Coleman Kane
2008-Apr-26 05:20 UTC
MFC Candidate: convert ffs_softdep.c over to callout(9)
Skipped content of type multipart/mixed-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20080426/9b056217/attachment.pgp
Ilya Bakulin
2008-Apr-26 13:57 UTC
MFC Candidate: convert ffs_softdep.c over to callout(9)
On Sat, 26 Apr 2008 01:19:01 -0400 Coleman Kane <cokane@FreeBSD.org> wrote: Hi, I have an ability to test this patch on RELENG_7 system (there is no mission-critical data stored on HDD :-) ) What sould I do? Just apply patch, recompile kernel and produce some disk load on that system?> Hi stable users, > > Recently I committed revision 1.219 of src/sys/ufs/ffs/ffs_softdep.c to > remove the use of old-style timeout(9) calls in the softdep code and > replace them with the new callout(9) API, to further MPSAFE-ness of the > softdep code. > > I am attaching a patch to RELENG_7 that I'd like some people to test, > which is just an MFC of this code. Can I get any testers (so that I can > MFC this)? > > -- > Coleman Kane >-- Ilya Bakulin <webmaster@kibab.com>
Coleman Kane
2008-Apr-28 21:35 UTC
MFC Candidate: convert ffs_softdep.c over to callout(9)
Ilya Bakulin spake thusly:> ? > Hi, > I have an ability to test this patch on RELENG_7 system (there is no mission-critical data stored on HDD :-) ) > What sould I do? Just apply patch, recompile kernel and produce some disk load on that system? > > > Hi stable users, > > > > Recently I committed revision 1.219 of src/sys/ufs/ffs/ffs_softdep.c to > > remove the use of old-style timeout(9) calls in the softdep code and > > replace them with the new callout(9) API, to further MPSAFE-ness of the > > softdep code. > > > > I am attaching a patch to RELENG_7 that I'd like some people to test, > > which is just an MFC of this code. Can I get any testers (so that I can > > MFC this)? > >Ilya, Yes just go ahead and apply the patch that was attached. Then rebuild your kernel. Make sure that your softupdates are enabled on the UFS partition(s). Then try hammering it with some sort of file test on that partition (e.g. unpack a somewhat large source tarball and try building it). The kernel did build fine for me when I built it. Unfortunately, due to my hardware, RELENG_7 doesn't "just work" for me from vanilla sources, so I'd like to see some testers on better "supported" hardware. -- Coleman Kane
Ilya Bakulin
2008-May-06 10:25 UTC
MFC Candidate: convert ffs_softdep.c over to callout(9)
On Mon, 28 Apr 2008 17:17:47 -0400 Coleman Kane <cokane@cokane.org> wrote:> Ilya, > > Yes just go ahead and apply the patch that was attached. Then rebuild > your kernel. Make sure that your softupdates are enabled on the UFS > partition(s). > > Then try hammering it with some sort of file test on that partition > (e.g. unpack a somewhat large source tarball and try building it). > > The kernel did build fine for me when I built it. Unfortunately, due to > my hardware, RELENG_7 doesn't "just work" for me from vanilla sources, > so I'd like to see some testers on better "supported" hardware. > > -- > Coleman KaneSo, everything is working pretty good. Kernel build succeeded today on machine running patched version of ffs_softdep and GENERIC kernel. ==================================================kibab@media%uname -a FreeBSD media.home 7.0-STABLE FreeBSD 7.0-STABLE #0: Tue May 6 10:56:39 MSD 2008 root@kibab-main.home:/usr/_OBJDIR/usr/src/sys/GENERIC i386 kibab@media%mount /dev/ad0s1a on / (ufs, local) devfs on /dev (devfs, local) /dev/ad0s1d on /tmp (ufs, local, soft-updates) /dev/ad0s1f on /usr (ufs, NFS exported, local, soft-updates) /dev/ad0s1e on /var (ufs, local, soft-updates) ================================================== -- Ilya Bakulin <webmaster@kibab.com>