Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 52669 - "make" goes into infinite loop in function dyn_test (portage-2.0.51_pre9)
Summary: "make" goes into infinite loop in function dyn_test (portage-2.0.51_pre9)
Status: RESOLVED FIXED
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: [OLD] Core system (show other bugs)
Hardware: All All
: High major (vote)
Assignee: Portage team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2004-06-01 03:03 UTC by Robert Moss (RETIRED)
Modified: 2004-06-09 17:41 UTC (History)
0 users

See Also:
Package list:
Runtime testing required: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Robert Moss (RETIRED) gentoo-dev 2004-06-01 03:03:59 UTC
I'm in Lynx at the moment so I can't give you "emerge info", but this is a relatively bleeding edge system - GCC 3.4, glibc-2.3.3_pre20040529, portage-2.0.51_pre9, linux-headers-2.6.5. CFLAGS="-O2 -march=athlon-xp -ftracer -fomit-frame-pointer -ftracer -pipe".
The issue is compiling libperl. The new portage-2.0.51 function "dyn_test", responsible for running "make check" (or its equivalent), sends make into an infinite loop. This is from a stage 1 install. Eventually, something really helpful will happen, like everything being killed!
The workaround is to insert an empty function into the ebuild:
dyn_test() { }
This stops the loop being entered. I'm unsure of a fix as I don't actually know what the function does.
Reproducible with each of libperl-5.8.{2,3,4}.
Comment 1 Robert Moss (RETIRED) gentoo-dev 2004-06-01 03:04:59 UTC
Oh, I should add that this is also reproducible with FEATURES="-*" USE="-*" CFLAGS="-O0" CXXFLAGS="-O0".
Comment 2 Nicholas Jones (RETIRED) gentoo-dev 2004-06-09 17:29:14 UTC
The default behaviour changed in _pre10.
Comment 3 Robert Moss (RETIRED) gentoo-dev 2004-06-09 17:41:29 UTC
Excellent! Ish. How to I re-enable the test phase? I found this quite helpful for certain packages whilst testing gcc-3.5 CVS HEAD, which I'm working on porting. Sure, all sorts of packages miscompile horribly, but it would be nice to get rid of the compilation errors so that when the runtime errors get fixed 3.5 will be usable straight out of the box. And I can't test for runtime errors without dyn_test...