Mickey, if you can get it working, and if some reasonable amount of testing has happened on it, we can include it in the glfs patched fuse. thanks, avati 2007/12/10, Mickey Mazarick <mic@xxxxxxxxxxxxxxxxxx>: > > I haven't tested this with tcp yet either, but I'm pretty sure it's fuse > that's preventing mmap from working fully. > I've turned direct-io off, but mmap functions still throw errors. I did > attempt to side-step the problem by creating a loop device and mounting > it but that didn't work for the same reasons. > > It's my understanding that fuse doesn't support mmap writing fully. > There is a questionable patch out there: > http://kerneltrap.org/mailarchive/linux-fsdevel/2007/11/15/417491 > > Is this something that's worth trying add to the patched glusterfs fuse > module? I was going to take a stab at it but wanted a second opinion > because my C skills are now half a decade rusty. > > > Thanks! > -- > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxx > http://lists.nongnu.org/mailman/listinfo/gluster-devel > -- It always takes longer than you expect, even when you take into account Hofstadter's Law. -- Hofstadter's Law
