On Sun, Jun 08, 2008 at 01:38:54PM -0700, [EMAIL PROTECTED] wrote: >Author: bernds >Date: 2008-06-08 13:38:53 -0700 (Sun, 08 Jun 2008) >New Revision: 22271 > >Log: >In string.h, move libc_hidden_proto next to the corresponding declaration. >Filter them out when installing headers. > >Tested by building up bfin-*linux toolchains and building our copy of >uClinux-dist.
>Modified: trunk/uClibc/include/string.h >=================================================================== >--- trunk/uClibc/include/string.h 2008-06-08 16:47:09 UTC (rev 22270) >+++ trunk/uClibc/include/string.h 2008-06-08 20:38:53 UTC (rev 22271) >@@ -38,10 +38,12 @@ > extern void *memcpy (void *__restrict __dest, > __const void *__restrict __src, size_t __n) > __THROW __nonnull ((1, 2)); >+libc_hidden_proto(memcpy) Hi, unfortunately this trips an assert on current gcc trunk for uClibc's DOMULTI (all honza's fault i assume, this definitely compiled fine some weeks ago ;) In file included from libc/sysdeps/linux/i386/sigaction.c:24: ./include/string.h:41: internal compiler error: in change_decl_assembler_name, at cgraph.c:1154 $ head -n 1154 cgraph.c | tail -n1 gcc_assert (!assembler_name_hash); Perhaps you can look into it? PS: oh, and i would be very grateful if you could eventually comment on my patch for PR31537 (an issue with duplicate weakrefs where i have no clear idea on how this was ment to work in the first place; perhaps you do or somebody here does?). That PR is not really related but IIRC both uClibc and libgfortran triggered it. _______________________________________________ uClibc mailing list [email protected] http://busybox.net/cgi-bin/mailman/listinfo/uclibc
