perl-devel July 2006

perl-devel@lists.fedoraproject.org

4 participants
47 discussions

Start a nNew thread

[Bug 175513] New: UTF-8 error from sa-learn
by Red Hat Bugzilla 19 Jan '07

19 Jan '07

1 7

[Bug 187974] New: selinux denials of spamd reading files
by Red Hat Bugzilla 20 Dec '06

20 Dec '06

Please do not reply directly to this email. All additional comments should be made in the comments box of this bug report. https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=187974 Summary: selinux denials of spamd reading files Product: Fedora Core Version: fc5 Platform: All OS/Version: Linux Status: NEW Severity: normal Priority: normal Component: spamassassin AssignedTo: wtogami(a)redhat.com ReportedBy: dbaron(a)dbaron.org CC: fedora-perl-devel- list@redhat.com,felicity@kluge.net,jm@jmason.org,parkerm @pobox.com,rcoker@redhat.com,reg+redhat@sidney.com,wtoga mi(a)redhat.com Description of problem: with the recent selinux and spamassassin updates to FC5 (which I picked up at the same time last week), there have started to be selinux denials of spamd, three at a time, when spamd starts: type=AVC msg=audit(1144179464.345:5): avc: denied { search } for pid=1768 comm="spamd" name="lib" dev=hda3 ino=423490 scontext=system_u:system_r:spamd_t:s0 tcontext=system_u:object_r:var_lib_t:s0 tclass=dir type=SYSCALL msg=audit(1144179464.345:5): arch=40000003 syscall=195 success=no exit=-13 a0=97843b0 a1=93dd0c8 a2=9bfff4 a3=97843b0 items=1 pid=1768 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 comm="spamd" exe="/usr/bin/perl" type=CWD msg=audit(1144179464.345:5): cwd="/" type=PATH msg=audit(1144179464.345:5): item=0 name="/var/lib/spamassassin/3.001001" flags=1 type=AVC msg=audit(1144179464.753:6): avc: denied { search } for pid=1768 comm="spamd" name="lib" dev=hda3 ino=423490 scontext=system_u:system_r:spamd_t:s0 tcontext=system_u:object_r:var_lib_t:s0 tclass=dir type=SYSCALL msg=audit(1144179464.753:6): arch=40000003 syscall=195 success=no exit=-13 a0=97843b0 a1=93dd0c8 a2=9bfff4 a3=97843b0 items=1 pid=1768 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 comm="spamd" exe="/usr/bin/perl" type=CWD msg=audit(1144179464.753:6): cwd="/" type=PATH msg=audit(1144179464.753:6): item=0 name="/var/lib/spamassassin/3.001001/languages" flags=101 type=AVC msg=audit(1144179466.234:7): avc: denied { search } for pid=1768 comm="spamd" name="lib" dev=hda3 ino=423490 scontext=system_u:system_r:spamd_t:s0 tcontext=system_u:object_r:var_lib_t:s0 tclass=dir type=SYSCALL msg=audit(1144179466.234:7): arch=40000003 syscall=195 success=no exit=-13 a0=97843b0 a1=93dd0c8 a2=9bfff4 a3=97843b0 items=1 pid=1768 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 comm="spamd" exe="/usr/bin/perl" type=CWD msg=audit(1144179466.234:7): cwd="/" type=PATH msg=audit(1144179466.234:7): item=0 name="/var/lib/spamassassin/3.001001/triplets.txt" flags=1 I'm not sure what this effects, but having selinux prevent spamd from doing things seems like it could break something. Version-Release number of selected component (if applicable): spamassassin-3.1.1-1.fc5 selinux-policy-2.2.25-3.fc5 selinux-policy-targeted-2.2.25-3.fc5 How reproducible: Always (when spamd starts/restarts). Steps to Reproduce: 1. tail -f /var/log/audit.log 2. /sbin/service spamassassin restart Actual results: selinux denials Expected results: no selinux denials Additional information: As a note, the directory /var/lib/spamassassin/ does not exist. And the files in question live in /usr/share/spamassassin/ ... which is why I'm filing this as a bug on spamassassin rather than selinux-policy-targeted. -- Configure bugmail: https://bugzilla.redhat.com/bugzilla/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.

1 19

[Bug 198461] New: Package spamassassin lacks IPv6 support
by Red Hat Bugzilla 14 Dec '06

14 Dec '06

1 3

[Bug 173646] New: Selinux denials for spamd
by Red Hat Bugzilla 14 Dec '06

14 Dec '06

1 1

[Bug 199285] New: Replace hardcoded .fc6 by %{?dist}
by Red Hat Bugzilla 11 Nov '06

11 Nov '06

1 3

[Bug 199286] New: Replace hardcoded .fc6 by %{?dist}
by Red Hat Bugzilla 10 Nov '06

10 Nov '06

1 2

[Bug 193818] New: spamd should start before sendmail
by Red Hat Bugzilla 08 Nov '06

08 Nov '06

1 7

[Bug 175439] New: [FC4 Regression]: spamc doesn't use localhost by default
by Red Hat Bugzilla 29 Oct '06

29 Oct '06

1 14

[Bug 200440] New: Building perl-5.8.8-8 without threading fails
by Red Hat Bugzilla 04 Oct '06

04 Oct '06

1 2

[Bug 191416] New: h2ph generates incorrect code for '#if defined A || defined B'
by Red Hat Bugzilla 04 Oct '06

04 Oct '06

Please do not reply directly to this email. All additional comments should be made in the comments box of this bug report. https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=191416 Summary: h2ph generates incorrect code for '#if defined A || defined B' Product: Fedora Core Version: devel Platform: All URL: http://rt.perl.org/rt3/Ticket/Display.html?id=39130 OS/Version: Linux Status: NEW Severity: normal Priority: normal Component: perl AssignedTo: jvdias(a)redhat.com ReportedBy: jvdias(a)redhat.com QAContact: dkl(a)redhat.com CC: fedora-perl-devel-list@redhat.com,prockai@redhat.com +++ This bug was initially created as a clone of Bug #191409 +++ Description of problem: This is perlbug #39130 - see URL . For cpp statements like : #if defined A || defined B h2ph would generate the code : if(defined (defined(&A) ? &A : 0) || defined (defined(&B) ? &B : 0)) { which is tautologous, as defined(0) is always true . This turns out to cause real problems on the Linux ppc64 platform, where "endian-ness" is selectable in the compiler; every inclusion of "endian.ph", which includes "bits/endian.ph", generates an error: $ perl -e 'require "sys/socket.ph";' Both BIG_ENDIAN and LITTLE_ENDIAN defined! at /usr/lib/perl5/5.8.8/ppc-linux-thread-multi/bits/endian.ph line 10. Compilation failed in require at... Because h2ph generated this code in ../bits/endian.ph: if(defined (defined(&__BIG_ENDIAN__) ? &__BIG_ENDIAN__ : 0) || defined (defined(&_BIG_ENDIAN) ? &_BIG_ENDIAN : 0)) { if(defined (defined(&__LITTLE_ENDIAN__) ? &__LITTLE_ENDIAN__ : 0) || defined (defined(&_LITTLE_ENDIAN) ? &_LITTLE_ENDIAN : 0)) { die("Both\ BIG_ENDIAN\ and\ LITTLE_ENDIAN\ defined\!"); } For the cpp code in /usr/include/bits/endian.h: #if defined __BIG_ENDIAN__ || defined _BIG_ENDIAN # if defined __LITTLE_ENDIAN__ || defined _LITTLE_ENDIAN # error Both BIG_ENDIAN and LITTLE_ENDIAN defined! ... The problem does not happen if the 'defined..' conditions are in parentheses - ie. #if defined(A) || defined(B) produces correct code: if( defined(&A) || defined(&B) ) A trivial fix for this is to simply replace the ': 0' with ': undef' in h2ph.PL line 517, as with this patch: --- perl-5.8.8/utils/h2ph.PL~ 2006-01-12 17:55:04.000000000 -0500 +++ perl-5.8.8/utils/h2ph.PL 2006-05-11 13:50:04.000000000 -0400 @@ -514,7 +514,7 @@ } } else { if ($inif && $new !~ /defined\s*\($/) { - $new .= '(defined(&' . $id . ') ? &' . $id . ' : 0)'; + $new .= '(defined(&' . $id . ') ? &' . $id . ' : undef)'; } elsif (/^\[/) { $new .= " \$$id"; } else { Version-Release number of selected component (if applicable): ALL How reproducible: 100% Steps to Reproduce: On a ppc64 platform with select-able endian-ness: # perl -e 'require "sys/socket.ph";' Actual results: Both BIG_ENDIAN and LITTLE_ENDIAN defined! at /usr/lib/perl5/5.8.8/ppc-linux-thread-multi/bits/endian.ph line 10. Compilation failed in require at... Expected results: no error -- Configure bugmail: https://bugzilla.redhat.com/bugzilla/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.

1 5

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

perl-devel July 2006