mfd: viperboard: remove redundant NULL check before kfree()
PATCH 1/1 mfd: viperboard: remove redundant NULL check before kfree()
From: Mohammad Shahid <[email protected]>
kfree() safely handles NULL pointers, so the explicit NULL check before calling kfree() is unnecessary.
This issue was reported by ifnullfree.cocci.
--- drivers/mfd/viperboard.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/mfd/viperboard.c +1 -2
@@ -96,8 +96,7 @@ static int vprbrd_probe(struct usb_interface *interface,return 0 ; error: -if (vb) -kfree (vb); +kfree (vb);return ret; }
Signature · 1 line
2.43.0
[email protected]
AI reviewThank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] DMA cacheline sharing memory corruption on struct vprbrd --
commit 226dd61211b7aeb126296ed3bc673a04067d3fee
Author: Mohammad Shahid <[email protected]>
mfd: viperboard: remove redundant NULL check before kfree()
This patch cleans up the error path in vprbrd_probe() by removing an unnecessary NULL check before calling kfree(). Since kfree() safely handles NULL pointers, the explicit check can be safely removed.
drivers/mfd/viperboard.c static int vprbrd_probe(struct usb_interface *interface, · line 99
Quoted 11 lines
drivers/mfd/viperboard.c +1
-2@@ -96,8 +96,7 @@ static int vprbrd_probe(struct usb_interface *interface,return 0 ; error: -if (vb) -kfree (vb); +kfree (vb);
[Severity: High]
This is a pre-existing issue and was not introduced by the patch under review, but while reviewing this cleanup, I noticed a potential DMA cacheline sharing issue with the vb structure being freed here.
In include/linux/mfd/viperboard.h, struct vprbrd embeds a buffer directly between a mutex and a platform_device:
struct vprbrd {
...
struct mutex lock;
u8 buf[sizeof(struct vprbrd_i2c_write_msg)];
struct platform_device pdev;
...
};During vprbrd_probe() in drivers/mfd/viperboard.c, this unaligned buffer is passed directly to usb_control_msg() for DMA transfers:
ret = usb_control_msg(vb->usb_dev, pipe, VPRBRD_USB_REQUEST_MAJOR, VPRBRD_USB_TYPE_IN, 0x0000, 0x0000, vb->buf, 1, VPRBRD_USB_TIMEOUT_MS);
Because buf is not explicitly cacheline-aligned (for example, using ____cacheline_aligned), it shares a cacheline with the preceding lock and the following pdev.
On non-coherent DMA architectures, when the DMA API invalidates the CPU cache lines covering buf during a receive transfer, it could discard concurrent modifications to the lock or pdev.
Can this memory layout cause silent memory corruption or deadlocks if another thread modifies vb->lock while a DMA receive operation like this one, or the runtime transfers in drivers/i2c/busses/i2c-viperboard.c:vprbrd_i2c_receive(), are in flight?
Quoted 3 lines
drivers/mfd/viperboard.c +0
-0return ret; }
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
Re: (subset) [PATCH] mfd: viperboard: remove redundant NULL check before kfree()
In reply to [email protected]Quoted 4 lines · On Fri, 03 Jul 2026 16:29:37 +0530, [email protected] wrote:
kfree()safely handles NULL pointers, so the explicit NULL check before callingkfree()is unnecessary.This issue was reported by
ifnullfree.cocci.
Applied, thanks!
[1/1] mfd: viperboard: remove redundant NULL check before kfree()
commit: cf9775233b4d1618dfac5e84009f36eba1e7fddc
--
Lee Jones [李琼斯]