USE=sanitize has not been ported to all supported architectures, having sys-devel/gcc failing to build when creating cross-compilers.
Note that it's not "sys-devel/gcc" but "cross-*/gcc" when building cross-compilers (thus http://sources.gentoo.org/profiles/base/package.use.mask?r1=1.617&r2=1.618 doesn't fix this bug).
(In reply to Andrew Church from comment #1) > Note that it's not "sys-devel/gcc" but "cross-*/gcc" when building > cross-compilers (thus > http://sources.gentoo.org/profiles/base/package.use.mask?r1=1.617&r2=1.618 > doesn't fix this bug). That's why this bug report doesn't mention that at all.
*** Bug 526784 has been marked as a duplicate of this bug. ***
*** Bug 526932 has been marked as a duplicate of this bug. ***
Created attachment 387766 [details, diff] 525910-crossdev-sanitize.patch
http://git.overlays.gentoo.org/gitweb/?p=proj/crossdev.git;a=commitdiff;h=09cc960610ff36e007e9fd26dc79edbdf4bfeb25
Created attachment 387856 [details] cross-hppa2.0-unknown-linux-gnu-gcc-stage2.log.xz It's masked on stage1 but not stage2 so it still fails
that's to be expected *** This bug has been marked as a duplicate of bug 370063 ***
This isn't a duplicate.
You didn't fix it at all.
crossdev writes and overwrites entries in package.use. Changes to those entries (for instance "cross-hppa2.0-unknown-linux-gnu/gcc -sanitize" are getting overwritten by the stage2 gcc run which therefore fails again and again.
i know how it works. i fixed the stage1 behavior because that matters everywhere. what you're talking about here is bug 370063, so yes, it is a dupe. *** This bug has been marked as a duplicate of bug 370063 ***
*** Bug 527976 has been marked as a duplicate of this bug. ***
*** Bug 527980 has been marked as a duplicate of this bug. ***
*** Bug 530786 has been marked as a duplicate of this bug. ***
*** Bug 532556 has been marked as a duplicate of this bug. ***
Please fix in main tree gentoo for architecture mips-linux-uclibc.
(In reply to SpanKY from comment #12) > i know how it works. i fixed the stage1 behavior because that matters > everywhere. what you're talking about here is bug 370063, so yes, it is a > dupe. > > *** This bug has been marked as a duplicate of bug 370063 *** Hi, just tried the overlay and tried to crossdev armv5tel-softfloat-linux-uclibceabi. During stage1 -sanitize is present, but during stage 2 it is not and the build fails just as it did before. So either I am doing something wrong or the fix is not enough.
Created attachment 396042 [details] failure of gcc stage2 with crossdev-99999999 Attached is the log for the stage 2 failure.
*** Bug 547226 has been marked as a duplicate of this bug. ***
*** Bug 536738 has been marked as a duplicate of this bug. ***
*** Bug 549510 has been marked as a duplicate of this bug. ***
*** Bug 550772 has been marked as a duplicate of this bug. ***
I'm also seeing compilation failure on libsanitize in gcc stage2 when building for mingw64 (using crossdev-20150602).
*** Bug 564314 has been marked as a duplicate of this bug. ***
*** Bug 567878 has been marked as a duplicate of this bug. ***
*** Bug 566976 has been marked as a duplicate of this bug. ***
*** Bug 581686 has been marked as a duplicate of this bug. ***
*** Bug 582298 has been marked as a duplicate of this bug. ***
*** Bug 582664 has been marked as a duplicate of this bug. ***
*** Bug 575170 has been marked as a duplicate of this bug. ***