The combination of these mods gets us to the point that Rich
Coe's and James Forris' test case now passes. The second mod
is the most critical one, but anyone who has seen or suspects
they've seen data corruption should use CVS (once it gets its
refresh from the internal tree).
BTW, I'll be away for four weeks, starting yesterday ;-), so I
wont be able to followup on any questions on this (or anything
else) until I'm back in the office.
cheers.
Do not do delalloc conversion on pages beyond EOF ever,
not just sometimes.
Date: Fri Apr 29 16:47:37 AEST 2005
Workarea: chook.melbourne.sgi.com:/build/nathans/xfs-linux
Inspected by: nathans
The following file(s) were checked into:
longdrop.melbourne.sgi.com:/isms/xfs-kern/xfs-linux-melb
Modid: xfs-linux-melb:xfs-kern:22376a
linux-2.6/xfs_aops.c - 1.85
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/linux-2.6/xfs_aops.c.diff?r1=text&tr1=1.85&r2=text&tr2=1.84&f=h
linux-2.4/xfs_aops.c - 1.88
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/linux-2.4/xfs_aops.c.diff?r1=text&tr1=1.88&r2=text&tr2=1.87&f=h
Use the right offset when ensuring a delayed allocate conversion has
covered the offset originally requested. Can cause data corruption
when multiple processes are performing writeout on different areas
of the same file. Quite difficult to hit though.
Date: Fri Apr 29 16:55:28 AEST 2005
Workarea: chook.melbourne.sgi.com:/build/nathans/xfs-linux
Inspected by: felixb@xxxxxxx
The following file(s) were checked into:
longdrop.melbourne.sgi.com:/isms/xfs-kern/xfs-linux-melb
Modid: xfs-linux-melb:xfs-kern:22377a
xfs_mount.h - 1.197
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_mount.h.diff?r1=text&tr1=1.197&r2=text&tr2=1.196&f=h
xfs_iomap.h - 1.6
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_iomap.h.diff?r1=text&tr1=1.6&r2=text&tr2=1.5&f=h
xfs_iomap.c - 1.32
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_iomap.c.diff?r1=text&tr1=1.32&r2=text&tr2=1.31&f=h
linux-2.6/xfs_aops.c - 1.86
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/linux-2.6/xfs_aops.c.diff?r1=text&tr1=1.86&r2=text&tr2=1.85&f=h
Cleanup use of loff_t vs xfs_off_t in the core code.
Date: Fri Apr 29 16:56:50 AEST 2005
Workarea: chook.melbourne.sgi.com:/build/nathans/xfs-linux
Inspected by: cattelan@xxxxxxx
The following file(s) were checked into:
longdrop.melbourne.sgi.com:/isms/xfs-kern/xfs-linux-melb
Modid: xfs-linux-melb:xfs-kern:22378a
xfs_dfrag.c - 1.46
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_dfrag.c.diff?r1=text&tr1=1.46&r2=text&tr2=1.45&f=h
xfs_mount.h - 1.198
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_mount.h.diff?r1=text&tr1=1.198&r2=text&tr2=1.197&f=h
xfs_iomap.h - 1.7
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_iomap.h.diff?r1=text&tr1=1.7&r2=text&tr2=1.6&f=h
xfs_iomap.c - 1.33
http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_iomap.c.diff?r1=text&tr1=1.33&r2=text&tr2=1.32&f=h
|