On Tue, 2010-12-07 at 16:53 -0500, Jeff Layton wrote: > On Tue, 07 Dec 2010 16:41:22 -0500 > Trond Myklebust <Trond.Myklebust@xxxxxxxxxx> wrote: > > I've been racking my brain for days now trying to remember why we clear > > the request in these cases. I think it was only about ensuring that we > > could throw the page out of the cache using invalidate_inode_pages() > > (and there is a comment about that in early 2.4.x kernels). > > > > As far as I can see, there no longer appear to be any side effects to > > deferring releasing the page or open_contexts but I'm still a bit > > nervous about doing so. > > Can we therefore please give it a good week of pounding upon before I > > pass it on to Linus? > > > > Cheers > > Trond > > > > Sounds good. I have the reporter of this problem testing a 2.6.18-based > kernel with this patch now to see whether it fixes it. We might as well > hold off until we have results from that. Their test takes several days > to run though so it may be a little while before we know whether this > helps. I'll reply once I hear from them. Looks like I was right to be sceptical. My setup hangs hard on the Connectathon test5 (read and write) when this patch is applied. Cheers Trond -- Trond Myklebust Linux NFS client maintainer NetApp Trond.Myklebust@xxxxxxxxxx www.netapp.com -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
