Bug 205976
Summary: | Missing gdb.ppc64 build (while gdb.ppc exists) | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Jan Kratochvil <jan.kratochvil> |
Component: | distribution | Assignee: | Bill Nottingham <notting> |
Status: | CLOSED RAWHIDE | QA Contact: | Bill Nottingham <notting> |
Severity: | medium | Docs Contact: | |
Priority: | medium | ||
Version: | rawhide | CC: | dcantrell, qiyaoltc, rth, rvokal, zhengyzy, zhouwu |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | ppc64 | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | gdb-6.5-8.fc6.ppc64 | Doc Type: | Bug Fix |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2006-09-30 09:05:01 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: | 205235 | ||
Bug Blocks: | 205333, 206488 |
Description
Jan Kratochvil
2006-09-11 03:53:09 UTC
As stated in the bug blocked by this, we can't really include this as multilib as the ppc and ppc64 versions conflict with eachother. In i386 / x86_64 land this isn't a problem so much as each arch gets their own gdb. But since we only ship one ppc tree that is both ppc and ppc64 we can't do that here. As I consider gdb.ppc64 as essential component of RHEL.ppc64 it should be resolved somehow. What to choose? (1) You will handle conflicting "application-multilibs" as some new feature of anaconda/yum till FC6-GA. (2) I should provide specific ppc64-conditional packaging in .spec to make gdb-6.5-X.ppc.rpm and gdb-6.5-X.ppc64.rpm non-conflicting (separate /usr/bin/gdb64 binary etc.). (3) I should provide specific ppc64-conditional packaging in .spec to make gdb-6.5-X.ppc.rpm and gdb64-6.5-X.ppc64.rpm non-conflicting (separate /usr/bin/gdb64 binary etc.). Not sure if it gets installed by anaconda/yum appropriately, I hope so. How is it that gcc handles this? Of course a better answer would be to stop trying to do silly things by offering only one ppc variant, but I don't think we can do that. In RHEL5 land, the ppc tree may be from a standpoint of we support only ppc64 machines, but 32 bit runs faster, so we could provide only the 64bit gdb. Can a ppc64 gdb debug 32bit applications? Yes, AFAIK. How do they conflict? If it's just /usr/bin/gdb, that should be handled fine by the normal elf32/elf64 handling. (In reply to comment #3) > Can a ppc64 gdb debug 32bit applications? Yes. ppc64 gdb can debug 32bit and 64bit application at the same time. Both RHEL4 and FC5 for ppc/ppc64 shipped a 64bit gdb, which can debug both 32-bit and 64bit application. I am not sure how these distribution build (and ship) it though. I'm going to add gdb back to the "whitelist" of multilib packages. Normal handling should make this work. (In reply to comment #6) > I'm going to add gdb back to the "whitelist" of multilib packages. Normal > handling should make this work. OK, understood it is in fact enough. (In reply to comment #6) > I'm going to add gdb back to the "whitelist" of multilib packages. http://armstrong.rhts.boston.redhat.com/distros/rawhide-ppc/Fedora/RPMS/ with timestamps "13-Sep-2006 05:34" still does not contain "gdb-*.ppc64.rpm". Should I wait more or it just still has not been done? *** Bug 168157 has been marked as a duplicate of this bug. *** *** Bug 205333 has been marked as a duplicate of this bug. *** Exists and it can debug itself: http://armstrong.rhts.boston.redhat.com/distros/rawhide-ppc/Fedora/RPMS/gdb-6.5-8.fc6.ppc64.rpm /usr/bin/gdb: ELF 64-bit MSB executable, cisco 7500, version 1 (SYSV), for GNU/Linux 2.6.9, dynamically linked (uses shared libs), for GNU/Linux 2.6.9, stripped |