Nathan S., merge to 2.5 , thanks :)
Andrew T., if you want, please try your program with
a larger inode and with this xfs change, and see if it
speeds things up.
Eric S., I think this should be good for XFS 1.1 .
And there will be another TAKE shortly for an error
code change for ACL/EAs from E2BIG to ERANGE to
be consistent with Andreas' convention (and doc'ed in
our man page and used in the user tools).
This will be a kernel and cmd change. Need to check
all repercussions first.
--Tim
Date: Sun Apr 14 18:48:42 PDT 2002
Workarea: snort.melbourne.sgi.com:/home/diskb/build4/tes/slinx-xfs
The following file(s) were checked into:
bonnie.engr.sgi.com:/isms/slinx/2.4.x-xfs
Modid: 2.4.x-xfs:slinx:116374a
linux/fs/xfs/xfs_acl.c - 1.15
- Set the length of the EA for the ACL to be the effective
length of the ACL (count + ACEs actually used) instead of the
length of the whole array. In practice, this will mean smaller
EAs for ACLs and may allow them to fit in the inode if it is
big enough. These small ACL/EAs on XFS were tested on IRIX
by copying the FS to IRIX and worked fine.
|