This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: RFA: enable LRA for rs6000 [patch for WRF]


I'm seeing a lot of failures with these changes in make check.

The first two that I noticed on a build that did not use --with-cpu=power7:

1) c-c++-common/dfp/call-by-value.c (and others in the directory) fails with -O0
for all targets before power7 because it can't spill SDmode.  Note, in the
earlier targets, we need to have a wider spill slot because we don't have
32-bit integer load/store to the FPR registers.

FAIL: c-c++-common/dfp/call-by-value.c (internal compiler error)
FAIL: c-c++-common/dfp/call-by-value.c (test for excess errors)
Excess errors:
/home/meissner/fsf-src/meissner-lra/gcc/testsuite/c-c++-common/dfp/call-by-value.c:43:1: internal compiler error: in assign_by_spills, at lra-assigns.c:1268
0x104ceff7 assign_by_spills
        /home/meissner/fsf-src/meissner-lra/gcc/lra-assigns.c:1268
0x104cfe43 lra_assign()
        /home/meissner/fsf-src/meissner-lra/gcc/lra-assigns.c:1425
0x104ca837 lra(_IO_FILE*)
        /home/meissner/fsf-src/meissner-lra/gcc/lra.c:2309
0x1047d6eb do_reload
        /home/meissner/fsf-src/meissner-lra/gcc/ira.c:4619
0x1047d6eb rest_of_handle_reload
        /home/meissner/fsf-src/meissner-lra/gcc/ira.c:4731

2) A lot of fortran tests are failing for all optimization levels due to
segmentation violations at runtime:

AIL: gfortran.dg/advance_1.f90  -O0  execution test
Executing on host: /home/meissner/fsf-build-ppc64/meissner-lra/gcc/testsuite/gfortran1/../../gfortran -B/home/meissner/fsf-build-ppc64/meissner-lra/gcc/testsuite/gfortran1/../../ -B/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unkno
wn-linux-gnu/./libgfortran/ /home/meissner/fsf-src/meissner-lra/gcc/testsuite/gfortran.dg/advance_1.f90  -fno-diagnostics-show-caret   -O1   -pedantic-errors  -B/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libg
fortran/.libs -L/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs -L/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs  -lm   -m32 -o ./advance_1.exe    (t
imeout = 300)
spawn /home/meissner/fsf-build-ppc64/meissner-lra/gcc/testsuite/gfortran1/../../gfortran -B/home/meissner/fsf-build-ppc64/meissner-lra/gcc/testsuite/gfortran1/../../ -B/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/
./libgfortran/ /home/meissner/fsf-src/meissner-lra/gcc/testsuite/gfortran.dg/advance_1.f90 -fno-diagnostics-show-caret -O1 -pedantic-errors -B/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs -L/ho
me/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs -L/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs -lm -m32 -o ./advance_1.exe
PASS: gfortran.dg/advance_1.f90  -O1  (test for excess errors)
Setting LD_LIBRARY_PATH to .:/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs:/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs:/home/meissner/fsf-build-
ppc64/meissner-lra/gcc:/home/meissner/fsf-build-ppc64/meissner-lra/gcc/32:.:/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu/32/libgfortran/.libs:/home/meissner/fsf-build-ppc64/meissner-lra/powerpc64-unknown-linux-gnu
/32/libgfortran/.libs:/home/meissner/fsf-build-ppc64/meissner-lra/gcc:/home/meissner/fsf-build-ppc64/meissner-lra/gcc/32:/home/meissner/tools/ppc64/lib:/home/meissner/tools/ppc32/lib:/home/meissner/tools-binutils/ppc64/lib:/home/meissner/to
ols-binutils/ppc32/lib
spawn [open ...]

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.


-- 
Michael Meissner, IBM
IBM, M/S 2506R, 550 King Street, Littleton, MA 01460, USA
email: meissner@linux.vnet.ibm.com, phone: +1 (978) 899-4797


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]