Messages in this thread | | | Date | Sun, 15 Jun 2008 15:31:45 -0700 | From | Arjan van de Ven <> | Subject | Re: [GIT]: Networking |
| |
On Sat, 14 Jun 2008 16:24:30 -0700 (PDT) David Miller <davem@davemloft.net> wrote:
> From: Arjan van de Ven <arjan@infradead.org> > Date: Sat, 14 Jun 2008 02:39:16 -0700 > > > the interesting thing is that there is a clear bias on which drivers > > are tripping this (r8169 is top when I counted yesterday, with > > sis900 second and then a long tail of nothing) that I think the > > WARN_ON() is useful in addition to the always-there printk. Eg it > > does help in seeing which driver is most likely to trigger this. > > (Andrew also thought this would trigger a *LOT*, so far it's only a > > rather modest amount, but it's waiting for Fedora or others to ship > > a kernel with this in to be sure) > > I agree that it's useful and should stay. > > One thing I noticed is that you can't tell which driver is > to blame just from the warning and backtrace. Somehow getting > a driver name in that warning message would be useful and help > diagnose problems as well as make it easier for you to compute > those statistics.
There's a patch in -mm to add a WARN_ON variant that takes printk like arguments... once that goes to mainline that'd be the thing to use here...
-- If you want to reach me at my work email, use arjan@linux.intel.com For development, discussion and tips for power savings, visit http://www.lesswatts.org
| |