On Thu, Mar 10, 2011 at 5:12 AM, walt <w41...@gmail.com> wrote: > On 03/09/2011 05:27 PM, Andrew Wilkinson wrote: > >> I have a VM image based on the amd64 hardened no-multilib profile. Today, >> when trying to emerge glibc-2.11.3 and gcc-4.4.5, the builds have been >> failing, and appear to be trying to build multilib versions, which of course >> it can't. Any idea why this would be happening? >> >> From glibc's build.log: >> >> * Building multilib glibc for ABIs: x86 amd64 >> >> * ABI: x86 >> * CBUILD: x86_64-pc-linux-gnu >> * CHOST: x86_64-pc-linux-gnu >> * CTARGET: x86_64-pc-linux-gnu >> * CBUILD_OPT: i686-pc-linux-gnu >> * CTARGET_OPT: i686-pc-linux-gnu >> * CC: >> * CFLAGS: -m32 -march=i686 -O2 -fno-strict-aliasing >> -fno-stack-protector >> * Manual CC: x86_64-pc-linux-gnu-gcc >> > > You might try emerge -d (debug). The Changelog is loaded with comments > about building > glibc on hardened systems, so maybe they introduced a recent bug? > > > Thanks for the tip! Unfortunately someone who knows more about what portage is doing than I do will need to look at this... in my glance over it I got the impression that it was magically conjuring the x86 ABI result from the ether, or something.
I've put it up on pastebin for now. If this is something bug worthy, I'll thus report. http://pastebin.com/vS6BMhJN Thanks again! -Andy
