Dale Ghent wrote:
On Aug 13, 2004, at 11:14 AM, Douglas E. Engert wrote:
Do you have a diff of the old and new header files? I have patch 112233-12 on my Solaris 9 workstation, but the AFS servers are running with 112233-11 The files are identical.
You know, this got me thinking, and did some digging around.
I just went through all of the Solaris 9 UFS and kernel patches, and discovered two patches, 113457 and 113073, which are the only patches which change /usr/include/sys/fs/ufs_fs.h.
113457 seems to have been withdrawn and superseded by 113073, which is a member of the Solaris 9 recommended patch cluster. This 113073 patch is a comprehensive UFS and LVM patch.
See, here our Solaris 9 boxes have S9 4/04 installed on them. The Solaris 9 box exhibiting the problems had the recommended cluster applied, which included 113073-13. We have another Solaris 9 box which is running AFS fine which has 113073-05 installed.
So it seems that somewhere after at least rev 05 of patch 113073 is when sun "broke" UFS as far as what AFS expects, and the kernel patches have no part in this.
Reading through the release history for the 113073 patch, nothing immediately sticks out as the cause of why 1) fs_interleave was removed from ufs_fs.h, and 2) why vfsck now has issues with block 0 of a UFS file system.
If they removed a field from the structure, all the following fields will be off by 4 bytes. any reference to these will cause problems. Do you have a diff of the older file, which I have and the one you have?
Can you check your S9 system for me and tell me what rev of 113073 you have installed, if it is installed?
according to patchdiag both the workstation and AFS server have 113073-05 (13 is listed as current) and called "ufs and fsck patch"
/dale
--
Douglas E. Engert <[EMAIL PROTECTED]> Argonne National Laboratory 9700 South Cass Avenue Argonne, Illinois 60439 (630) 252-5444 _______________________________________________ OpenAFS-info mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-info
