Skip to contentSkip to conversation

drm/amd/display: Use kmalloc_array() instead of kmalloc()

PATCH Applied James Flowers 3 participants 3 messages Sep 13, 2025 – Sep 16, 2025

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

PATCH 1/1 drm/amd/display: Use kmalloc_array() instead of kmalloc()

Documentation/process/deprecated.rst recommends against the use of kmalloc with dynamic size calculations due to the risk of overflow and smaller allocation being made than the caller was expecting. This could lead to buffer overflow in code similar to the memcpy in amdgpu_dm_plane_add_modifier().

---
I see that in amdgpu_dm_plane_get_plane_modifiers, capacity is initialized to only 128, but it is probably preferable to refactor.

Tested on a Steam Deck OLED with no apparent regressions using these test suites from igt-gpu-tools:
1) kms_plane
2) amd_plane
3) amd_fuzzing
4) testdisplay

 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +2 -2
index b7c6e8d13435..b587d2033f0b 100644
@@ -146,7 +146,7 @@ static void amdgpu_dm_plane_add_modifier(uint64_t **mods, uint64_t *size, uint64
 
 	if (*cap - *size < 1) {
 		uint64_t new_cap = *cap * 2;
-		uint64_t *new_mods = kmalloc(new_cap * sizeof(uint64_t), GFP_KERNEL);
+		uint64_t *new_mods = kmalloc_array(new_cap, sizeof(uint64_t), GFP_KERNEL);
 
 		if (!new_mods) {
 			kfree(*mods);
@@ -732,7 +732,7 @@ static int amdgpu_dm_plane_get_plane_modifiers(struct amdgpu_device *adev, unsig
 	if (adev->family < AMDGPU_FAMILY_AI)
 		return 0;
 
-	*mods = kmalloc(capacity * sizeof(uint64_t), GFP_KERNEL);
+	*mods = kmalloc_array(capacity, sizeof(uint64_t), GFP_KERNEL);
 
 	if (plane_type == DRM_PLANE_TYPE_CURSOR) {
 		amdgpu_dm_plane_add_modifier(mods, &size, &capacity, DRM_FORMAT_MOD_LINEAR);
Signature · 1 line
2.51.0
Applied
#1 lore raw

Applied. Thanks!

Alex

Quoted 47 lines · James Flowers

Documentation/process/deprecated.rst recommends against the use of kmalloc with dynamic size calculations due to the risk of overflow and smaller allocation being made than the caller was expecting. This could lead to buffer overflow in code similar to the memcpy in amdgpu_dm_plane_add_modifier().

---
I see that in amdgpu_dm_plane_get_plane_modifiers, capacity is initialized to only 128, but it is probably preferable to refactor.

Tested on a Steam Deck OLED with no apparent regressions using these test suites from igt-gpu-tools:
1) kms_plane
2) amd_plane
3) amd_fuzzing
4) testdisplay

 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +2 -3
index b7c6e8d13435..b587d2033f0b 100644
@@ -146,7 +146,7 @@ static void amdgpu_dm_plane_add_modifier(uint64_t **mods, uint64_t *size, uint64
 
        if (*cap - *size < 1) {
                uint64_t new_cap = *cap * 2;
-               uint64_t *new_mods = kmalloc(new_cap * sizeof(uint64_t), GFP_KERNEL);
+               uint64_t *new_mods = kmalloc_array(new_cap, sizeof(uint64_t), GFP_KERNEL);
 
                if (!new_mods) {
                        kfree(*mods);
@@ -732,7 +732,7 @@ static int amdgpu_dm_plane_get_plane_modifiers(struct amdgpu_device *adev, unsig
        if (adev->family < AMDGPU_FAMILY_AI)
                return 0;
 
-       *mods = kmalloc(capacity * sizeof(uint64_t), GFP_KERNEL);
+       *mods = kmalloc_array(capacity, sizeof(uint64_t), GFP_KERNEL);
 
        if (plane_type == DRM_PLANE_TYPE_CURSOR) {
                amdgpu_dm_plane_add_modifier(mods, &size, &capacity, DRM_FORMAT_MOD_LINEAR);
--

2.51.0

Reviewed-by on v1 1/1
#2 lore raw
In reply to James Flowers
Quoted 43 lines · James Flowers

Documentation/process/deprecated.rst recommends against the use of kmalloc with dynamic size calculations due to the risk of overflow and smaller allocation being made than the caller was expecting. This could lead to buffer overflow in code similar to the memcpy in amdgpu_dm_plane_add_modifier().

---
I see that in amdgpu_dm_plane_get_plane_modifiers, capacity is initialized to only 128, but it is probably preferable to refactor.

Tested on a Steam Deck OLED with no apparent regressions using these test suites from igt-gpu-tools:
1) kms_plane
2) amd_plane
3) amd_fuzzing
4) testdisplay

  drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +2 -2
index b7c6e8d13435..b587d2033f0b 100644
@@ -146,7 +146,7 @@ static void amdgpu_dm_plane_add_modifier(uint64_t **mods, uint64_t *size, uint64
 
  	if (*cap - *size < 1) {
  		uint64_t new_cap = *cap * 2;
-		uint64_t *new_mods = kmalloc(new_cap * sizeof(uint64_t), GFP_KERNEL);
+		uint64_t *new_mods = kmalloc_array(new_cap, sizeof(uint64_t), GFP_KERNEL);
 
  		if (!new_mods) {
  			kfree(*mods);
@@ -732,7 +732,7 @@ static int amdgpu_dm_plane_get_plane_modifiers(struct amdgpu_device *adev, unsig
  	if (adev->family < AMDGPU_FAMILY_AI)
  		return 0;
 
-	*mods = kmalloc(capacity * sizeof(uint64_t), GFP_KERNEL);
+	*mods = kmalloc_array(capacity, sizeof(uint64_t), GFP_KERNEL);
 
  	if (plane_type == DRM_PLANE_TYPE_CURSOR) {
  		amdgpu_dm_plane_add_modifier(mods, &size, &capacity, DRM_FORMAT_MOD_LINEAR);