| Summary: | app-backup/tsm clobbers $ROOTPATH (was: dev-util/cmake-3.14.1 with app-backup/tsm installed: bin/cmake: error while loading shared libraries: | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Toralf Förster <toralf> |
| Component: | Current packages | Assignee: | Andreas K. Hüttel <dilfridge> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | kde, slawomir.nizio, supersonic112 |
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: |
emerge-info.txt
dev-util:cmake-3.14.1:20190331-143947.log emerge-history.txt environment etc.portage.tbz2 logs.tbz2 temp.tbz2 |
||
|
Description
Toralf Förster
2019-03-31 15:07:34 UTC
Created attachment 571390 [details]
emerge-info.txt
Created attachment 571392 [details]
dev-util:cmake-3.14.1:20190331-143947.log
Created attachment 571394 [details]
emerge-history.txt
Created attachment 571396 [details]
environment
Created attachment 571398 [details]
etc.portage.tbz2
Created attachment 571400 [details]
logs.tbz2
Created attachment 571402 [details]
temp.tbz2
This is almost certainly a race condition (eg. openssl version change at the same time) or some breakage in the underlying environment - cmake doesn't use libcrypto at all. If the issue persists, could you please run the build again with V=1 VERBOSE=1 in the environment? (In reply to Michael Palimaka (kensington) from comment #8) > If the issue persists, could you please run the build again with V=1 > VERBOSE=1 in the environment? It is reproducible (at this image) Exact same issue here on multiple devices; do you happen to have app-backup/tsm-7.1.8.0 installed? For me, the issue could be circumvented by unmerging it temporarily. On a side note, sys-cluster/ceph-12.2.8-r1 also has/had an issue connected with tsm, but from its build log the problem was somewhat more obvious; there it shadowed the 'libcurl.so.4'. If you want, I can provide it, but from my (admittedly unexperienced) point of view the issue rather seems to come from TSM. Unfortunately I lack the experience and (more importantly) time to investigate the actual source of the issue right now, but if there's any way I can help, feel free to ask me. hhm, scary, but at least tsm is installed here: Sun Mar 31 08:13:37 2019 >>> app-backup/tsm-7.1.8.0 (In reply to Aldo from comment #10) > Exact same issue here on multiple devices; do you happen to have > app-backup/tsm-7.1.8.0 installed? For me, the issue could be circumvented by > unmerging it temporarily. On a side note, sys-cluster/ceph-12.2.8-r1 also > has/had an issue connected with tsm, but from its build log the problem was > somewhat more obvious; there it shadowed the 'libcurl.so.4'. If you want, I > can provide it, but from my (admittedly unexperienced) point of view the > issue rather seems to come from TSM. > > Unfortunately I lack the experience and (more importantly) time to > investigate the actual source of the issue right now, but if there's any way > I can help, feel free to ask me. Thanks for the tip, the build is definitely broken by cmake somehow picking up the stuff in /opt I don't know if it can be the reason or not, anyhow it's interesting that /opt/tivoli/tsm/client/ba/bin ends up as an element of PATH in ebuild's temp/environment file during build. …which is, by the look of it, due to ROOTPATH="/opt/tivoli/tsm/client/ba/bin" in /etc/env.d/80tivoli. I didn't try to remove that as I can't mess with that system I had it on. (In reply to Sławomir Nizio from comment #14) > …which is, by the look of it, due to > ROOTPATH="/opt/tivoli/tsm/client/ba/bin" in /etc/env.d/80tivoli. I didn't > try to remove that as I can't mess with that system I had it on. Thanks, I have tested this and agree it's the cause It seems like a bad idea for tsm to clobber $ROOTPATH like this, reassigning this bug to tsm maintainer. I've unset ROOTPATH on my system and will check over the next days if tsm still runs properly. Updates soon. Same problem here. Temporal removal of /etc/env.d/80tivoli followed by 'env-update' and 'source /etc/profile' works for me to build cmake *** Bug 690228 has been marked as a duplicate of this bug. *** any news here ?
what about using
start-stop-daemon --start --background --nicelevel 15 \
--env PATH="${PATH}:/opt/tivoli/tsm/client/ba/bin" \
....
for daemon (to ensure the the existing functional is kept).
AFAIS are relevant binaries are symlinked anyway:
/usr/bin/dsmadmc -> ../../opt/tivoli/tsm/client/ba/bin/dsmadmc
/usr/bin/dsmagent -> ../../opt/tivoli/tsm/client/ba/bin/dsmagent
/usr/bin/dsmc -> ../../opt/tivoli/tsm/client/ba/bin/dsmc
/usr/bin/dsmcad -> ../../opt/tivoli/tsm/client/ba/bin/dsmcad
/usr/bin/dsmcert -> ../../opt/tivoli/tsm/client/ba/bin/dsmcert
/usr/bin/dsmj -> ../../opt/tivoli/tsm/client/ba/bin/dsmj
/usr/bin/dsmswitch -> ../../opt/tivoli/tsm/client/ba/bin/dsmswitch
/usr/bin/dsmtrace -> ../../opt/tivoli/tsm/client/ba/bin/dsmtrace
/usr/bin/gskcapicmd_64 -> ../../opt/ibm/gsk8_64/bin/gsk8capicmd_64
/usr/bin/gskver_64 -> ../../opt/ibm/gsk8_64/bin/gsk8ver_64
*** Bug 716932 has been marked as a duplicate of this bug. *** *** Bug 716980 has been marked as a duplicate of this bug. *** The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=9ada041e9ece7eea13f52a12e55155ea8bdc9d70 commit 9ada041e9ece7eea13f52a12e55155ea8bdc9d70 Author: Andreas K. Hüttel <dilfridge@gentoo.org> AuthorDate: 2020-05-06 20:06:56 +0000 Commit: Andreas K. Hüttel <dilfridge@gentoo.org> CommitDate: 2020-05-06 21:15:08 +0000 app-backup/tsm: Drop ROOTPATH from env file Closes: https://bugs.gentoo.org/682142 Package-Manager: Portage-2.3.99, Repoman-2.3.22 Signed-off-by: Andreas K. Hüttel <dilfridge@gentoo.org> app-backup/tsm/{tsm-7.1.8.0.ebuild => tsm-7.1.8.0-r1.ebuild} | 1 - 1 file changed, 1 deletion(-) |
