Hi, I am not sure the patch will help because the race is _very_ unlikely. But I would be glad to know if you can reproduce on latest compat-wireless. How reproducible is that on 3.3.3 ? Is that 100% ? I guess my machine doesn't have IOMMU, but honestly, I am not sure I know how to check. On Tue, May 1, 2012 at 19:52, Venkataraman, Meenakshi <meenakshi.venkataraman@xxxxxxxxx> wrote: > > Hi John, Antonio, > > Emmanuel found a race condition in rx buffer allocation/deallocation, and > attached is a patch that fixes it. > > >> > since linux-3.0 (I don't remember if it was the case even before), > >> > I'm getting this output in dmesg: > >> > > >> > [ 53.662990] ------------[ cut here ]------------ > >> > [ 53.663016] WARNING: at drivers/iommu/intel-iommu.c:2911 > >> > intel_unmap_page+0x14b/0x180() > >> > [ 53.663047] Hardware name: Latitude E5420 > >> > [ 53.663066] Driver unmaps unmatched page at PFN ffd1a000 > >> > >> > >> I see that I always get the same addr passed to the unmap function: > >> > >> [ 23.950535] Driver unmaps unmatched page at PFN ffd1a000 > >> > >> is it a coincidence? > >> I'm running linux-3.3.3 right now. I can try wireless-testing if it > >> would help. > > [MV] Antonio -- can you try out the patch attached and let us know if you > continue to see the problem? > > >Just for the record, we are seeing a similar problem on RHEL 6.3 (which > > has an > >iwlwifi driver derived from linux-3.2). > > > > https://bugzilla.redhat.com/show_bug.cgi?id=815998 > > [MV] John -- I'll update the redhat bugzilla as well. > > Thanks > Meenakshi -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
