Skip to contentSkip to conversation

drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

PATCH In review Felix Kuehling 5 participants 8 messages Jul 28, 2020 – Jul 30, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

---
 drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo,
 int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
 		     void *buf, int len, int write)
 {
-	unsigned long offset = (addr) - vma->vm_start;
 	struct ttm_buffer_object *bo = vma->vm_private_data;
+	unsigned long offset = (addr) - vma->vm_start +
+		((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+		 << PAGE_SHIFT);
 	int ret;
 
 	if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)
Signature · 6 lines
2.17.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Reviewed-by on v1 1/1
#1 lore raw

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Quoted 10 lines
---
  drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo,
  int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, · line 414

Quoted 10 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
  		     void *buf, int len, int write)
  {
-	unsigned long offset = (addr) - vma->vm_start;
  	struct ttm_buffer_object *bo = vma->vm_private_data;
+	unsigned long offset = (addr) - vma->vm_start +
+		((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+		 << PAGE_SHIFT);
  	int ret;
 
  	if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

#2 lore raw

Am 2020-07-29 um 4:08 a.m. schrieb Christian König:

Quoted 11 lines

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Thanks. I'll submit it to amd-staging-drm-next so it makes its way into our DKMS branch quickly.

Alex, would you push this to drm-fixes?

Regards,
  Felix

Quoted 14 lines
Quoted 14 lines

---
  drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-

  1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct
 ttm_buffer_object *bo,
   int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
                void *buf, int len, int write)
Quoted 9 lines

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, · line 414

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
   {
-    unsigned long offset = (addr) - vma->vm_start;
       struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
       int ret;
         if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

#3 lore raw

[AMD Public Use]

Christian, Can you cc stable when you apply it to drm-misc?

Alex
________________________________
From: Kuehling, Felix <[email protected]>
Sent: Wednesday, July 29, 2020 10:15 AM
To: Koenig, Christian <[email protected]>; [email protected] <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>
Cc: Morichetti, Laurent <[email protected]>
Subject: Re: [PATCH 1/1] drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

Am 2020-07-29 um 4:08 a.m. schrieb Christian König:

Quoted 11 lines

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Thanks. I'll submit it to amd-staging-drm-next so it makes its way into our DKMS branch quickly.

Alex, would you push this to drm-fixes?

Regards,
Felix

Quoted 14 lines
Quoted 14 lines
---
  drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct
 ttm_buffer_object *bo,
  int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
               void *buf, int len, int write)
Quoted 9 lines

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, · line 414

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
  {
-    unsigned long offset = (addr) - vma->vm_start;
      struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
      int ret;
        if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#4 lore raw

Sure.

Christian.

Am 29.07.2020 17:30 schrieb "Deucher, Alexander" <[email protected]>:

[AMD Public Use]

Christian, Can you cc stable when you apply it to drm-misc?

Alex
________________________________
From: Kuehling, Felix <[email protected]>
Sent: Wednesday, July 29, 2020 10:15 AM
To: Koenig, Christian <[email protected]>; [email protected] <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>
Cc: Morichetti, Laurent <[email protected]>
Subject: Re: [PATCH 1/1] drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

Am 2020-07-29 um 4:08 a.m. schrieb Christian König:

Quoted 11 lines

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Thanks. I'll submit it to amd-staging-drm-next so it makes its way into our DKMS branch quickly.

Alex, would you push this to drm-fixes?

Regards,
Felix

Quoted 14 lines
Quoted 14 lines
---
  drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct
 ttm_buffer_object *bo,
  int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
               void *buf, int len, int write)
Quoted 9 lines

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, · line 414

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
  {
-    unsigned long offset = (addr) - vma->vm_start;
      struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
      int ret;
        if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

#5 lore raw
Quoted 2 lines · Christian

Sure.

Note that drm-misc-next isn't the right branch for cc: stable stuff, see

https://drm.pages.freedesktop.org/maintainer-tools/committer-drm-misc.html#where-do-i-apply-my-patch

Just to avoid confusion and needless cherrypicking across branches. -Daniel

Quoted 67 lines

Christian.

Am 29.07.2020 17:30 schrieb "Deucher, Alexander" <[email protected]>:

[AMD Public Use]

Christian, Can you cc stable when you apply it to drm-misc?

Alex
________________________________
From: Kuehling, Felix <[email protected]>
Sent: Wednesday, July 29, 2020 10:15 AM
To: Koenig, Christian <[email protected]>; [email protected] <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>
Cc: Morichetti, Laurent <[email protected]>
Subject: Re: [PATCH 1/1] drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

Am 2020-07-29 um 4:08 a.m. schrieb Christian König:

Quoted 11 lines

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Thanks. I'll submit it to amd-staging-drm-next so it makes its way into our DKMS branch quickly.

Alex, would you push this to drm-fixes?

Regards,
Felix

Quoted 14 lines
Quoted 14 lines
---
  drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct
 ttm_buffer_object *bo,
  int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
               void *buf, int len, int write)
Quoted 9 lines

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, · line 414

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
  {
-    unsigned long offset = (addr) - vma->vm_start;
      struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
      int ret;
        if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#6 lore raw

Am 30.07.20 um 13:40 schrieb Daniel Vetter:

Quoted 6 lines
Quoted 1 line · Christian

Sure.

Note that drm-misc-next isn't the right branch for cc: stable stuff, see

https://drm.pages.freedesktop.org/maintainer-tools/committer-drm-misc.html#where-do-i-apply-my-patch

So this should go into drm-misc-fixes? The bug is there for quite a while.

Christian.

Quoted 68 lines

Just to avoid confusion and needless cherrypicking across branches. -Daniel

Quoted 62 lines

Christian.

Am 29.07.2020 17:30 schrieb "Deucher, Alexander" <[email protected]>:

[AMD Public Use]

Christian, Can you cc stable when you apply it to drm-misc?

Alex
________________________________
From: Kuehling, Felix <[email protected]>
Sent: Wednesday, July 29, 2020 10:15 AM
To: Koenig, Christian <[email protected]>; [email protected] <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>
Cc: Morichetti, Laurent <[email protected]>
Subject: Re: [PATCH 1/1] drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

Am 2020-07-29 um 4:08 a.m. schrieb Christian König:

Quoted 10 lines

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Thanks. I'll submit it to amd-staging-drm-next so it makes its way into our DKMS branch quickly.

Alex, would you push this to drm-fixes?

Regards,
Felix

Quoted 13 lines
Quoted 13 lines
---
   drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-
   1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct
 ttm_buffer_object *bo,
   int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
                void *buf, int len, int write)
Quoted 8 lines

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, · line 414

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
   {
-    unsigned long offset = (addr) - vma->vm_start;
       struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
       int ret;
         if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

#7 lore raw

Am 30.07.20 um 13:50 schrieb Christian König:

Quoted 10 lines

Am 30.07.20 um 13:40 schrieb Daniel Vetter:

Quoted 7 lines
Quoted 1 line · Christian

Sure.

Note that drm-misc-next isn't the right branch for cc: stable stuff, see

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdrm.pages.freedesktop.org%2Fmaintainer-tools%2Fcommitter-drm-misc.html%23where-do-i-apply-my-patch&amp;data=02%7C01%7CChristian.Koenig%40amd.com%7Cfdb80f31688743c1279f08d8347ece50%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637317066517801827&amp;sdata=gKCTEgJLaP9f96f3wJUGu0yfaV%2FyG3U3WvXDJ5%2FYrOc%3D&amp;reserved=0

So this should go into drm-misc-fixes? The bug is there for quite a while.

Ok, so I pushed this to drm-misc-fixes, but forgot to add the CC stable tag.

To much to do and not enough time,
Christian.

Quoted 73 lines

Christian.

Quoted 69 lines

Just to avoid confusion and needless cherrypicking across branches. -Daniel

Quoted 63 lines

Christian.

Am 29.07.2020 17:30 schrieb "Deucher, Alexander" <[email protected]>:

[AMD Public Use]

Christian, Can you cc stable when you apply it to drm-misc?

Alex
________________________________
From: Kuehling, Felix <[email protected]>
Sent: Wednesday, July 29, 2020 10:15 AM
To: Koenig, Christian <[email protected]>; [email protected] <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>
Cc: Morichetti, Laurent <[email protected]>
Subject: Re: [PATCH 1/1] drm/ttm: fix offset in VMAs with a pg_offs in ttm_bo_vm_access

Am 2020-07-29 um 4:08 a.m. schrieb Christian König:

Quoted 10 lines

Am 28.07.20 um 20:27 schrieb Felix Kuehling:

Quoted 6 lines

VMAs with a pg_offs that's offset from the start of the vma_node need to adjust the offset within the BO accordingly. This matches the offset calculation in ttm_bo_vm_fault_reserved.

Going to pick that up for inclusion in drm-misc-next.

Thanks. I'll submit it to amd-staging-drm-next so it makes its way into our DKMS branch quickly.

Alex, would you push this to drm-fixes?

Regards,
   Felix

Quoted 13 lines
Quoted 13 lines

---
   drivers/gpu/drm/ttm/ttm_bo_vm.c | 4 +++-

   1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
index 389128b8c4dd..60b41447bec8 100644
@@ -405,8 +405,10 @@ static int ttm_bo_vm_access_kmap(struct
 ttm_buffer_object *bo,
    int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
                 void *buf, int len, int write)
Quoted 8 lines

drivers/gpu/drm/ttm/ttm_bo_vm.c static int ttm_bo_vm_access_kmap(struct · line 414

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo_vm.c +3 -1
    {
-    unsigned long offset = (addr) - vma->vm_start;
        struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
        int ret;
          if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages)

_______________________________________________
dri-devel mailing list
[email protected]
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Fdri-devel&amp;data=02%7C01%7CChristian.Koenig%40amd.com%7Cfdb80f31688743c1279f08d8347ece50%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637317066517801827&amp;sdata=ZQofPqGAHNmglMcDPQzjpq0AzGbzaUx5RAfwr9zfJ%2FQ%3D&amp;reserved=0

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx