Skip to contentSkip to conversation

ARM: Fix MAX_DMA_ADDRESS overflow

PATCH In review Florian Fainelli 2 participants 4 messages Jul 6, 2022 – Jul 19, 2022 Arm (32-bit)

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 ARM: Fix MAX_DMA_ADDRESS overflow

Commit 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis") added a check to determine whether arm_dma_zone_size is exceeding the amount of kernel virtual address space available between the upper 4GB virtual address limit and PAGE_OFFSET in order to provide a suitable definition of MAX_DMA_ADDRESS that should fit within the 32-bit virtual address space. The quantity used for comparison was off by a missing trailing 0, leading to MAX_DMA_ADDRESS to be overflowing a 32-bit quantity.

This was caught thanks to CONFIG_DEBUG_VIRTUAL on the bcm2711 platform where we define a dma_zone_size of 1GB and we have a PAGE_OFFSET value of 0xc000_0000 (CONFIG_VMSPLIT_3G) leading to MAX_DMA_ADDRESS being 0x1_0000_0000 which overflows the unsigned long type used throughout __pa() and then __virt_addr_valid(). Because the virtual address passed to __virt_addr_valid() would now be 0, the function would loudly warn and flood the kernel log, thus making the platform unable to boot properly.

Fixes 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis")

---
Changes in v3:

- added ULL suffix to the constant
- reworded the commit message to be clearer

Changes in v2:

- simplify the patch and drop the first patch that attempted to fix an off by one in the calculation.

 arch/arm/include/asm/dma.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/include/asm/dma.h +1 -1
index a81dda65c576..45180a2cc47c 100644
@@ -10,7 +10,7 @@
 #else
 #define MAX_DMA_ADDRESS	({ \
 	extern phys_addr_t arm_dma_zone_size; \
-	arm_dma_zone_size && arm_dma_zone_size < (0x10000000 - PAGE_OFFSET) ? \
+	arm_dma_zone_size && arm_dma_zone_size < (0x100000000ULL - PAGE_OFFSET) ? \
 		(PAGE_OFFSET + arm_dma_zone_size) : 0xffffffffUL; })
 #endif
 
Signature · 1 line
2.25.1
#1 lore raw
Quoted 18 lines · Florian Fainelli

Commit 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis") added a check to determine whether arm_dma_zone_size is exceeding the amount of kernel virtual address space available between the upper 4GB virtual address limit and PAGE_OFFSET in order to provide a suitable definition of MAX_DMA_ADDRESS that should fit within the 32-bit virtual address space. The quantity used for comparison was off by a missing trailing 0, leading to MAX_DMA_ADDRESS to be overflowing a 32-bit quantity.

This was caught thanks to CONFIG_DEBUG_VIRTUAL on the bcm2711 platform where we define a dma_zone_size of 1GB and we have a PAGE_OFFSET value of 0xc000_0000 (CONFIG_VMSPLIT_3G) leading to MAX_DMA_ADDRESS being 0x1_0000_0000 which overflows the unsigned long type used throughout __pa() and then __virt_addr_valid(). Because the virtual address passed to __virt_addr_valid() would now be 0, the function would loudly warn and flood the kernel log, thus making the platform unable to boot properly.

Quoted 2 lines
Fixes 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis")

Shall I send this to RMK's patch system?

Signature · 1 line
Florian
Reviewed-by
#2 lore raw
Quoted 19 lines · Florian Fainelli
Quoted 18 lines · Florian Fainelli

Commit 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis") added a check to determine whether arm_dma_zone_size is exceeding the amount of kernel virtual address space available between the upper 4GB virtual address limit and PAGE_OFFSET in order to provide a suitable definition of MAX_DMA_ADDRESS that should fit within the 32-bit virtual address space. The quantity used for comparison was off by a missing trailing 0, leading to MAX_DMA_ADDRESS to be overflowing a 32-bit quantity.

This was caught thanks to CONFIG_DEBUG_VIRTUAL on the bcm2711 platform where we define a dma_zone_size of 1GB and we have a PAGE_OFFSET value of 0xc000_0000 (CONFIG_VMSPLIT_3G) leading to MAX_DMA_ADDRESS being 0x1_0000_0000 which overflows the unsigned long type used throughout __pa() and then __virt_addr_valid(). Because the virtual address passed to __virt_addr_valid() would now be 0, the function would loudly warn and flood the kernel log, thus making the platform unable to boot properly.

Quoted 4 lines
Quoted 2 lines
Fixes 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis")

Shall I send this to RMK's patch system?

I think so!

Yours,
Linus Walleij

#3 lore raw
Quoted 20 lines · Linus Walleij
Quoted 19 lines · Florian Fainelli
Quoted 18 lines · Florian Fainelli

Commit 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis") added a check to determine whether arm_dma_zone_size is exceeding the amount of kernel virtual address space available between the upper 4GB virtual address limit and PAGE_OFFSET in order to provide a suitable definition of MAX_DMA_ADDRESS that should fit within the 32-bit virtual address space. The quantity used for comparison was off by a missing trailing 0, leading to MAX_DMA_ADDRESS to be overflowing a 32-bit quantity.

This was caught thanks to CONFIG_DEBUG_VIRTUAL on the bcm2711 platform where we define a dma_zone_size of 1GB and we have a PAGE_OFFSET value of 0xc000_0000 (CONFIG_VMSPLIT_3G) leading to MAX_DMA_ADDRESS being 0x1_0000_0000 which overflows the unsigned long type used throughout __pa() and then __virt_addr_valid(). Because the virtual address passed to __virt_addr_valid() would now be 0, the function would loudly warn and flood the kernel log, thus making the platform unable to boot properly.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Fixes 26f09e9b3a06 ("mm/memblock: add memblock memory allocation apis")

Shall I send this to RMK's patch system?

I think so!

Thanks:

https://www.armlinux.org.uk/developer/patches/viewpatch.php?id=9216/1

Signature · 1 line
Florian