Bug 123697

Summary: portmap contains name provide causing triggers to fire incorrectly
Product: [Fedora] Fedora Reporter: Glen <glen>
Component: rpmAssignee: Jeff Johnson <jbj>
Status: CLOSED RAWHIDE QA Contact:
Severity: medium Docs Contact:
Priority: medium    
Version: 2CC: jakub, jval, nobody+pnasrat, pmjcovello, rdieter, shishz, steved, wtogami
Target Milestone: ---   
Target Release: ---   
Hardware: i386   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2004-06-16 13:12:32 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: 123680    
Attachments:
Description Flags
Patch to filter out DT_DEBUG none

Description Glen 2004-05-20 03:59:25 UTC
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4.1)
Gecko/20031114

Description of problem:
When I run "rpm -e portmap", I get these errors:

error reading information on service portmap: No such file or directory
error:%prerun(portmap-4.0-59) scriptlet failed, exit status 1

Version-Release number of selected component (if applicable):
portmap-4.0-59

How reproducible:
Always

Steps to Reproduce:
1. Run "rpm -e portmap"
2.
3.
    

Actual Results:  It gives me the error I mentioned.

Expected Results:  I expected portmap to be removed.

Additional info:

Comment 1 Paul Nasrat 2004-05-20 05:56:24 UTC
Path failure on fc2 use /sbin/service rather than service

[pauln@enki tmp]$ rpm --scripts -q portmap
postinstall scriptlet (using /bin/sh):
/sbin/chkconfig --add portmap
/usr/sbin/groupadd -g 32 rpc > /dev/null 2>&1
/usr/sbin/useradd -c "Portmapper RPC user" -d / -g 32 -M -s
/sbin/nologin -u 32 rpc > /dev/null 2>&1
exit 0
preuninstall scriptlet (using /bin/sh):
if [ $1 = 0 ]; then
  service portmap stop > /dev/null 2>&1
  /sbin/chkconfig --del portmap
fi
postuninstall scriptlet (using /bin/sh):
if [ "$1" -ge "1" ]; then
  service portmap condrestart > /dev/null 2>&1
fi


Comment 2 Paul Nasrat 2004-05-20 06:24:26 UTC
This is not a path issue - SCRIPT_PATH is unaltered looks to be a
trigger misfiring:

rpm --triggers -q portmap
triggerpostun scriptlet (using /bin/sh) -- portmap <= portmap-4.0-22
/sbin/chkconfig --add portmap

from rpm -e -vv --nodeps portmap I get:

D:     erase: %trigger(portmap-4.0-59)	execv(/bin/sh) pid 4466

+ /sbin/chkconfig --add portmap

error reading information on service portmap: No such file or directory

D:     erase: waitpid(4466) rc 4466 status 100 secs 0.073

error: %trigger(portmap-4.0-59) scriptlet failed, exit status 1



Comment 3 Paul Nasrat 2004-05-20 11:31:22 UTC
portmap seems to have a %{name} level provides as triggers match
against all provides and Provides: %{name} matches all EVR it's being
fired

rpm -q --provides portmap
config(portmap) = 4.0-59
portmap
portmap = 4.0-59

I can't see Provides: %{name} or anything in the spec so need to
figure why the virtual is there.


Comment 4 Paul Nasrat 2004-05-20 13:36:49 UTC
Change component to rpm

Comment 5 Paul Nasrat 2004-05-20 17:28:12 UTC
bingo:

[pauln@enki pauln]$ file /sbin/portmap
/sbin/portmap: ELF 32-bit LSB shared object, Intel 80386, version 1
(SYSV), stripped

The provide here is due to SONAME adding:

  /* For DSO's, provide the basename of the file if DT_SONAME not
found. */
    if (!fc->skipProv && isDSO && !gotSONAME) {

All the following binaries are probably the same:

exim
finger
httpd
nscd
perl
perl-URI
portmap
postgresql-odbc
routed
rusers
squid
sudo
telnet
vsftpd


Comment 6 Paul Nasrat 2004-05-20 17:31:13 UTC
Note this is probably due to the executables being PIE.  Probably need
to teach rpm how to tell the difference if we can

Comment 7 Paul Nasrat 2004-05-21 09:52:13 UTC
Created attachment 100411 [details]
Patch to filter out DT_DEBUG

Tested this against rpm-4_3 branch and filtering for DT_DEBUG skips adding a
basename (usually == %{name}) of the PIE executable to Provides:

Comment 8 Paul Nasrat 2004-05-21 09:59:16 UTC
Jeff does this all look sane to you?

Comment 9 Thomas Woerner 2004-05-21 11:25:01 UTC
*** Bug 123680 has been marked as a duplicate of this bug. ***

Comment 10 Jeff Johnson 2004-06-16 13:12:32 UTC
The PIE patch is in latest rpm-4.3.2.

Comment 11 Bill Nottingham 2004-06-28 21:17:01 UTC
*** Bug 123680 has been marked as a duplicate of this bug. ***

Comment 12 Bill Nottingham 2004-07-09 20:55:01 UTC
*** Bug 127571 has been marked as a duplicate of this bug. ***

Comment 13 Jarkko 2004-07-10 08:31:04 UTC
What should I do to get ppp updated? Update rpm? Why the fix is not in
the updates?

Comment 14 Matthew Miller 2004-07-10 16:55:57 UTC
*** Bug 127554 has been marked as a duplicate of this bug. ***

Comment 15 Paul Nasrat 2004-07-10 18:16:18 UTC
Any chance of getting 4.3.2 as errata for fc2 and into buildroots?

Comment 19 Stefan Neufeind 2005-02-12 00:15:50 UTC
Would really appreciate a release of a fixed rpm for fc2. Any chance?