[Bugs] [Bug 1472961] [GNFS+EC] lock is being granted to 2 different client for the same data range at a time after performing lock acquire /release from the clients1
bugzilla at redhat.com
bugzilla at redhat.com
Tue Sep 26 10:24:42 UTC 2017
https://bugzilla.redhat.com/show_bug.cgi?id=1472961
--- Comment #6 from Worker Ant <bugzilla-bot at gluster.org> ---
COMMIT: https://review.gluster.org/17826 committed in master by Pranith Kumar
Karampuri (pkarampu at redhat.com)
------
commit 572b4bf889d903dcaed49a57a75270a763dc259d
Author: Pranith Kumar K <pkarampu at redhat.com>
Date: Fri Sep 22 12:50:43 2017 +0530
features/locks: Maintain separation of lock->client_pid, flock->l_pid
Problem:
grant_blocked_locks() constructs flock from lock. Locks xlator uses
frame->root->pid interchangeably flock->l_pid. With gNFS frame->root->pid
(which translates to lock->client_pid) is not same as flock->l_pid, this
leads
to lk's cbk returning flock with l_pid from lock->client_pid instead of
input
flock->l_pid. This triggers EC's error code path leading to failure of lk
call,
because the response' flock->l_pid is different from request's
flock->l_pid.
Fix:
Maintain separation of lock->client_pid, flock->l_pid. Always unwind with
flock with correct pid.
BUG: 1472961
Change-Id: Ifab35c458662cf0082b902f37782f8c5321d823d
Signed-off-by: Pranith Kumar K <pkarampu at redhat.com>
--
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.
More information about the Bugs
mailing list