libafs: dont handle outstatus on write error

if a write errored, we can't trust the OutStatus we got. don't.

Change-Id: I74934c32426c6f25aace06f6ee59c1c1b547af90
Reviewed-on: http://gerrit.openafs.org/6714
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Reviewed-by: Alistair Ferguson <alistair.ferguson@mac.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
This commit is contained in:
Derrick Brashear 2012-02-14 21:58:09 -05:00
parent 7a650025c5
commit 4c8a3cd789

View File

@ -543,6 +543,7 @@ afs_CacheStoreDCaches(struct vcache *avc, struct dcache **dclist,
if (!code) {
code = (*ops->close)(rock, OutStatus, doProcessFS);
/* if this succeeds, dv has been bumped. */
if (*doProcessFS) {
hadd32(*anewDV, 1);
}
@ -550,6 +551,11 @@ afs_CacheStoreDCaches(struct vcache *avc, struct dcache **dclist,
}
if (ops)
code = (*ops->destroy)(&rock, code);
/* if we errored, can't trust this. */
if (code)
*doProcessFS = 0;
return code;
}