Hey, here are the fixups for the abort issues we have been talking about.
Well, almost all of them. The sync exch_done issues is not handled,
becuase Vasu was going to rework his patch.
The following patch were made over this last commit:
commit c3a9998682299fc52d3d1478d3fbbab3db7ada6c
Author: Vasu Dev <[EMAIL PROTECTED]>
Date: Fri Aug 8 15:04:22 2008 -0700
libfc: removed checks for timer_pending and then updating exch ref count.
The first 7 patches:
0001-libfc-use-spin_lock_bh-for-scsi-pkt-lock.patch
0002-libfc-fix-return-value-from-fc_seq_exch_abort.patch
0003-libfc-remove-exrta-hold-release-sequences.patch
0004-libfc-don-t-pass-wait_for_completion_timeout-msecs.patch
0005-libfc-don-t-process-frames-when-aborting-commands.patch
0006-libfc-don-t-WARN_ON-if-fc_seq_start_next-and-fc_seq.patch
0007-libfc-fix-RRQ-handling.patch
should be pretty safe at least compared to the 8th one.
Maybe the one exception is
0005-libfc-don-t-process-frames-when-aborting-commands.patch
because I am not sure if I am reading the SPEC right. The current
code definately does not look right though, because of how we
drop some data and not other bits.
0008-libfc-have-fc_exch-notify-fc_fcp-of-abort-responses.patch needs
more testing. I have not been able to get the target to return
a abts response, so the code is not well tested in that path.
_______________________________________________
devel mailing list
[email protected]
http://www.open-fcoe.org/mailman/listinfo/devel
