This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, i386] Optionally use %xmm0 to return float and/or doublevalues (take 2)
Paolo Bonzini wrote:
However, tomorrow (unless Uros beats me to it :-) I'll benchmark
povray: a much simpler patch that only does this for local functions
might still provide a good bang for the buck.
A compile error is here:
+ /* Most things go in %eax, except (unless -mno-fp-ret-in-387) fp values. */
+ if (GET_MODE_CLASS (mode) != MODE_FLOAT || !TARGET_FLOAT_RETURNS_IN_80387))
+ return 0;
When (fixed) patch is applied, povray-3.50c (-mfpmath=sse) compilation
fails with:
c++ -DPREFIX=\"/usr/local\"
-DPOV_LIB_DIR=\"/usr/local/share/povray-3.5\"
-DCOMPILER_VER=\".Linux.gcc\" -DSYSCONFDIR=\"/usr/local/etc\"
-DUSE_IO_RESTRICTIONS=\"\" `if [ "Xgcc" = "Xgcc" ]; then echo
"-Wno-multichar"; fi ` -O3 -march=pentium4 -mfpmath=sse -ffast-math
-D__NO_MATH_INLINES -finline-functions -fomit-frame-pointer
-funroll-loops -fexpensive-optimizations -malign-double
-foptimize-sibling-calls -minline-all-stringops -Wno-multichar -c chi2.cpp
chi2.cpp: In function âdouble lgam(double)â:
chi2.cpp:347: warning: converting to âintâ from âdoubleâ
chi2.cpp:417: internal compiler error: in cgraph_local_info, at cgraph.c:443
Please submit a full bug report,
Uros.