Bug 618602

Summary:	core_pattern handler truncates parameters
Product:	Red Hat Enterprise Linux 6	Reporter:	Denys Vlasenko <dvlasenk>
Component:	kernel	Assignee:	Danny Feng <dfeng>
Status:	CLOSED ERRATA	QA Contact:	Mike Gahagan <mgahagan>
Severity:	medium	Docs Contact:
Priority:	low
Version:	6.0	CC:	jmoskovc, mnowak, mzywusko, rvokal
Target Milestone:	rc	Keywords:	RHELNAK
Target Release:	---	Flags:	mnowak: needinfo?
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:	kernel-2.6.32-83.el6	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2011-05-23 20:43:31 UTC	Type:	---
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:	614486

Description Denys Vlasenko 2010-07-27 11:06:22 UTC

Development version of abrt intended to fix bug 614486 sets /proc/sys/kernel/core_pattern to:

echo "|/usr/libexec/abrt-hook-ccpp /var/spool/abrt %s %c %p %u %g %t 11 1234567890123456789012345678901234567890" >/proc/sys/kernel/core_pattern

but on invocation, the parameters are:

Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[0]='/usr/libexec/abrt-hook-ccpp'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[1]='/var/spool/abrt'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[2]='6'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[3]='18446744073709551615'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[4]='25677'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[5]='0'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[6]='0'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[7]='1280228357'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[8]='11'
Jul 27 12:59:17 dhcp-25-227 abrt[25678]: argv[9]='123456789012345678901234567890'

Note that argv[9] is truncated.

Please raise the limit of core_pattern handler parameter length in the relevant part of kernel code, it is too small.

Comment 2 RHEL Program Management 2010-07-27 11:37:50 UTC

This issue has been proposed when we are only considering blocker
issues in the current Red Hat Enterprise Linux release.

** If you would still like this issue considered for the current
release, ask your support representative to file as a blocker on
your behalf. Otherwise ask that it be considered for the next
Red Hat Enterprise Linux release. **

Comment 8 RHEL Program Management 2010-11-11 01:39:29 UTC

This request was evaluated by Red Hat Product Management for inclusion
in a Red Hat Enterprise Linux maintenance release. Product Management has 
requested further review of this request by Red Hat Engineering, for potential
inclusion in a Red Hat Enterprise Linux Update release for currently deployed 
products. This request is not yet committed for inclusion in an Update release.

Comment 9 Aristeu Rozanski 2010-11-17 19:47:22 UTC

Patch(es) available on kernel-2.6.32-83.el6

Comment 11 Denys Vlasenko 2011-01-06 14:03:40 UTC

Testing the fix for bug 614486 with kernel-2.6.32-83.el6.

I set /proc/sys/kernel/core_pattern to:

echo really_long_long_long_long_long_long_long_long_corename_s%s_c%c_p%p_u%u_g%g_t%t_h%h_e%e >/proc/sys/kernel/core_pattern

When abrt starts, it sets core_pattern to:

|/usr/libexec/abrt-hook-ccpp /var/spool/abrt %s %c %p %u %g %t %h %e 7265616c6c795f6c6f6e675f6c6f6e675f6c6f6e675f6c6f6e675f6c6f6e675f6c6f6e675f6c6f6e675f6c6f6e675f636f72656e616d655f7325735f6325635f7025705f7525755f6725675f7425745f6825685f652565

The long hex string is a hex-encoded "really_long_long_long_long_long_long_long_long_corename_s%s_c%c_p%p_u%u_g%g_t%t_h%h_e%e"

But only "|/usr/libexec/abrt-hook-ccpp /var/spool/abrt %s %c %p %u %g %t %h %e 7265616c6c795f6c6f6e675f6c6f6e675f6c6f6e675f6c6f6e675f6c6f" makes it (127 chars).

As a result, abrt-hook-ccpp saves coredump under truncated name "really_long_long_long_long_lo"

I'd say that for most real-world cases it will work, but raising the 128 char limit would be appreciated.

Comment 13 errata-xmlrpc 2011-05-23 20:43:31 UTC

An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHSA-2011-0542.html

Comment 14 Michal Nowak 2011-07-28 12:51:31 UTC

We have a test case for this issue but I not sure it's correct. What are the expectations on core_pattern here to be?

Here comes the important lines from the test:

    rlPhaseStartSetup
        rlServiceStop abrt-ccpp
        core_pattern_bkp="$(cat /proc/sys/kernel/core_pattern)"
        rlLog "Former core_pattern was: $core_pattern_bkp"
    rlPhaseEnd

    rlPhaseStartTest
        rlRun "echo really_long_long_long_long_long_long_long_long_corename_s%s_c%c_p%p_u%u_g%g_t%t_h%h_e%e > /proc/sys/kernel/core_pattern" 0 "Set core_pattern to long garbage"
        rlServiceStart abrtd
        chars_in_core_pattern="$[$(cat /proc/sys/kernel/core_pattern | wc -c) - 1]"
        rlAssertGreaterOrEqual "core_pattern length should be >= 127" $chars_in_core_pattern 127
    rlPhaseEnd

    rlPhaseStartCleanup
        rlServiceRestore abrt-ccpp
        rlLog "core_pattern is $(cat /proc/sys/kernel/core_pattern)"
    rlPhaseEnd


What I know is wrong is that core_patter length was not extended beyond 128 chars.

How should be the test written?