[PATCH] staging: gpib: use int instead of u32 to store error codes

[PATCH] staging: gpib: use int instead of u32 to store error codes

[Qianfeng Rong]

Use int instead of size_t for 'ret' variable to store negative error codes
returned by bb_write().

Signed-off-by: Qianfeng Rong <rongqianfeng@vivo.com>
---
 drivers/staging/gpib/gpio/gpib_bitbang.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/staging/gpib/gpio/gpib_bitbang.c b/drivers/staging/gpib/gpio/gpib_bitbang.c
index 22a55f3f794d..374cd61355e9 100644
--- a/drivers/staging/gpib/gpio/gpib_bitbang.c
+++ b/drivers/staging/gpib/gpio/gpib_bitbang.c
@@ -726,7 +726,7 @@ static irqreturn_t bb_SRQ_interrupt(int irq, void *arg)
 static int bb_command(struct gpib_board *board, u8 *buffer,
 		      size_t length, size_t *bytes_written)
 {
-	size_t ret;
+	int ret;
 	struct bb_priv *priv = board->private_data;
 	int i;
 
-- 
2.34.1

Re: [PATCH] staging: gpib: use int instead of u32 to store error codes

[Dan Carpenter]

On Tue, Aug 26, 2025 at 05:58:59PM +0800, Qianfeng Rong wrote:
> Use int instead of size_t for 'ret' variable to store negative error codes
> returned by bb_write().
> 

The commit message needs to give a motivation.  I would have assumed from
reading the patch that this caused a signedness bug but it doesn't.
The commit has no effect on runtime but fixing the type is a nice
cleanup.  My commit message would say:

The "ret" variable is used to store the return from bb_write() returns
either zero on success or negative error codes on failure.  Storing the
error codes in size_t which is an unsigned long, doesn't cause an issue
at runtime but it's ugly as pants.  Change "ret" from size_t to int
type.  No effect on runtime.

regards,
dan carpenter

Re: [PATCH] staging: gpib: use int instead of u32 to store error codes

[Qianfeng Rong]

在 2025/8/26 18:19, Dan Carpenter 写道:
> On Tue, Aug 26, 2025 at 05:58:59PM +0800, Qianfeng Rong wrote:
>> Use int instead of size_t for 'ret' variable to store negative error codes
>> returned by bb_write().
>>
> The commit message needs to give a motivation.  I would have assumed from
> reading the patch that this caused a signedness bug but it doesn't.
> The commit has no effect on runtime but fixing the type is a nice
> cleanup.  My commit message would say:
>
> The "ret" variable is used to store the return from bb_write() returns
> either zero on success or negative error codes on failure.  Storing the
> error codes in size_t which is an unsigned long, doesn't cause an issue
> at runtime but it's ugly as pants.  Change "ret" from size_t to int
> type.  No effect on runtime.


Thanks very much for the detailed comments.  I will send v2 with a commit
message similar to the one you suggested.

Best regards,
Qianfeng