Le 13.11.2005 13:18, Laurent Riffard a écrit : > Le 13.11.2005 01:55, Artur Makówka a écrit : > >>Hans Reiser wrote: >> >> >>>John Gilmore wrote: >>> > > [snip] > >>>> >>>>Oh, BTW. "The slowdown" as I called it is still there. I guess I >>>>spoke to soon. The specific symptom is that the effected process >>>>locks for a time, usually just a second or two, but sometimes a >>>>minute or two and and at least once for many many minutes. I think >>>>that the crash (soft lockup) that I reported earlier is related as >>>>well. And it sounds like the comment that rvalles had about lockups >>>>with mmaped files, except that it doesn't lock up permanently. Just >>>>for a second or three usually. >>>> >>>> >>>> >>> >>>zam, please comment. >>> >> >>one more thing im pretty sure of - the 2.6.13mm3 without any reiser4 >>additional patches (just clean 2.6.13mm3 as it has reiser4 already >>built) is working fine. >> >>i mean, im not sure if this bug still exists here, but im 100% sure i >>can write vim files easy without any downtime, so this is big difference. >> >>so for everyone with this bug - try clean 2.6.13mm3 from kernel.org. It >>worked for me. i will wait with this kernel for a patch to stable line. >> >>Also - i will test it a little more tomorrow to be sure that this >>version is bug free. >> > > > Hello, > > It seems that fsync could be really slow when there is a lot of > activity on the FS. > > I've done the following test : > - on a reiser4 FS, unpack a fresh kernel tarball and start a compilation > - open a new terminal, wait 1 minute and start editing a file on the > same FS with vi. > - hit ":w" and wait... > > $ strace -T vi foo 2>strace_vi.log > $ grep fsync strace_vi.log > fsync(3) = 0 <30.923808> > $ > > My computer is a desktop with an Athlon XP 1600 and 512 Mb RAM. > Write cache is disabled on the disk (hdparm -W0 /dev/hda). >
I forgot to mention: kernel version is 2.6.14-mm2. -- laurent
