On Fri, 19 Aug 2011, Richard Guenther wrote:

> On Fri, 19 Aug 2011, Richard Guenther wrote:
> 
> > 
> > This is the fix for the testcase in PR50067.  We strip outermost
> > (yes, outermost only, which makes it very inefficient) MEM_REFs
> > which causes the DR base objects in the PR to agree for two
> > conflicting DRs, but with the issues we have with how we
> > compose access functions they still get disambiguated.
> > 
> > Bootstrap and regtest pending on x86_64-unknown-linux-gnu.
> 
> Updated patch, which exposes some latent wide-multiply GIMPLE
> type issues at least.  Fixes all known testcases I have.
> 
> We can't really use the MEM_REF (or maybe any?) offset from
> the base object as independent access-function.  Nor can
> we replace the base with scev_not_known - the alias oracle
> will assume funny things about this (a latent issue for sure).
> 
> Bootstrap and regtest running on x86_64-unknown-linux-gnu,
> I'll have to dig into the latent issue exposed first, but that's
> for next week only.

Bah, bootstrap fails very early when building stage1 libgcc:

/abuild/rguenther/obj3/./gcc/xgcc -B/abuild/rguenther/obj3/./gcc/ 
-B/usr/local/x86_64-unknown-linux-gnu/bin/ 
-B/usr/local/x86_64-unknown-linux-gnu/lib/ -isystem 
/usr/local/x86_64-unknown-linux-gnu/include -isystem 
/usr/local/x86_64-unknown-linux-gnu/sys-include    -g -O2 -m32 -O2  -I. 
-I. -I/space/rguenther/src/svn/trunk/gcc 
-I/space/rguenther/src/svn/trunk/gcc/. 
-I/space/rguenther/src/svn/trunk/gcc/../include 
-I/space/rguenther/src/svn/trunk/gcc/../libdecnumber 
-I/space/rguenther/src/svn/trunk/gcc/../libdecnumber/bid -I../libdecnumber 
-I/space/rguenther/src/svn/trunk/gcc/../libgcc -g -O2 -DIN_GCC   -W -Wall 
-Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes 
-Wold-style-definition  -isystem ./include  -fPIC -g -DHAVE_GTHR_DEFAULT 
-DIN_LIBGCC2 -fbuilding-libgcc -fno-stack-protector   -I. -I. 
-I../../.././gcc -I/space/rguenther/src/svn/trunk/libgcc 
-I/space/rguenther/src/svn/trunk/libgcc/. 
-I/space/rguenther/src/svn/trunk/libgcc/../gcc 
-I/space/rguenther/src/svn/trunk/libgcc/../include 
-I/space/rguenther/src/svn/trunk/libgcc/config/libbid 
-DENABLE_DECIMAL_BID_FORMAT -DHAVE_CC_TLS  -DUSE_TLS -o bid64_div.o -MT 
bid64_div.o -MD -MP -MF bid64_div.dep -c 
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c: In 
function '__bid64dq_div':
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c:523:51: 
warning: variable 'Ql' set but not used [-Wunused-but-set-variable]
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c: In 
function '__bid64qd_div':
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c:937:51: 
warning: variable 'Ql' set but not used [-Wunused-but-set-variable]
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c: In 
function '__bid64qq_div':
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c:1374:51: 
warning: variable 'Ql' set but not used [-Wunused-but-set-variable]
PD_359 = digit_22 w* 109951163;

/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c: In 
function '__bid64_div':
/space/rguenther/src/svn/trunk/libgcc/config/libbid/bid64_div.c:80:1: 
internal compiler error: verify_gimple failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

I suppose it's exposed by

2011-08-19  Andrew Stubbs  <a...@codesourcery.com>

       * tree-ssa-math-opts.c (build_and_insert_cast): New function.
        (is_widening_mult_rhs_p): Allow widening by more than one mode.
        Explicitly disallow mis-matched input types.
        (convert_mult_to_widen): Use find_widening_optab_handler, and cast
        input types to fit the new handler.
        (convert_plusminus_to_widen): Likewise.

Andrew - appearantly you broke bootstrap on x86_64.

Richard.

Reply via email to