Summary: | sci-chemistry/gromacs-4.6_beta1 with dev-util/nvidia-cuda-toolkit-5.0.35 - nvcc fatal : redefinition of argument 'compiler-bindir' | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Juergen Rose <rose> |
Component: | Current packages | Assignee: | Alexey Shvetsov <alexxy> |
Status: | RESOLVED UPSTREAM | ||
Severity: | normal | CC: | sci-chemistry |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
URL: | http://redmine.gromacs.org/issues/1051 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | /var/tmp/portage/sci-chemistry/gromacs-4.6_beta1/temp/build.log |
Description
Juergen Rose
2012-12-01 14:54:01 UTC
I have no cuda card, let's see if alexxy can help. I will take a look on that. Currently I can't see the origin of the first definition of compiler-bindir, only the second which is in our control. (In reply to comment #2) > I will take a look on that. Currently I can't see the origin of the first > definition of compiler-bindir, only the second which is in our control. It comes from adding -ccbin in gmxManageNvccConfig.cmake: <http://repo.or.cz/w/gromacs.git/blob/HEAD:/cmake/gmxManageNvccConfig.cmake#l97> If you find a general solution to be merged upstream upload it here: https://gerrit.gromacs.org (In reply to comment #3) > (In reply to comment #2) > > I will take a look on that. Currently I can't see the origin of the first > > definition of compiler-bindir, only the second which is in our control. > > It comes from adding -ccbin in gmxManageNvccConfig.cmake: > <http://repo.or.cz/w/gromacs.git/blob/HEAD:/cmake/gmxManageNvccConfig. > cmake#l97> Thats the second, which we can control. but somehow it got added earlier. I will try to find the source for that. Got fixed. |