Bug 5878 – initscripts hosed on Alpha/Jensen

Bugzilla will be upgraded to version 5.0 on a still to be determined date in the near future. The original upgrade date has been delayed.

Bug 5878 - initscripts hosed on Alpha/Jensen

Summary:	initscripts hosed on Alpha/Jensen

Status:	CLOSED DUPLICATE of bug 5946

Aliases:	None

Product:	Red Hat Linux
Classification:	Retired
Component:	initscripts (Show other bugs)
Sub Component:
Version:	6.0
Hardware:	alpha Linux

Priority	high Severity high
Target Milestone:	---
Target Release:	---
Assigned To:	Bill Nottingham
QA Contact:
Docs Contact:

URL:
Whiteboard:
Keywords:

Depends On:
Blocks:
	Show dependency tree / graph

Reported:	1999-10-12 11:25 EDT by mchampig
Modified:	2014-03-16 22:10 EDT (History)
CC List:	1 user (show)

See Also:
Fixed In Version:
Doc Type:	Bug Fix
Doc Text:
Story Points:	---
Clone Of:
Environment:
Last Closed:	1999-10-14 11:22:32 EDT
Type:	---
Regression:	---
Mount Type:	---
Documentation:	---
CRM:
Verified Versions:
Category:	---
oVirt Team:	---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---

Attachments	(Terms of Use)
Add an attachment (proposed patch, testcase, etc.)

Groups: None (edit)

Description mchampig 1999-10-12 11:25:17 EDT

The daemon() and killproc() functions defined in the
the file "functions" from initscripts is causing hangs
at bootup on Alpha/Jensen. I am looking into this. As
a horrible workaround, I've just returned immediately from
these function and started/stopped system services by
hand. Yuck! I'll try to narrow down the problem. I don't
know why this just happens on Jensen since my other Alpha
boxes work fine with these functions. Might have something
to do with the generic kernel for RH6.0 not working on
Jensen. I'll have more details soon to help you resolve
this since I know you don't have a Jensen box.

Comment 1 Bill Nottingham 1999-10-12 11:50:59 EDT

Does the Jensen have some sort of weird console?
Out of curiousity, if you try the latest initscripts from Raw Hide
(unfortunately, this might pull in some other requires), does
that fare better?

Comment 2 Bill Nottingham 1999-10-12 11:54:59 EDT

One thing that should narrow the hang down fairly quickly is to
add 'set -x' to the beginning of daemon() & killproc().

Comment 3 mchampig 1999-10-13 14:07:59 EDT

I've narrowed the problem down. There is nothing wrong with daemon()
or killproc() per se. It's the call to pidofproc() in each that is
causing the hang. In particular, the call to awk is hanging. Sure
enough, I tried running awk alone from the command line. It hangs
every time on Jensen. :-( Looks like it's broke...maybe a compiler
bug in EV4 machines? Hmmm. Anyway, several options here...you could
use grep/cut/sed/perl instead of awk to parse the field in ps output.
Better yet, why not just dump the calls to awk? You are already
checking for the process id with pidof. There is really no reason
to be redundant. I'm sure pidof takes care of checking the process
table better than searching the output of "ps ax".

A workaround for Jensen owners is to return "0" just before the
awk invokation. More seriously, awk is broken on Jensen. Maybe
building without optimizations could fix it.

Comment 4 Bill Nottingham 1999-10-13 14:11:59 EDT

Out of curiosity, if you do 'strace awk', what exactly
*is* it doing?

Does the awk currently in Raw Hide (gawk-3.0.4-1)
work any better?  If awk is broken, more things are
going to break, probably in bizarre ways.

Comment 5 Bill Nottingham 1999-10-14 11:22:59 EDT

*** This bug has been marked as a duplicate of 5946 ***

Note You need to log in before you can comment on or make changes to this bug.