[PATCH] minimal fixes for drivers/usb/gadget/m66592-udc.c

From: Al Viro
Date: Sun Jul 15 2007 - 15:59:30 EST



still looks racy (and definitely leaks)

Signed-off-by: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
---
drivers/usb/gadget/m66592-udc.c | 27 ++++-----------------------
1 files changed, 4 insertions(+), 23 deletions(-)

diff --git a/drivers/usb/gadget/m66592-udc.c b/drivers/usb/gadget/m66592-udc.c
index c0a962b..0174a32 100644
--- a/drivers/usb/gadget/m66592-udc.c
+++ b/drivers/usb/gadget/m66592-udc.c
@@ -937,6 +937,7 @@ static void get_status(struct m66592 *m66592, struct usb_ctrlrequest *ctrl)
*m66592->ep0_buf = status;
m66592->ep0_req->buf = m66592->ep0_buf;
m66592->ep0_req->length = 2;
+ /* AV: what happens if we get called again before that gets through? */
m66592_queue(m66592->gadget.ep0, m66592->ep0_req, GFP_KERNEL);
}

@@ -1254,24 +1255,6 @@ static void m66592_free_request(struct usb_ep *_ep, struct usb_request *_req)
kfree(req);
}

-static void *m66592_alloc_buffer(struct usb_ep *_ep, unsigned bytes,
- dma_addr_t *dma, gfp_t gfp_flags)
-{
- void *buf;
-
- buf = kzalloc(bytes, gfp_flags);
- if (dma)
- *dma = virt_to_bus(buf);
-
- return buf;
-}
-
-static void m66592_free_buffer(struct usb_ep *_ep, void *buf,
- dma_addr_t dma, unsigned bytes)
-{
- kfree(buf);
-}
-
static int m66592_queue(struct usb_ep *_ep, struct usb_request *_req,
gfp_t gfp_flags)
{
@@ -1378,9 +1361,6 @@ static struct usb_ep_ops m66592_ep_ops = {
.alloc_request = m66592_alloc_request,
.free_request = m66592_free_request,

- .alloc_buffer = m66592_alloc_buffer,
- .free_buffer = m66592_free_buffer,
-
.queue = m66592_queue,
.dequeue = m66592_dequeue,

@@ -1603,11 +1583,12 @@ static int __init m66592_probe(struct platform_device *pdev)

the_controller = m66592;

+ /* AV: leaks */
m66592->ep0_req = m66592_alloc_request(&m66592->ep[0].ep, GFP_KERNEL);
if (m66592->ep0_req == NULL)
goto clean_up;
- m66592->ep0_buf = m66592_alloc_buffer(&m66592->ep[0].ep, 2, NULL,
- GFP_KERNEL);
+ /* AV: leaks, and do we really need it separately allocated? */
+ m66592->ep0_buf = kzalloc(2, GFP_KERNEL);
if (m66592->ep0_buf == NULL)
goto clean_up;

--
1.5.3.GIT


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/