Summary: | sys-cluster/torque-2.5.12 - qstat-qstat.o: In function `tcl_run': qstat.c:(.text+0x1ac5): undefined reference to `Tcl_GetErrorLine' | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Volkmar Glauche <volkmar.glauche> |
Component: | [OLD] Core system | Assignee: | Justin Bronder (RETIRED) <jsbronder> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | cluster, justin.t.riley, quantheory |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
build log
patch for src/cmds/qstat.c that fixes the issue for me |
Description
Volkmar Glauche
2013-01-23 09:05:41 UTC
Created attachment 336558 [details]
build log
Created attachment 338118 [details, diff]
patch for src/cmds/qstat.c that fixes the issue for me
This patch adds forwards/backwards compatibility for Tcl_GetErrorLine() to src/cmds/qstat.c. Still needs to be reported upstream...
I'm having this problem too. But isn't this caused by a patch in portage, not upstream? There are these three: torque-2.5.12-tcl8.6.patch torque-2.5.9-tcl8.6.patch torque-3.0.6-tcl8.6.patch These were created recently for bug 453154. AFAICT these are always being applied, but should only be applied if the "tk" use flag is on. Or the patch from comment 2 could be added to these, but it makes more sense to me to not apply the patch in the first place if "tk" is not on. Sorry for the delay guys, I admit to slacking. I've fixed the immediate problem for USE=-tk people with: 08 Mar 2013; Justin Bronder <jsbronder@gentoo.org> torque-2.5.9-r1.ebuild, torque-2.5.12.ebuild, torque-3.0.6-r1.ebuild: Toggle --enable-tcl-qstat based on USE=tk, resolves #453644 However, #453154 has been reopened as the patches break with USE=tk and <tk-8.6 Thanks for your patience and for reporting. I am sorry. I wasn't aware of the backward incompatibility. Actually I checker for Tcl_GetStringResult() which is fine, but Tcl_GetErrorLine() is new in 8.6 |