A variety of fixes and a couple of feature type things from the irix
version of xfs.
Date: Fri Jun 16 13:57:56 PDT 2000
Workarea: jen.cray.com:/data/clink/io/lord/slinx
The following file(s) were checked into:
bonnie.engr.sgi.com:/isms/slinx/2.4.0-test1-xfs
Modid: 2.4.0-test1-xfs:slinx:64480a
cmd/xfs/mkfs/xfs_mkfs.c - 1.166
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/cmd/xfs/mkfs/xfs_mkfs.c.diff?r1=text&tr1=1.166&r2=text&tr2=1.165&f=h
- Add new "agsize=", "su=", & "sw=" options. Modify the AG alignment,
and others, to be stripe width aligned rather than stripe unit
aligned. Fix problems caused by the cvtnum() routine trying to
understand the "b" suffix before blksize has been defined. Add the
capability to pre-allocate file space via the prototype file.
cmd/xfs/repair/dir.c - 1.56
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/cmd/xfs/repair/dir.c.diff?r1=text&tr1=1.56&r2=text&tr2=1.55&f=h
- Remove assert on bad root inode. Still need to recover the
rest of the filesystem if root is bad.
linux/fs/xfs/xfs_trans_dquot.c - 1.24
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_trans_dquot.c.diff?r1=text&tr1=1.24&r2=text&tr2=1.23&f=h
- The DQ_DIRTY flag needs to be copied forward in transactions
associated
with the initial reservation. Otherwise the dquot isn't updated
latter.
linux/fs/xfs/xfs_rw.c - 1.317
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_rw.c.diff?r1=text&tr1=1.317&r2=text&tr2=1.316&f=h
linux/fs/xfs/xfs_sb.h - 1.47
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_sb.h.diff?r1=text&tr1=1.47&r2=text&tr2=1.46&f=h
- Define values needed for logging sb_uuid with xfs_mod_sb.
linux/fs/xfs/xfs_vnodeops.c - 1.456
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_vnodeops.c.diff?r1=text&tr1=1.456&r2=text&tr2=1.455&f=h
- Rearrange some #ifdef's to make the xfs_alloc_file_space() visible to
the SIM library, also involved using a couple of specific cast's to
keep the compiler happy when generating four different versions of the
SIM library.
linux/fs/xfs/xfs_qm_syscalls.c - 1.38
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_qm_syscalls.c.diff?r1=text&tr1=1.38&r2=text&tr2=1.37&f=h
- modified so as to recognize some ID's as project ID's.
Previously, if the users ID didn't match the ID for the
quota query, you'd get EPERM. In most cases, the user's
ID wont match the project ID they are querying.
linux/fs/xfs/xfs_vfsops.c - 1.271
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_vfsops.c.diff?r1=text&tr1=1.271&r2=text&tr2=1.270&f=h
linux/fs/xfs/xfs_mount.h - 1.114
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_mount.h.diff?r1=text&tr1=1.114&r2=text&tr2=1.113&f=h
- Add m_newuuid to xfs_mount_t. If m_sb.sb_uuid is updated because
its a duplicate, a copy is saved here and the uuid is restored after
xlog_do_recover overwrites it.
linux/fs/xfs/xfs_mount.c - 1.226
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_mount.c.diff?r1=text&tr1=1.226&r2=text&tr2=1.225&f=h
- When generating a new filesystem uuid, save a copy in m_newuuid.
If the filesystem is dirty, xlog_do_recover will overwrite m_sb;
restore and log m_sb.sb_uuid after xfs_log_mount has run.
sb_uuid is logged along with sb_unit & sb_width.
linux/fs/xfs/xfs_qm.c - 1.50
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_qm.c.diff?r1=text&tr1=1.50&r2=text&tr2=1.49&f=h
- fixed so that when the inode for the project dquot is
created, the superblock pointer to the user dquot inode
isn't reinitialized to -1.
The message indicating that quotas are being calculated only takes
place
when the dquot space is very large (more than 500 disk blocks). On a
file
system with many files this can take a lot of time even if the dquot
space
is small. Therefore printing the message shouldn't be dependant on
the
size of the dquot space. Since the bulkstat part of the process can
take
a long time, the Done part of the message should be printed after the
bulkstat is done. This is a fix for PV 699842
linux/fs/xfs/xfs_inode.c - 1.290
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_inode.c.diff?r1=text&tr1=1.290&r2=text&tr2=1.289&f=h
- if the inode has a non-zero project id, a version 2 inode
is written to disk. This is only the case for create. The chproj
call
already writes version 2 inodes if necessary. PV 785513
linux/fs/xfs/xfs_trans.h - 1.106
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_trans.h.diff?r1=text&tr1=1.106&r2=text&tr2=1.105&f=h
- the transaction reservation size needs to be big enough
to accomidate the superblock. If we write a version 2 inode, the
superblock version has to support version 2 inodes... we might need to
update the superblock. Therefore the transaction reservation needs to
be big enough for the superblock.
linux/fs/xfs/xfs_bmap.h - 1.73
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_bmap.h.diff?r1=text&tr1=1.73&r2=text&tr2=1.72&f=h
- Add a "prealloc" variable to the xfs_bmalloca structure.
linux/fs/xfs/xfs_bmap.c - 1.254
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/xfs_bmap.c.diff?r1=text&tr1=1.254&r2=text&tr2=1.253&f=h
- Modify xfs_bmap_alloc() to attempt stripe width alignment first.
linux/fs/xfs/linux/xfs_lrw.c - 1.43
http://oss.sgi.com/cgi-bin/cvsweb.cgi/linux-2.4-xfs/linux/fs/xfs/linux/xfs_lrw.c.diff?r1=text&tr1=1.43&r2=text&tr2=1.42&f=h
- In xfs_iomap_write(), roundup the allocation request to a stripe width
iff the existing file size is already at least one stripe width, and
we're allocating at the eof.
|