With this diff $ diff --recursive -u gcc-4.0-20050116/ gcc-play/ diff --recursive -u gcc-4.0-20050116/gcc/toplev.c gcc-play/gcc/toplev.c --- gcc-4.0-20050116/gcc/toplev.c 2005-01-16 16:28:15.000000000 +0100 +++ gcc-play/gcc/toplev.c 2005-01-17 17:10:13.000000000 +0100 @@ -269,7 +269,7 @@ 1 means wide ranges of inputs must work for complex divide. 2 means C99-like requirements for complex divide (not yet implemented). */
-int flag_complex_divide_method = 0; +int flag_complex_divide_method = 1; /* Nonzero means that we don't want inlining by virtue of -fno-inline, not just because the tree inliner turned us off. */ to the 20051116 snapshot, compiling a BLAS routine from netlib (to be attached) leads to an ICE: gfortran -O0 -g -c crotg.f crotg.f: In function 'crotg': crotg.f:19: internal compiler error: in purge_dead_edges, at cfgrtl.c:2460 Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://gcc.gnu.org/bugs.html> for instructions. make[1]: *** [crotg.o] Error 1 make[1]: Leaving directory `/home/zfkts/zfkts/LAPACK/BLAS/SRC' make: *** [blaslib] Error 2 $ gfortran -v ; gfortran -dumpmachine Using built-in specs. Configured with: ../gcc-4.0-20050116/configure --prefix=/home/zfkts --enable-languages=c,f95 Thread model: posix gcc version 4.0.0 20050116 (experimental) ia64-unknown-linux-gnu -- Summary: flags_complex_divide_method=1 doesn't work Product: gcc Version: 4.0.0 Status: UNCONFIRMED Severity: enhancement Priority: P2 Component: middle-end AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: Thomas dot Koenig at online dot de CC: gcc-bugs at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19486