On Wed, Sep 26, 2007 at 10:46:56AM -0400, Jesse Keating wrote:
> On Wed, 26 Sep 2007 09:37:43 -0500
> Matt Domsch <Matt_Domsch@xxxxxxxx> wrote:
>
> > i re-ran the failures overnight, so some previously false failures are
> > now good. I'll have the stats email out shortly. Down to 395
> > failures on i386, 408 on x86_64.
>
> Did your build set include glibc-2.6.90-14? That introduces
> -D_FORTIFY_SOURCE{,=2} support for C++ and there may be some warnings in the build output about potential buffer overflows (which would result in runtime aborts and glibc logging). I'm curious if we can check the C++ builds for any warnings like this.
>
> Warnings such as "warning: call to __builtin___memcpy_chk will always
> overflow destination buffer"
I re-ran all the failed builds and all the C++ applications
overnight - there were no additional warnings reported, so the list
stands at:
dnsmasq-2.40-1.fc8.src.rpm
freedroidrpg-0.10.3-1.fc8.src.rpm
hping3-0.0.20051105-7.fc7.src.rpm
Io-language-20070710-2.fc8.src.rpm
libtirpc-0.1.7-10.fc8.src.rpm
mcrypt-2.6.6-2.fc8.src.rpm
openslp-1.2.1-7.fc8.src.rpm
perl-Imager-0.60-1.fc8.src.rpm
none of which are actually C++. dnsmasq upstream is already patched
to fix this (thanks jima).
Thanks,
Matt
--
Matt Domsch
Linux Technology Strategist, Dell Office of the CTO
linux.dell.com & www.dell.com/linux
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
