The problem is that "req->actual" is a u32, "req->status" is an int, and
iocb->ki_complete() takes a long. We would expect that a negative error
code in "req->status" would translate to a negative long value.
But what actually happens is that because "req->actual" is a u32, the
error codes is type promoted to a high positive value and then remains
a positive value when it is cast to long. (No sign expansion).
We can fix this by casting "req->status" to long.
Signed-off-by: Dan Carpenter <[email protected]>
---
drivers/usb/gadget/legacy/inode.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c
index 71e7d10dd76b..cd8e2737947b 100644
--- a/drivers/usb/gadget/legacy/inode.c
+++ b/drivers/usb/gadget/legacy/inode.c
@@ -498,7 +498,8 @@ static void ep_aio_complete(struct usb_ep *ep, struct usb_request *req)
iocb->private = NULL;
/* aio_complete() reports bytes-transferred _and_ faults */
- iocb->ki_complete(iocb, req->actual ? req->actual : req->status,
+ iocb->ki_complete(iocb,
+ req->actual ? req->actual : (long)req->status,
req->status);
} else {
/* ep_copy_to_user() won't report both; we hide some faults */
--
2.30.2
Dan Carpenter <[email protected]> writes:
> The problem is that "req->actual" is a u32, "req->status" is an int, and
> iocb->ki_complete() takes a long. We would expect that a negative error
> code in "req->status" would translate to a negative long value.
>
> But what actually happens is that because "req->actual" is a u32, the
> error codes is type promoted to a high positive value and then remains
> a positive value when it is cast to long. (No sign expansion).
>
> We can fix this by casting "req->status" to long.
>
> Signed-off-by: Dan Carpenter <[email protected]>
I'm just going to assume your type promotion rank is correct :-)
Acked-by: Felipe Balbi <[email protected]>
--
balbi