Bug 2131905

Summary:	autofs: missing unlock in sasl_do_kinit_ext_cc()
Product:	Red Hat Enterprise Linux 8	Reporter:	Ian Kent <ikent>
Component:	autofs	Assignee:	Ian Kent <ikent>
Status:	CLOSED WONTFIX	QA Contact:	Kun Wang <kunwan>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	8.7	CC:	esandeen, xzhou
Target Milestone:	rc	Keywords:	Triaged
Target Release:	---	Flags:	pm-rhel: mirror+
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:
Clones:	2131907 (view as bug list)		Environment:
Last Closed:	2023-09-11 23:26:49 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:
Bug Blocks:	2131907

Description Ian Kent 2022-10-04 02:45:50 UTC

Description of problem:
A missing mutex unlock has been reported upstream.
Not sure when it crept in but it needs to be fixed.

Comment 1 Ian Kent 2022-10-05 05:05:26 UTC

(In reply to Ian Kent from comment #0)
> Description of problem:
> A missing mutex unlock has been reported upstream.
> Not sure when it crept in but it needs to be fixed.

Ok, so this is a problem in RHEL-9 but RHEL-8 doesn't have upstream
commit:

Comment 2 Ian Kent 2022-10-05 05:12:48 UTC

(In reply to Ian Kent from comment #1)
> (In reply to Ian Kent from comment #0)
> > Description of problem:
> > A missing mutex unlock has been reported upstream.
> > Not sure when it crept in but it needs to be fixed.
> 
> Ok, so this is a problem in RHEL-9 but RHEL-8 doesn't have upstream
> commit:

commit d0d99cfc0737f34af326ba3f8f35f45625300576
Author: Ian Kent <raven>
Date:   Sat Jun 6 14:17:02 2020 +0800

    autofs-5.1.6 - fix ldap sasl reconnect problem
    
    When performing an ldap sasl connection a two step initialisation
    was being done in an attempt to partially reuse existing connection
    setup.
    
    But if a network connectivity problem occurs the connection can end
    up only half initialized and recovery after connectivity is restored
    fails.
    
    So get rid of the two step initialization, as it's benefit was at best
    questionable, so that connection attempts either succeed or completely
    fail. This leaves the connection completely uninitialized if there's a
    network conectivity problem, ready for a new connection attempt.
    
    Signed-off-by: Ian Kent <raven>

which sounds useful even though we haven't had any problems reported.

I'm tempted to go ahead with this anyway ... anyone have any objections ...

At least RHEL-8 will be closer to RHEL-9, which is a good thing, and
autofs is always closer to upstream than the RHEL base version indicates
since RHEL is most often the source of bug fixes and development anyway.

Comment 9 Ian Kent 2022-12-16 07:07:47 UTC

I think it best to defer this to 8.9 if in fact we do this at all.

Comment 11 Ian Kent 2023-09-11 23:26:49 UTC

The fix here is actually for a bug introduced by a later change not present in RHEL autofs.

I was originally going to apply both patches but I think it wiser to not make these changes unless we get a report of the original problem.

Closing WONTFIX for now.