Bug 809917

Summary: s390x strace problems with 31-bit executables
Product: Red Hat Enterprise Linux 6 Reporter: Jeff Law <law>
Component: straceAssignee: Jeff Law <law>
Status: CLOSED ERRATA QA Contact: qe-baseos-tools-bugs
Severity: medium Docs Contact:
Priority: low    
Version: 6.2CC: bgollahe, dgregor, dhorak, dsmith, fche, law, mcermak, mgracik, mnowak, mpetlan, mpolacek, notting, roland, tlavigne
Target Milestone: rc   
Target Release: ---   
Hardware: s390x   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Previously, strace on s390x architecture couldn't handle the 31-bit binaries correctly. This has been fixed so that now there is a strace32 binary, which correctly handles 31-bit binaries even on s390x architecture.
Story Points: ---
Clone Of: 495935
: 867021 (view as bug list) Environment:
Last Closed: 2013-02-21 07:51:22 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: 495935    
Bug Blocks: 840528, 867021    

Comment 7 Bill Nottingham 2012-10-15 20:38:58 UTC
Ah, never mind - see it's built as a separate subpackage on the compat arch.

On RHEL 5, we have:
       <packagereq arch="ppc" type="mandatory">strace64</packagereq>
We could do the same for power & z on 6 and 7.

diff -u -u -r1.212 comps-rhel7.xml.in
--- comps-rhel7.xml.in  4 Oct 2012 20:14:33 -0000       1.212
+++ comps-rhel7.xml.in  15 Oct 2012 20:38:13 -0000
@@ -217,6 +217,8 @@
       <packagereq type="default">smartmontools</packagereq>
       <packagereq type="default">sos</packagereq>
       <packagereq type="default">strace</packagereq>
+      <packagereq type="default" arch="ppc64">strace32</packagereq>
+      <packagereq type="default" arch="s390x">strace32</packagereq>
       <packagereq type="default">subscription-manager</packagereq>
       <packagereq type="default">sysstat</packagereq>
       <packagereq type="default">systemtap-runtime</packagereq>
@@ -537,6 +539,8 @@
       <packagereq type="mandatory">latrace</packagereq>
       <packagereq type="mandatory">libreport-cli</packagereq>
       <packagereq type="mandatory">strace</packagereq>
+      <packagereq type="mandatory" arch="ppc64">strace32</packagereq>
+      <packagereq type="mandatory" arch="s390x">strace32</packagereq>
       <packagereq type="mandatory">systemtap-runtime</packagereq>
       <packagereq type="default">abrt-addon-ccpp</packagereq>
       <packagereq type="default">abrt-addon-python</packagereq>

as an example.

Comment 8 Michael Petlan 2013-01-18 12:41:10 UTC
Hi,

I am creating a test plan for this Errata and it seems to me, that this bug and the "semtimedop" one ( https://bugzilla.redhat.com/show_bug.cgi?id=759566 ) are based on the same problem. Am I right? Can I test it the same way?

OLD = 4.5.19-1.11.el6_3.2
NEW = 4.5.19-1.17.el6

so I run the strace, both OLD and NEW on:

1) semtimedop_app compiled with "-m31" and both fails:

semtimedop(4393752297474, 0x3ff7fdcc1ca, 4393751543809, {...}) = -1 EAGAIN (Resource temporarily unavailable)

2) semtimedop_app compiled normally, 64bit and both are OK:

semtimedop(688130, {{0, -1, 0}}, 1, {0, 0}) = -1 EAGAIN (Resource temporarily unavailable)

-----

Then I run the strace32 (NEW, because OLD does not exist) on semtimedop_app compiled with "-m31" and it runs well:
semtimedop(851970, {{0, -1, 0}}, 1, {0, 0}) = -1 EAGAIN (Resource temporarily unavailable)

-----

So it seems to me, that the strace32 works well on 31 bit apps with semtimedop and also others syscalls and the 64bit strace does not, but it is planned and expectable. So I should test somehow that the strace32 is present in the update and decodes well the 31bit app with semtimedop call or whatever else and not to care about the "normal" strace and its compatibility with 31bit apps.

Is it like this?

Thanks.

Michael

Comment 15 errata-xmlrpc 2013-02-21 07:51:22 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHBA-2013-0282.html