On Fri, Mar 5, 2010 at 1:07 PM, <[email protected]> wrote: > > Joonwoo Park: >> > Try setting udba=none. >> > In udba=reval (default), aufs calls ubifs_getattr() and gets the correct >> > value. >> >> Is it possible to use udba=none in this case? >> I was under the impression if the branch partition (in this case >> ubifs) can be changed directly, I had to set udba as not none (I've >> used inotify) >> Please correct me if I'm wrong. > > Correct. > The udba=none is just for testing your patch. > With udba=reval or udba=inotify, aufs always returns the correct > ubifs i_blocks value. Since you want to see how your patch changed the > behaviour, you need to set udba=none. > With udba=none and without your patch, aufs returns the incorrect value > of ubifs. Applying your patch, even if you specify udba=none, you can > see the correct value. > > >> Also I was *not* able to access this file if I try to open this file >> through aufs. it complains no such file or directory even though I >> can 'ls' & 'stat' the file through aufs and ubifs both. (I'm able to >> open file through ubifs though) > ::: >> Does this mean there is another issue beside au_test_fs_bad_iattr_size()? > > I guess so. > Did you correctly build the aufs module? > Haven't you succeeded such operation before?
I double checked but looks everything okay on my build. I cannot answer that question because I've always used aufs with udba=inotify. However with udba=inotify those operations were successful. Thanks, Joonwoo. > > > J. R. Okajima > ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev
