[U-Boot] [PATCH] USB-CDC: wrong ep status used
In case a status ep is requested, it is always allocated a request for the ep0, instead of the correct one saved in the dev structure.
Signed-off-by: Stefano Babic sbabic@denx.de --- drivers/usb/gadget/ether.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/gadget/ether.c b/drivers/usb/gadget/ether.c index 9f9b093..1481d76 100644 --- a/drivers/usb/gadget/ether.c +++ b/drivers/usb/gadget/ether.c @@ -1731,14 +1731,14 @@ autoconf_fail: /* ... and maybe likewise for status transfer */ #if defined(DEV_CONFIG_CDC) if (dev->status_ep) { - dev->stat_req = usb_ep_alloc_request(gadget->ep0, GFP_KERNEL); - dev->stat_req->buf = status_req; + dev->stat_req = usb_ep_alloc_request(dev->status_ep, GFP_KERNEL); if (!dev->stat_req) { dev->stat_req->buf=NULL; - usb_ep_free_request (gadget->ep0, dev->req); + usb_ep_free_request (dev->status_ep, dev->req);
goto fail; } + dev->stat_req->buf = status_req; dev->stat_req->context = NULL; } #endif
Hi Stefano,
On 08/12/2010 01:41 AM, Stefano Babic wrote:
#if defined(DEV_CONFIG_CDC) if (dev->status_ep) {
dev->stat_req = usb_ep_alloc_request(gadget->ep0, GFP_KERNEL);dev->stat_req->buf = status_req;
dev->stat_req = usb_ep_alloc_request(dev->status_ep, GFP_KERNEL);
We get oops here!
usb_ep_free_request (gadget->ep0, dev->req);
usb_ep_free_request (dev->status_ep, dev->req); goto fail;}
dev->stat_req->buf = status_req;dev->stat_req->context = NULL; }
#endif
Vitaly Kuzmichev wrote:
Hi Stefano,
On 08/12/2010 01:41 AM, Stefano Babic wrote:
#if defined(DEV_CONFIG_CDC) if (dev->status_ep) {
dev->stat_req = usb_ep_alloc_request(gadget->ep0, GFP_KERNEL);dev->stat_req->buf = status_req;
dev->stat_req = usb_ep_alloc_request(dev->status_ep, GFP_KERNEL);We get oops here!
Agree, and the issue is not related to this patch, I missed to correct it, thanks. If no one complains, I will send a single patch to fix both problems (wrong ep status + null pointer access).
Best regards, Stefano
Hello.
Stefano Babic wrote:
On 08/12/2010 01:41 AM, Stefano Babic wrote:
#if defined(DEV_CONFIG_CDC) if (dev->status_ep) {
dev->stat_req = usb_ep_alloc_request(gadget->ep0, GFP_KERNEL);dev->stat_req->buf = status_req;
dev->stat_req = usb_ep_alloc_request(dev->status_ep, GFP_KERNEL);We get oops here!
Agree, and the issue is not related to this patch, I missed to correct it, thanks. If no one complains, I will send a single patch to fix both problems (wrong ep status + null pointer access).
Looks like Vitaly has done the same already:
http://lists.denx.de/pipermail/u-boot/2010-August/075419.html
Best regards, Stefano
WBR, Sergei
Hi Sergei,
On 08/12/2010 09:05 PM, Sergei Shtylyov wrote:
Hello.
Stefano Babic wrote:
We get oops here!
Agree, and the issue is not related to this patch, I missed to correct it, thanks. If no one complains, I will send a single patch to fix both problems (wrong ep status + null pointer access).
Looks like Vitaly has done the same already:
Yes, but I have not seen Stefano's comment when sent.
Stefano, I would like to add you 'signed-off' in this patch because I took this part from your patch: - usb_ep_free_request (gadget->ep0, dev->req); + usb_ep_free_request (dev->status_ep, dev->req);
Do you allow me to add your signature?
Vitaly Kuzmichev wrote:
Hi Sergei,
On 08/12/2010 09:05 PM, Sergei Shtylyov wrote:
Hello.
Stefano Babic wrote:
We get oops here!
Agree, and the issue is not related to this patch, I missed to correct it, thanks. If no one complains, I will send a single patch to fix both problems (wrong ep status + null pointer access).
Looks like Vitaly has done the same already:
Yes, but I have not seen Stefano's comment when sent.
Stefano, I would like to add you 'signed-off' in this patch because I took this part from your patch:
usb_ep_free_request (gadget->ep0, dev->req);
usb_ep_free_request (dev->status_ep, dev->req);Do you allow me to add your signature?
Yes, of course !
Stefano
participants (3)
-
Sergei Shtylyov -
Stefano Babic -
Vitaly Kuzmichev