Messages in this thread |  | | From | Oliver Neukum <> | Subject | Re: USB autosuspend vs. URB submission | Date | Thu, 10 Jan 2013 16:37:06 +0100 |
| |
On Thursday 10 January 2013 10:20:42 Alan Stern wrote: > On Thu, 10 Jan 2013, Oliver Neukum wrote:
> > In the long run it is probably a good idea to pass duplicated URBs to usbmon by > > a special code path. > > I'd prefer to add extra information to the WARN_ONCE message. Even > though it would require the extra effort of correlating the dmesg > output with the usbmon output.
A stack_trace() I presume. But what is the use of needing two logs?
> You know, it's possible that the URB really was not submitted before > but instead the urb->hcpriv field got overwritten. Of course, that > would also be a bug.
We could log a corrupted URB generically speaking.
Regards Oliver
|  |