context *
@@ -2754,11 +2758,6 @@ svm_range_best_prefetch_location(struct svm_range *prange) p = container_of(prange->svms, struct kfd_process, svms); -/* xnack on */ -if (p->xnack_enabled) - goto out; - -/* xnack off */ if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) goto out;
@@ -2768,8 +2767,12 @@ svm_range_best_prefetch_location(struct svm_range *prange) best_loc =0 ; goto out; } -bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, -MAX_GPU_INSTANCE ); + +if (p->xnack_enabled) +bitmap_copy (bitmap, prange->bitmap_aip,MAX_GPU_INSTANCE ); +else +bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, +MAX_GPU_INSTANCE );for_each_set_bit (gpuidx, bitmap,MAX_GPU_INSTANCE ) { pdd =kfd_process_device_from_gpuidx (p, gpuidx);
Signature · 1 line
2.17.1
Am 2021-08-09 um 6:21 p.m. schrieb Philip Yang:
Quoted 5 lines
For xnack on, if range ACCESS or
ACCESS_IN_PLACE(AIP) by single GPU, or range isACCESS_IN_PLACEby mGPUs and all mGPUs connection on xgmi same hive, the best prefetch location isprefetch_locGPU. Otherwise, the best prefetch location is always CPU because GPU can not map vram of other GPUs through small bar PCIe.
I don't think small-bar is really a factor here. Even with large-BAR, our P2P mappings are not coherent like XGMI mappings are. So we wouldn't be able to use P2P even on large-BAR systems. So I would modify this sentence:
Quoted 27 lines
Otherwise, the best
prefetch location is always CPU because GPU can not coherently map vram of other GPUs through PCIe.Signed-off-by Philip Yang <[email protected]>--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 35 +++++++++++++++------------- 1 file changed, 19 insertions(+), 16 deletions(-)drivers/gpu/drm/amd/amdkfd/kfd_svm.c +3
-2@@ -2719,22 +2719,26 @@ svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size,return 0 ; } -/* svm_range_best_prefetch_location - decide the best prefetch location +/** +* svm_range_best_prefetch_location - decide the best prefetch location * @prange: svm range structure * * For xnack off: -* If range map to single GPU, the best acutal location is prefetch loc, which +* If range map to single GPU, the best prefetch location is prefetch_loc, which
drivers/gpu/drm/amd/amdkfd/kfd_svm.c svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size, · line 2733
Quoted 10 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +4
-4* can be CPU or GPU. * -* If range map to multiple GPUs, only if mGPU connection on xgmi same hive, -* the best actual location could be prefetch_loc GPU. If mGPU connection on -* PCIe, the best actual location is always CPU, because GPU cannot access vram -* of other GPUs, assuming PCIe small bar (large bar support is not upstream). +* If range is ACCESS or ACCESS_IN_PLACE by mGPUs, only if mGPU connection on +* XGMI same hive, the best prefetch location is prefetch_loc GPU, othervise +* the best prefetch location is always CPU, because GPU can not map vram of +* other GPUs, assuming PCIe small bar (large bar support is not upstream).
Same as above. With that fixed, the patch is
Quoted 41 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +12
-10* * For xnack on: -* The best actual location is prefetch location. If mGPU connection on xgmi -* same hive, range map to multiple GPUs. Otherwise, the range only map to -* actual location GPU. Other GPU access vm fault will trigger migration. +* If range is not ACCESS_IN_PLACE by mGPUs, the best prefetch location is +* prefetch_loc, other GPU access will generate vm fault and trigger migration. +* +* If range is ACCESS_IN_PLACE by mGPUs, only if mGPU connection on XGMI same +* hive, the best prefetch location is prefetch_loc GPU, otherwise the best +* prefetch location is always CPU, because GPU cannot map vram of other GPUs. * * Context: Process context * @@ -2754,11 +2758,6 @@ svm_range_best_prefetch_location(struct svm_range *prange) p = container_of(prange->svms, struct kfd_process, svms); -/* xnack on */ -if (p->xnack_enabled) - goto out; - -/* xnack off */ if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) goto out;@@ -2768,8 +2767,12 @@ svm_range_best_prefetch_location(struct svm_range *prange) best_loc =0 ; goto out; } -bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, -MAX_GPU_INSTANCE ); + +if (p->xnack_enabled) +bitmap_copy (bitmap, prange->bitmap_aip,MAX_GPU_INSTANCE ); +else +bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, +MAX_GPU_INSTANCE );for_each_set_bit (gpuidx, bitmap,MAX_GPU_INSTANCE ) { pdd =kfd_process_device_from_gpuidx (p, gpuidx);
PATCH v3 1/1 drm/amdkfd: AIP mGPUs best prefetch location for xnack on
In reply to Philip YangFor xnack on, if range ACCESS or ACCESS_IN_PLACE (AIP) by single GPU, or range is ACCESS_IN_PLACE by mGPUs and all mGPUs connection on XGMI same hive, the best prefetch location is prefetch_loc GPU. Otherwise, the best prefetch location is always CPU because GPU does not have coherent mapping VRAM of other GPUs even with large-BAR PCIe connection.
--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 35 +++++++++++++++------------- 1 file changed, 19 insertions(+), 16 deletions(-)
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +19 -16
@@ -2719,22 +2719,26 @@ svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size,return 0 ; } -/* svm_range_best_prefetch_location - decide the best prefetch location +/** +* svm_range_best_prefetch_location - decide the best prefetch location * @prange: svm range structure * * For xnack off: -* If range map to single GPU, the best acutal location is prefetch loc, which +* If range map to single GPU, the best prefetch location is prefetch_loc, which * can be CPU or GPU. * -* If range map to multiple GPUs, only if mGPU connection on xgmi same hive, -* the best actual location could be prefetch_loc GPU. If mGPU connection on -* PCIe, the best actual location is always CPU, because GPU cannot access vram -* of other GPUs, assuming PCIe small bar (large bar support is not upstream). +* If range is ACCESS or ACCESS_IN_PLACE by mGPUs, only if mGPU connection on +* XGMI same hive, the best prefetch location is prefetch_loc GPU, othervise +* the best prefetch location is always CPU, because GPU can not have coherent +* mapping VRAM of other GPUs even with large-BAR PCIe connection. * * For xnack on: -* The best actual location is prefetch location. If mGPU connection on xgmi -* same hive, range map to multiple GPUs. Otherwise, the range only map to -* actual location GPU. Other GPU access vm fault will trigger migration. +* If range is not ACCESS_IN_PLACE by mGPUs, the best prefetch location is +* prefetch_loc, other GPU access will generate vm fault and trigger migration. +* +* If range is ACCESS_IN_PLACE by mGPUs, only if mGPU connection on XGMI same +* hive, the best prefetch location is prefetch_loc GPU, otherwise the best +* prefetch location is always CPU. * * Context: Process context *
@@ -2754,11 +2758,6 @@ svm_range_best_prefetch_location(struct svm_range *prange) p = container_of(prange->svms, struct kfd_process, svms); -/* xnack on */ -if (p->xnack_enabled) - goto out; - -/* xnack off */ if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) goto out;
@@ -2768,8 +2767,12 @@ svm_range_best_prefetch_location(struct svm_range *prange) best_loc =0 ; goto out; } -bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, -MAX_GPU_INSTANCE ); + +if (p->xnack_enabled) +bitmap_copy (bitmap, prange->bitmap_aip,MAX_GPU_INSTANCE ); +else +bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, +MAX_GPU_INSTANCE );for_each_set_bit (gpuidx, bitmap,MAX_GPU_INSTANCE ) { pdd =kfd_process_device_from_gpuidx (p, gpuidx);
Signature · 1 line
2.17.1
Am 2021-08-11 um 10:35 a.m. schrieb Philip Yang:
Quoted 7 lines
For xnack on, if range ACCESS or
ACCESS_IN_PLACE(AIP) by single GPU, or range isACCESS_IN_PLACEby mGPUs and all mGPUs connection on XGMI same hive, the best prefetch location isprefetch_locGPU. Otherwise, the best prefetch location is always CPU because GPU does not have coherent mapping VRAM of other GPUs even with large-BAR PCIe connection.Signed-off-by Philip Yang <[email protected]>
Quoted 71 lines
--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 35 +++++++++++++++------------- 1 file changed, 19 insertions(+), 16 deletions(-)drivers/gpu/drm/amd/amdkfd/kfd_svm.c +19
-16@@ -2719,22 +2719,26 @@ svm_range_add(struct kfd_process *p, uint64_t start, uint64_t size,return 0 ; } -/* svm_range_best_prefetch_location - decide the best prefetch location +/** +* svm_range_best_prefetch_location - decide the best prefetch location * @prange: svm range structure * * For xnack off: -* If range map to single GPU, the best acutal location is prefetch loc, which +* If range map to single GPU, the best prefetch location is prefetch_loc, which * can be CPU or GPU. * -* If range map to multiple GPUs, only if mGPU connection on xgmi same hive, -* the best actual location could be prefetch_loc GPU. If mGPU connection on -* PCIe, the best actual location is always CPU, because GPU cannot access vram -* of other GPUs, assuming PCIe small bar (large bar support is not upstream). +* If range is ACCESS or ACCESS_IN_PLACE by mGPUs, only if mGPU connection on +* XGMI same hive, the best prefetch location is prefetch_loc GPU, othervise +* the best prefetch location is always CPU, because GPU can not have coherent +* mapping VRAM of other GPUs even with large-BAR PCIe connection. * * For xnack on: -* The best actual location is prefetch location. If mGPU connection on xgmi -* same hive, range map to multiple GPUs. Otherwise, the range only map to -* actual location GPU. Other GPU access vm fault will trigger migration. +* If range is not ACCESS_IN_PLACE by mGPUs, the best prefetch location is +* prefetch_loc, other GPU access will generate vm fault and trigger migration. +* +* If range is ACCESS_IN_PLACE by mGPUs, only if mGPU connection on XGMI same +* hive, the best prefetch location is prefetch_loc GPU, otherwise the best +* prefetch location is always CPU. * * Context: Process context * @@ -2754,11 +2758,6 @@ svm_range_best_prefetch_location(struct svm_range *prange) p = container_of(prange->svms, struct kfd_process, svms); -/* xnack on */ -if (p->xnack_enabled) - goto out; - -/* xnack off */ if (!best_loc || best_loc ==KFD_IOCTL_SVM_LOCATION_UNDEFINED ) goto out;@@ -2768,8 +2767,12 @@ svm_range_best_prefetch_location(struct svm_range *prange) best_loc =0 ; goto out; } -bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, -MAX_GPU_INSTANCE ); + +if (p->xnack_enabled) +bitmap_copy (bitmap, prange->bitmap_aip,MAX_GPU_INSTANCE ); +else +bitmap_or (bitmap, prange->bitmap_access, prange->bitmap_aip, +MAX_GPU_INSTANCE );for_each_set_bit (gpuidx, bitmap,MAX_GPU_INSTANCE ) { pdd =kfd_process_device_from_gpuidx (p, gpuidx);