Skip to contentSkip to conversation

drm/amdgpu: validate the parameters of amdgpu_vm_bo_clear_mappings

PATCH In review xinhui pan 5 participants 6 messages Apr 11, 2024 – Apr 12, 2024

Subsystems (2)
lore
Share & tools
lore

xinhui pan

author
#0 lore raw

PATCH 1/1 drm/amdgpu: validate the parameters of amdgpu_vm_bo_clear_mappings

Ensure there is no address overlapping.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +6 -0
index 8af3f0fd3073..f1315a854192 100644
@@ -1852,6 +1852,12 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 	LIST_HEAD(removed);
 	uint64_t eaddr;
 
+	/* validate the parameters */
+	if (saddr & ~PAGE_MASK || size & ~PAGE_MASK)
+		return -EINVAL;
+	if (saddr + size <= saddr)
+		return -EINVAL;
+
 	eaddr = saddr + size - 1;
 	saddr /= AMDGPU_GPU_PAGE_SIZE;
 	eaddr /= AMDGPU_GPU_PAGE_SIZE;
Signature · 1 line
2.34.1
#1 lore raw

Am 11.04.24 um 05:28 schrieb xinhui pan:

Quoted 13 lines

Ensure there is no address overlapping.

---
  drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++
  1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index 8af3f0fd3073..f1315a854192 100644

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1860

Quoted 9 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +6 -0
@@ -1852,6 +1852,12 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
  	LIST_HEAD(removed);
  	uint64_t eaddr;
 
+	/* validate the parameters */
+	if (saddr & ~PAGE_MASK || size & ~PAGE_MASK)
+		return -EINVAL;
+	if (saddr + size <= saddr)
+		return -EINVAL;
+

Why the heck should we do that? Looks invalid to me.

Regards,
Christian.

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
  	eaddr = saddr + size - 1;
  	saddr /= AMDGPU_GPU_PAGE_SIZE;
  	eaddr /= AMDGPU_GPU_PAGE_SIZE;
maintainer
#2 lore raw
In reply to xinhui pan

Am 11.04.24 um 05:28 schrieb xinhui pan:

Quoted 13 lines

Ensure there is no address overlapping.

---
  drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++
  1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index 8af3f0fd3073..f1315a854192 100644

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1857

Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
@@ -1852,6 +1852,12 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
  	LIST_HEAD(removed);
  	uint64_t eaddr;
 
+	/* validate the parameters */
+	if (saddr & ~PAGE_MASK || size & ~PAGE_MASK)
+		return -EINVAL;

Well as general rule: *never* use PAGE_MASK and other PAGE_* macros here. This is GPUVM and not related to the CPUVM space.

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1860

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
+	if (saddr + size <= saddr)
+		return -EINVAL;
+

Mhm, so basically size is not checked for a wraparound?

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1863

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
  	eaddr = saddr + size - 1;
  	saddr /= AMDGPU_GPU_PAGE_SIZE;
  	eaddr /= AMDGPU_GPU_PAGE_SIZE;

If that's the case then I would rather check for saddr < eaddr here.

But that actually shouldn't matter since this code here:

        /* Now gather all removed mappings */
        tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr); while (tmp) {

Then shouldn't return anything, so the operation is basically a NO-OP then.

Regards,
Christian.

#3 lore raw
Quoted 24 lines · Christian König

Am 11.04.24 um 05:28 schrieb xinhui pan:

Quoted 13 lines

Ensure there is no address overlapping.

---
  drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++
  1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index 8af3f0fd3073..f1315a854192 100644

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1857

Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
@@ -1852,6 +1852,12 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
      LIST_HEAD(removed);
      uint64_t eaddr;
 
+     /* validate the parameters */
+     if (saddr & ~PAGE_MASK || size & ~PAGE_MASK)
+             return -EINVAL;

Well as general rule: *never* use PAGE_MASK and other PAGE_* macros here. This is GPUVM and not related to the CPUVM space.

Quoted 5 lines

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1860

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
+     if (saddr + size <= saddr)
+             return -EINVAL;
+

Mhm, so basically size is not checked for a wraparound?

Yeah, exactly.

Quoted 5 lines

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1863

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
      eaddr = saddr + size - 1;
      saddr /= AMDGPU_GPU_PAGE_SIZE;
      eaddr /= AMDGPU_GPU_PAGE_SIZE;

If that's the case then I would rather check for saddr < eaddr here.

FWIW, it would probably a good idea to keep the added check analogous to other functions called from amdgpu_gem_va_ioctl() like amdgpu_vm_bo_replace_map(), which also checks "if (saddr + size <= saddr || offset + size <= offset)" before the division.

Quoted 7 lines

But that actually shouldn't matter since this code here:

         /* Now gather all removed mappings */
         tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr);
         while (tmp) {

Then shouldn't return anything, so the operation is basically a NO-OP then.

That's not how it works; the interval tree is not designed to be fed bogus ranges that end before they start. (Or at least I don't think it is - if it is, it is buggy.) I think basically if the specified start and end addresses are both within an rbtree node, this rbtree node is returned from the lookup, even if the specified end address is before the specified start address.

A more verbose example:
Let's assume the interval tree contains a single entry from address A to address D.
Looking at the _iter_first implementation in interval_tree_generic.h, when it is called with a start address C which is between A and D, and an end address B (result of an addition that wraps around to an address below C but above A), it does the following:

1. bails out if "node->ITSUBTREE < start" (meaning if the specified start address C is after the range covered by the root node - which is not the case)
2. bails out if "ITSTART(leftmost) > last" (meaning if the specified end address is smaller than the entry start address A - which is not the case)
3. enters _subtree_search. in there:
4. the root node has no children, so "node->ITRB.rb_left" is NULL
5. the specified end address B is after the node's start address A, so "Cond1" is fulfilled
6. the specified start address C is before the node's end address D, so "Cond2" is fulfilled
7. the root node is returned from the lookup

maintainer
#4 lore raw

Am 11.04.24 um 17:44 schrieb Jann Horn:

Quoted 31 lines
Quoted 23 lines · Christian König

Am 11.04.24 um 05:28 schrieb xinhui pan:

Quoted 13 lines

Ensure there is no address overlapping.

---
   drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++
   1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index 8af3f0fd3073..f1315a854192 100644

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1857

Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
@@ -1852,6 +1852,12 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
       LIST_HEAD(removed);
       uint64_t eaddr;
 
+     /* validate the parameters */
+     if (saddr & ~PAGE_MASK || size & ~PAGE_MASK)
+             return -EINVAL;

Well as general rule: *never* use PAGE_MASK and other PAGE_* macros here. This is GPUVM and not related to the CPUVM space.

Quoted 4 lines

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1860

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
+     if (saddr + size <= saddr)
+             return -EINVAL;
+

Mhm, so basically size is not checked for a wraparound?

Yeah, exactly.

Quoted 8 lines
Quoted 4 lines

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1863

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
       eaddr = saddr + size - 1;
       saddr /= AMDGPU_GPU_PAGE_SIZE;
       eaddr /= AMDGPU_GPU_PAGE_SIZE;

If that's the case then I would rather check for saddr < eaddr here.

FWIW, it would probably a good idea to keep the added check analogous to other functions called from amdgpu_gem_va_ioctl() like amdgpu_vm_bo_replace_map(), which also checks "if (saddr + size <= saddr || offset + size <= offset)" before the division.

I would also change that function as well.

The eaddr needs to be checked against the max_pfn as well and we currently shift that around for this check which looks quite ugly.

Only the overflow check can probably be before it.

Quoted 7 lines
Quoted 7 lines

But that actually shouldn't matter since this code here:

          /* Now gather all removed mappings */
          tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr);
          while (tmp) {
Quoted 7 lines
Quoted 1 line

Then shouldn't return anything, so the operation is basically a NO-OP then.

That's not how it works; the interval tree is not designed to be fed bogus ranges that end before they start. (Or at least I don't think it is - if it is, it is buggy.) I think basically if the specified start and end addresses are both within an rbtree node, this rbtree node is returned from the lookup, even if the specified end address is before the specified start address.

Ah, yeah that makes sense. The search functions checks if a node only partially intersects with start and end and not if it is covered by it.

Thanks,
Christian.

Quoted 22 lines

A more verbose example:
Let's assume the interval tree contains a single entry from address A to address D.
Looking at the _iter_first implementation in interval_tree_generic.h, when it is called with a start address C which is between A and D, and an end address B (result of an addition that wraps around to an address below C but above A), it does the following:

1. bails out if "node->ITSUBTREE < start" (meaning if the specified start address C is after the range covered by the root node - which is not the case)
2. bails out if "ITSTART(leftmost) > last" (meaning if the specified end address is smaller than the entry start address A - which is not the case)
3. enters _subtree_search. in there:
4. the root node has no children, so "node->ITRB.rb_left" is NULL
5. the specified end address B is after the node's start address A, so "Cond1" is fulfilled
6. the specified start address C is before the node's end address D, so "Cond2" is fulfilled
7. the root node is returned from the lookup

#5 lore raw
In reply to Christian König

+Jann Horn <[email protected]> for his thoughts

Quoted 49 lines · Christian König

Am 11.04.24 um 05:28 schrieb xinhui pan:

Review

Quoted 9 lines

Ensure there is no address overlapping.

---
  drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++
  1 file changed, 6 insertions(+)
(quoted diff) +0 -0

b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct

Quoted 4 lines

index 8af3f0fd3073..f1315a854192 100644

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
@@ -1852,6 +1852,12 @@ int amdgpu_vm_bo_clear_mappings(struct

amdgpu_device *adev,

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1857

Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
      LIST_HEAD(removed);
      uint64_t eaddr;
 
+     /* validate the parameters */
+     if (saddr & ~PAGE_MASK || size & ~PAGE_MASK)
+             return -EINVAL;

Well as general rule: *never* use PAGE_MASK and other PAGE_* macros here. This is GPUVM and not related to the CPUVM space.

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1860

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +3 -0
+     if (saddr + size <= saddr)
+             return -EINVAL;
+

Mhm, so basically size is not checked for a wraparound?

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, · line 1863

Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
      eaddr = saddr + size - 1;
      saddr /= AMDGPU_GPU_PAGE_SIZE;
      eaddr /= AMDGPU_GPU_PAGE_SIZE;

If that's the case then I would rather check for saddr < eaddr here.

But that actually shouldn't matter since this code here:

         /* Now gather all removed mappings */
         tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr);
         while (tmp) {

Then shouldn't return anything, so the operation is basically a NO-OP then.

Regards,
Christian.