Skip to contentSkip to conversation

hw/display/exynos4210_fimd: Fix potential NULL pointer dereference

PATCH In review AlexChen 2 participants 4 messages Oct 30, 2020 – Oct 31, 2020 Arm Samsung Exynos

lore
Share & tools
lore

AlexChen

author
#0 lore raw

PATCH 1/1 hw/display/exynos4210_fimd: Fix potential NULL pointer dereference

In exynos4210_fimd_update(), the pointer s is dereferenced before being check if it is valid, which may lead to NULL pointer dereference. So move the assignment to global_width after checking that the s is valid

---
 hw/display/exynos4210_fimd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/display/exynos4210_fimd.c +1 -1
index 4c16e1f5a0..a1179d2f89 100644
@@ -1275,12 +1275,12 @@ static void exynos4210_fimd_update(void *opaque)
     bool blend = false;
     uint8_t *host_fb_addr;
     bool is_dirty = false;
-    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
 
     if (!s || !s->console || !s->enabled ||
         surface_bits_per_pixel(qemu_console_surface(s->console)) == 0) {
         return;
     }
+    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
     exynos4210_update_resolution(s);
     surface = qemu_console_surface(s->console);
 
Signature · 1 line
2.19.1
#1 lore raw
Quoted 18 lines · AlexChen

In exynos4210_fimd_update(), the pointer s is dereferenced before being check if it is valid, which may lead to NULL pointer dereference. So move the assignment to global_width after checking that the s is valid

---
 hw/display/exynos4210_fimd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/display/exynos4210_fimd.c +0 -0
index 4c16e1f5a0..a1179d2f89 100644
@@ -1275,12 +1275,12 @@ static void exynos4210_fimd_update(void *opaque)
     bool blend = false;
     uint8_t *host_fb_addr;

hw/display/exynos4210_fimd.c static void exynos4210_fimd_update(void *opaque) · line 1285

Quoted 10 lines
hw/display/exynos4210_fimd.c +1 -1
     bool is_dirty = false;
-    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
 
     if (!s || !s->console || !s->enabled ||
         surface_bits_per_pixel(qemu_console_surface(s->console)) == 0) {
         return;
     }
+    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
     exynos4210_update_resolution(s);
     surface = qemu_console_surface(s->console);

Yep, this is a bug, but QEMU's coding style doesn't permit variable declarations in the middle of functions. You need to split the declaration (which stays where it is) and the initialization (which can moved down below the !s test.

thanks
-- PMM

author
#2 lore raw
Quoted 22 lines · Peter Maydell
Quoted 21 lines · AlexChen

In exynos4210_fimd_update(), the pointer s is dereferenced before being check if it is valid, which may lead to NULL pointer dereference. So move the assignment to global_width after checking that the s is valid

---
 hw/display/exynos4210_fimd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/display/exynos4210_fimd.c +0 -1
index 4c16e1f5a0..a1179d2f89 100644
@@ -1275,12 +1275,12 @@ static void exynos4210_fimd_update(void *opaque)
     bool blend = false;
     uint8_t *host_fb_addr;
     bool is_dirty = false;
-    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
 
Quoted 13 lines

hw/display/exynos4210_fimd.c static void exynos4210_fimd_update(void *opaque) · line 1285

Quoted 7 lines
hw/display/exynos4210_fimd.c +1 -0
     if (!s || !s->console || !s->enabled ||
         surface_bits_per_pixel(qemu_console_surface(s->console)) == 0) {
         return;
     }
+    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
     exynos4210_update_resolution(s);
     surface = qemu_console_surface(s->console);

Yep, this is a bug, but QEMU's coding style doesn't permit variable declarations in the middle of functions. You need to split the declaration (which stays where it is) and the initialization (which can moved down below the !s test.

Thans for your review. I have also considered this modification to be incompatible with the QEMU's coding style. But the type of global_width is const int which cannot be assigned once they are defined.
Could we define the global_width as int, such as this modification:

hw/display/exynos4210_fimd.c +3 -1
index 4c16e1f5a0..34a960a976 100644
@@ -1275,12 +1275,14 @@ static void exynos4210_fimd_update(void *opaque)
     bool blend = false;
     uint8_t *host_fb_addr;
     bool is_dirty = false;
-    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
+    int global_width;
 
     if (!s || !s->console || !s->enabled ||
         surface_bits_per_pixel(qemu_console_surface(s->console)) == 0) {
         return;
     }
+
+    global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
     exynos4210_update_resolution(s);
     surface = qemu_console_surface(s->console);
 
#3 lore raw
Quoted 51 lines · AlexChen
Quoted 22 lines · Peter Maydell
Quoted 21 lines · AlexChen

In exynos4210_fimd_update(), the pointer s is dereferenced before being check if it is valid, which may lead to NULL pointer dereference. So move the assignment to global_width after checking that the s is valid

---
 hw/display/exynos4210_fimd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/display/exynos4210_fimd.c +0 -1
index 4c16e1f5a0..a1179d2f89 100644
@@ -1275,12 +1275,12 @@ static void exynos4210_fimd_update(void *opaque)
     bool blend = false;
     uint8_t *host_fb_addr;
     bool is_dirty = false;
-    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
 
Quoted 13 lines

hw/display/exynos4210_fimd.c static void exynos4210_fimd_update(void *opaque) · line 1287

Quoted 7 lines
hw/display/exynos4210_fimd.c +1 -0
     if (!s || !s->console || !s->enabled ||
         surface_bits_per_pixel(qemu_console_surface(s->console)) == 0) {
         return;
     }
+    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
     exynos4210_update_resolution(s);
     surface = qemu_console_surface(s->console);

Yep, this is a bug, but QEMU's coding style doesn't permit variable declarations in the middle of functions. You need to split the declaration (which stays where it is) and the initialization (which can moved down below the !s test.

Thans for your review. I have also considered this modification to be incompatible with the QEMU's coding style. But the type of global_width is const int which cannot be assigned once they are defined.
Could we define the global_width as int, such as this modification:

hw/display/exynos4210_fimd.c +0 -1
index 4c16e1f5a0..34a960a976 100644
@@ -1275,12 +1275,14 @@ static void exynos4210_fimd_update(void *opaque)
     bool blend = false;
     uint8_t *host_fb_addr;
     bool is_dirty = false;
-    const int global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;

hw/display/exynos4210_fimd.c static void exynos4210_fimd_update(void *opaque) · line 1287

Quoted 10 lines
hw/display/exynos4210_fimd.c +3 -0
+    int global_width;
 
     if (!s || !s->console || !s->enabled ||
         surface_bits_per_pixel(qemu_console_surface(s->console)) == 0) {
         return;
     }
+
+    global_width = (s->vidtcon[2] & FIMD_VIDTCON2_SIZE_MASK) + 1;
     exynos4210_update_resolution(s);
     surface = qemu_console_surface(s->console);

Yes, I think that would be fine.

thanks
-- PMM