drm/amdkfd: Optimize svm range map to GPU with XNACK on
Subsystems (3)
PATCH 1/1 drm/amdkfd: Optimize svm range map to GPU with XNACK on
With XNACK on if svm_range_set_attr set the range access or access_in_place attribute, we don't call svm_range_validate_and_map to update GPU mapping. This avoids prefaulting the range pages on system memory if the range is not prefetch to VRAM and not mapped to GPUs.
--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +3 -1
@@ -734,7 +734,9 @@ svm_range_apply_attrs(struct kfd_process *p, struct svm_range *prange,case KFD_IOCTL_SVM_ATTR_ACCESS :case KFD_IOCTL_SVM_ATTR_ACCESS_IN_PLACE :case KFD_IOCTL_SVM_ATTR_NO_ACCESS : - *update_mapping = true; +if (!p->xnack_enabled) + *update_mapping = true; + gpuidx =kfd_process_gpuidx_from_gpuid (p, attrs[i].value);if (attrs[i].type ==KFD_IOCTL_SVM_ATTR_NO_ACCESS ) {
Signature · 1 line
2.35.1
Felix Kuehling
maintainerQuoted 15 lines · Philip Yang
With XNACK on if
svm_range_set_attrset the range access oraccess_in_placeattribute, we don't callsvm_range_validate_and_mapto update GPU mapping. This avoids prefaulting the range pages on system memory if the range is not prefetch to VRAM and not mapped to GPUs.Signed-off-by Philip Yang <[email protected]>--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)drivers/gpu/drm/amd/amdkfd/kfd_svm.c +0
-0
drivers/gpu/drm/amd/amdkfd/kfd_svm.c svm_range_apply_attrs(struct kfd_process *p, struct svm_range *prange, · line 739
Quoted 7 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +3
-1@@ -734,7 +734,9 @@ svm_range_apply_attrs(struct kfd_process *p, struct svm_range *prange,case KFD_IOCTL_SVM_ATTR_ACCESS :case KFD_IOCTL_SVM_ATTR_ACCESS_IN_PLACE :case KFD_IOCTL_SVM_ATTR_NO_ACCESS : - *update_mapping = true; +if (!p->xnack_enabled) + *update_mapping = true; +
For NO_ACCESS we need to update the mapping to ensure that the PTEs are invalidated. For ACCESS or ACCESS_IN_PLACE we can leave it for a page fault.
Regards,
Felix
Quoted 3 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +0
-0gpuidx =kfd_process_gpuidx_from_gpuid (p, attrs[i].value);if (attrs[i].type ==KFD_IOCTL_SVM_ATTR_NO_ACCESS ) {
Quoted 12 lines · Felix Kuehling
Quoted 10 lines · Philip Yang
With XNACK on if
svm_range_set_attrset the range access oraccess_in_placeattribute, we don't callsvm_range_validate_and_mapto update GPU mapping. This avoids prefaulting the range pages on system memory if the range is not prefetch to VRAM and not mapped to GPUs.Signed-off-by Philip Yang <[email protected]>---
drivers/gpu/drm/amd/amdkfd/kfd_svm.c| 4 +++-1 file changed, 3 insertions(+), 1 deletion(-)
Quoted 15 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +0
-0@@ -734,7 +734,9 @@ svm_range_apply_attrs(struct kfd_process *p, struct svm_range *prange,Quoted 7 lines
case
KFD_IOCTL_SVM_ATTR_ACCESS:
caseKFD_IOCTL_SVM_ATTR_ACCESS_IN_PLACE:
caseKFD_IOCTL_SVM_ATTR_NO_ACCESS:
-*update_mapping= true;
+ if (!p->xnack_enabled)
+*update_mapping= true;
+For
NO_ACCESSwe need to update the mapping to ensure that the PTEs are invalidated. For ACCESS orACCESS_IN_PLACEwe can leave it for a page fault.
For NO_ACCESS, there is todo comment "TODO: unmap ranges from GPU that lost access", this should be handled by another patch. This patch only address that for XNACK ON, we can setup GPU mapping after page fault, without prefaulting and mapping to GPU for ACCESS or ACCESS_IN_PLACE.
Regards,
Philip
Quoted 8 lines
Regards,
FelixQuoted 3 lines
gpuidx =
kfd_process_gpuidx_from_gpuid(p, attrs[i].value); if (attrs[i].type ==KFD_IOCTL_SVM_ATTR_NO_ACCESS) {
Quoted 30 lines · Philip Yang
Quoted 12 lines · Felix Kuehling
Quoted 10 lines · Philip Yang
With XNACK on if
svm_range_set_attrset the range access oraccess_in_placeattribute, we don't callsvm_range_validate_and_mapto update GPU mapping. This avoids prefaulting the range pages on system memory if the range is not prefetch to VRAM and not mapped to GPUs.Signed-off-by Philip Yang <[email protected]>---
drivers/gpu/drm/amd/amdkfd/kfd_svm.c| 4 +++-1 file changed, 3 insertions(+), 1 deletion(-)Quoted 15 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdkfd/kfd_svm.c +0
-0@@ -734,7 +734,9 @@ svm_range_apply_attrs(struct kfd_process *p, struct svm_range *prange,Quoted 7 lines
case
KFD_IOCTL_SVM_ATTR_ACCESS:
caseKFD_IOCTL_SVM_ATTR_ACCESS_IN_PLACE:
caseKFD_IOCTL_SVM_ATTR_NO_ACCESS:
-*update_mapping= true;
+ if (!p->xnack_enabled)
+*update_mapping= true;
+For
NO_ACCESSwe need to update the mapping to ensure that the PTEs are invalidated. For ACCESS orACCESS_IN_PLACEwe can leave it for a page fault.
Quoted 1 line
For
NO_ACCESS, there is todo comment "TODO: unmap ranges from GPU that lost access", this should be handled by another patch. This patch only address that for XNACK ON, we can setup GPU mapping after page fault, without prefaulting and mapping to GPU for ACCESS orACCESS_IN_PLACE.
OK. Then you'll probably change the same code again in that follow-up patch. This patch is
Quoted 13 lines
Regards,
Philip
Quoted 8 lines
Regards,
FelixQuoted 3 lines
gpuidx =
kfd_process_gpuidx_from_gpuid(p, attrs[i].value); if (attrs[i].type ==KFD_IOCTL_SVM_ATTR_NO_ACCESS) {