| Message ID | yddpq5john5.fsf@manam.CeBiTec.Uni-Bielefeld.DE |
|---|---|
| State | New |
| Headers | show |
On Tue, Sep 18, 2012 at 10:32 AM, Rainer Orth <ro@cebitec.uni-bielefeld.de> wrote: > The libbacktrace integration broke Solaris 10 and 11 bootstrap when > using gcc 4.4 (any version of gcc without __sync_* support actually): Ouch, that's bad. > stage1 config.h has > > /* #undef HAVE_SYNC_FUNCTIONS */ > > and fileline.c and mmap.c fail to compile: > > /vol/gcc/src/hg/trunk/local/libbacktrace/fileline.c: In function 'fileline_init > alize': > /vol/gcc/src/hg/trunk/local/libbacktrace/fileline.c:58: error: implicit declarat > ion of function 'abort' > > The following patch fixes this by including <stdlib.h> for the abort() > declaration in the affected files. It allows the Solaris 11 bootstrap > to continue. > > Ok for mainline? Ok. Thanks, Richard. > Unfortunately, Solaris 10 (and certainly Solaris 9, too) bootstrap is still > broken: > > /vol/gcc/src/hg/trunk/local/libbacktrace/dwarf.c:652: error: implicit declaration of function 'strnlen' > make[1]: *** [dwarf.lo] Error 1 > > Both completely lack strnlen(). I haven't done anything about this yet. > > Rainer > > > 2012-09-18 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE> > > * fileline.c: Include <stdlib.h>. > * mmap.c: Likewise. > > > > > -- > ----------------------------------------------------------------------------- > Rainer Orth, Center for Biotechnology, Bielefeld University >
On Tue, Sep 18, 2012 at 10:54 AM, Richard Guenther <richard.guenther@gmail.com> wrote: > On Tue, Sep 18, 2012 at 10:32 AM, Rainer Orth > <ro@cebitec.uni-bielefeld.de> wrote: >> The libbacktrace integration broke Solaris 10 and 11 bootstrap when >> using gcc 4.4 (any version of gcc without __sync_* support actually): > > Ouch, that's bad. Btw, why do we need to build libbacktrace during stage1? >> stage1 config.h has >> >> /* #undef HAVE_SYNC_FUNCTIONS */ >> >> and fileline.c and mmap.c fail to compile: >> >> /vol/gcc/src/hg/trunk/local/libbacktrace/fileline.c: In function 'fileline_init >> alize': >> /vol/gcc/src/hg/trunk/local/libbacktrace/fileline.c:58: error: implicit declarat >> ion of function 'abort' >> >> The following patch fixes this by including <stdlib.h> for the abort() >> declaration in the affected files. It allows the Solaris 11 bootstrap >> to continue. >> >> Ok for mainline? > > Ok. > > Thanks, > Richard. > >> Unfortunately, Solaris 10 (and certainly Solaris 9, too) bootstrap is still >> broken: >> >> /vol/gcc/src/hg/trunk/local/libbacktrace/dwarf.c:652: error: implicit declaration of function 'strnlen' >> make[1]: *** [dwarf.lo] Error 1 >> >> Both completely lack strnlen(). I haven't done anything about this yet. >> >> Rainer >> >> >> 2012-09-18 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE> >> >> * fileline.c: Include <stdlib.h>. >> * mmap.c: Likewise. >> >> >> >> >> -- >> ----------------------------------------------------------------------------- >> Rainer Orth, Center for Biotechnology, Bielefeld University >>
On Tue, Sep 18, 2012 at 1:32 AM, Rainer Orth <ro@cebitec.uni-bielefeld.de> wrote: > The libbacktrace integration broke Solaris 10 and 11 bootstrap when > using gcc 4.4 (any version of gcc without __sync_* support actually): The patch is fine and should fix the problem, but GCC 4.4 does have __sync_* support. Might be worth looking into why the test failed. > Unfortunately, Solaris 10 (and certainly Solaris 9, too) bootstrap is still > broken: > > /vol/gcc/src/hg/trunk/local/libbacktrace/dwarf.c:652: error: implicit declaration of function 'strnlen' > make[1]: *** [dwarf.lo] Error 1 > > Both completely lack strnlen(). I haven't done anything about this yet. This should be fixed now. Sorry about the problems. Ian
On Tue, Sep 18, 2012 at 1:55 AM, Richard Guenther <richard.guenther@gmail.com> wrote: > On Tue, Sep 18, 2012 at 10:54 AM, Richard Guenther > <richard.guenther@gmail.com> wrote: >> On Tue, Sep 18, 2012 at 10:32 AM, Rainer Orth >> <ro@cebitec.uni-bielefeld.de> wrote: >>> The libbacktrace integration broke Solaris 10 and 11 bootstrap when >>> using gcc 4.4 (any version of gcc without __sync_* support actually): >> >> Ouch, that's bad. > > Btw, why do we need to build libbacktrace during stage1? It's easier for users. The library is designed to quietly do nothing if it does not have the required support. This was just a bug in that design. Ian
# HG changeset patch # Parent a22dd5d7246fa4e8a73de2e66db7594cf9ae9f5a Fix bootstrap with gcc 4.4 diff --git a/libbacktrace/fileline.c b/libbacktrace/fileline.c --- a/libbacktrace/fileline.c +++ b/libbacktrace/fileline.c @@ -35,6 +35,7 @@ POSSIBILITY OF SUCH DAMAGE. */ #include <sys/types.h> #include <sys/stat.h> #include <fcntl.h> +#include <stdlib.h> #include "backtrace.h" #include "internal.h" diff --git a/libbacktrace/mmap.c b/libbacktrace/mmap.c --- a/libbacktrace/mmap.c +++ b/libbacktrace/mmap.c @@ -34,6 +34,7 @@ POSSIBILITY OF SUCH DAMAGE. */ #include <errno.h> #include <string.h> +#include <stdlib.h> #include <unistd.h> #include <sys/mman.h>