This is the mail archive of the crossgcc@sourceware.org mailing list for the crossgcc project.
See crosstool-NG for lots more information.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
Hi Yann, Johannes, all, On Saturday, February 16, 2013 3:19:37 PM, Johannes Stezenbach wrote: > Hi Yann, > > On Sat, Feb 16, 2013 at 12:37:28AM +0100, Yann E. MORIN wrote: > > On Saturday 16 February 2013 Yann E. MORIN wrote: > > > On Tuesday 05 February 2013 BenoÃt ThÃbaudeau wrote: > > > [--SNIP--] > > > > +CT_LIBC_GLIBC_EXTRA_CONFIG_ARRAY="--disable-multi-arch" > > > Ah... The multi-arch debianism... :-( > > > > OK, forget it, I need some rest. Seems it's unrelated to Debian multi-arch. > > > > But what does it mean to have a "single DSO with optimizations for multiple > > architectures" ? > > I'm guessing it is the STT_GNU_IFUNC thing. > http://udrepper.livejournal.com/20948.html This is also something I took from the Sourcery build script. I think that Johannes' guess is the right one. As we are building for only a single architecture (no multi-lib), there is no reason to enable this multi-arch DSO optimization mechanism, which could only slow things down. This is probably not strictly needed, but better without multi-lib (which does not work for ARM anyway). I think that this option could be automated in CT-NG for (e)glibc in all cases without multi-lib. Best regards, BenoÃt -- For unsubscribe information see http://sourceware.org/lists.html#faq
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |