[Bug c++/44473] iterators already defined for std::vector when using std::decimal
--- Comment #7 from janis at gcc dot gnu dot org 2010-06-10 21:58 --- The new decimal* classes are sometimes treated as classes, sometimes as scalars. From a first look, something might be going wrong with the use of __are_same in bits/std_iterator.h so that it needs special casing for these odd classes. I'll continue digging slowly, but don't let that stop anyone else from jumping in here. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44473
[Bug objc++/23613] obj-c++.dg/isa-field-1.mm fails with the GNU runtime
--- Comment #8 from janis at gcc dot gnu dot org 2010-03-31 15:48 --- Subject: Bug 23613 Author: janis Date: Wed Mar 31 15:47:42 2010 New Revision: 157872 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157872 Log: 2010-03-31 Iain Sandoe iain.san...@sandoe-acoustics.co.uk PR objc++/23613 * gcc/testsuite/objc-obj-c++-shared/next-mapping.h (CLASSPTRFIELD): New macro. * gcc/testsuite/obj-c++.dg/isa-field-1.mm: Use new CLASSPTRFIELD macro. Remove dg-xfail-if. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/obj-c++.dg/isa-field-1.mm trunk/gcc/testsuite/objc-obj-c++-shared/next-mapping.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23613
[Bug objc/35165] Massive failures of objc on i686-apple-darwin9
--- Comment #14 from janis at gcc dot gnu dot org 2010-03-31 15:51 --- Subject: Bug 35165 Author: janis Date: Wed Mar 31 15:50:45 2010 New Revision: 157873 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157873 Log: 2010-03-31 Iain Sandoe iain.san...@sandoe-acoustics.co.uk PR testsuite/35165 * obj-c++.dg/try-catch-9.mm: Don't XFAIL m32 NeXT runtime. * obj-c++.dg/try-catch-2.mm: Ditto. * obj-c++.dg/lookup-2.mm: Ditto. * obj-c++.dg/encode-8.m: Ditto. * obj-c++.dg/cxx-ivars-2.mm: Ditto. * obj-c++.dg/cxx-ivars-3.mm: Skip for GNU, XFail for m64 NeXT. * obj-c++.dg/const-str-10.mm: Skip for GNU, match .quad for m64 NeXT. * obj-c++.dg/const-str-11.mm: Ditto. * obj-c++.dg/const-str-9.mm: Ditto. * obj-c++.dg/bitfield-3.mm: Skip for GNU runtime. * obj-c++.dg/bitfield-2.mm: XFAIL run for m64 NeXT runtime. * obj-c++.dg/except-1.mm: Ditto. * obj-c++.dg/const-str-7.mm: Ditto. * obj-c++.dg/cxx-ivars-1.mm: Ditto. * obj-c++.dg/const-str-3.mm: Ditto. * obj-c++.dg/const-str-7.mm: Ditto. * obj-c++.dg/stubify-2.mm: Change dump file name. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/obj-c++.dg/bitfield-2.mm trunk/gcc/testsuite/obj-c++.dg/bitfield-3.mm trunk/gcc/testsuite/obj-c++.dg/const-str-10.mm trunk/gcc/testsuite/obj-c++.dg/const-str-11.mm trunk/gcc/testsuite/obj-c++.dg/const-str-3.mm trunk/gcc/testsuite/obj-c++.dg/const-str-7.mm trunk/gcc/testsuite/obj-c++.dg/const-str-9.mm trunk/gcc/testsuite/obj-c++.dg/cxx-ivars-1.mm trunk/gcc/testsuite/obj-c++.dg/cxx-ivars-2.mm trunk/gcc/testsuite/obj-c++.dg/cxx-ivars-3.mm trunk/gcc/testsuite/obj-c++.dg/encode-8.mm trunk/gcc/testsuite/obj-c++.dg/except-1.mm trunk/gcc/testsuite/obj-c++.dg/lookup-2.mm trunk/gcc/testsuite/obj-c++.dg/stubify-2.mm trunk/gcc/testsuite/obj-c++.dg/try-catch-2.mm trunk/gcc/testsuite/obj-c++.dg/try-catch-9.mm -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35165
[Bug testsuite/43512] [4.5 regression] Many objc test failures
--- Comment #2 from janis at gcc dot gnu dot org 2010-03-25 18:07 --- Iain, these tests pass if the tests are run after make install because then the objc header files are found in the install tree. I tested your patch but installed first, that's why I didn't see the failures. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43512
[Bug objc/35165] Massive failures of objc on i686-apple-darwin9
--- Comment #8 from janis at gcc dot gnu dot org 2010-03-25 22:25 --- Subject: Bug 35165 Author: janis Date: Thu Mar 25 22:25:05 2010 New Revision: 157731 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157731 Log: 2010-03-25 Iain Sandoe iain.san...@sandoe-acoustics.co.uk PR objc/35165 PR testsuite/43512 * objc-obj-c++-shared: New directory. * objc-obj-c++-shared/Object1-implementation.h: New file. * objc-obj-c++-shared/Object1.h: New file. * objc-obj-c++-shared/Protocol1.h: New file. * objc-obj-c++-shared/next-abi.h: New file. * objc-obj-c++-shared/next-mapping.h: New file. * objc/execute/next_mapping.h: Delete. * objc.dg/special/special.exp: For all targets run the tests with -fgnu-runtime, for darwin targets also run the tests with -fnext-runtime. * objc.dg/dg.exp: Ditto. * obj-c++.dg/dg.exp: Ditto. * objc/execute/forward-1.m: Use shared wrapper headers (Object1.h, Protocol1.h) and next-mapping.h as required. Amend testcase to include use of updated NeXT interface. * objc/execute/formal_protocol-5.m: Ditto. * objc/execute/protocol-isEqual-2.m: Ditto. * objc/execute/protocol-isEqual-4.m: Ditto. * objc/execute/class-11.m: Use shared wrapper headers (Object1.h, Protocol1.h) and next-mapping.h as required. * objc/execute/object_is_class.m: Ditto. * objc/execute/enumeration-1.m: Ditto. * objc/execute/class-13.m: Ditto. * objc/execute/formal_protocol-2.m: Ditto. * objc/execute/formal_protocol-4.m: Ditto. * objc/execute/class-1.m: Ditto. * objc/execute/bycopy-1.m: Ditto. * objc/execute/formal_protocol-6.m: Ditto. * objc/execute/bycopy-3.m: Ditto. * objc/execute/class-3.m: Ditto. * objc/execute/bf-11.m: Ditto. * objc/execute/class-5.m: Ditto. * objc/execute/bf-13.m: Ditto. * objc/execute/class-7.m: Ditto. * objc/execute/bf-15.m: Ditto. * objc/execute/class-9.m: Ditto. * objc/execute/bf-17.m: Ditto. * objc/execute/bf-19.m: Ditto. * objc/execute/IMP.m: Ditto. * objc/execute/exceptions/catchall-1.m: Ditto. * objc/execute/exceptions/trivial.m: Ditto. * objc/execute/exceptions/finally-1.m: Ditto. * objc/execute/exceptions/local-variables-1.m: Ditto. * objc/execute/exceptions/foward-1.m: Ditto. * objc/execute/bf-2.m: Ditto. * objc/execute/string1.m: Ditto. * objc/execute/bf-4.m: Ditto. * objc/execute/informal_protocol.m: Ditto. * objc/execute/string3.m: Ditto. * objc/execute/bf-6.m: Ditto. * objc/execute/bf-8.m: Ditto. * objc/execute/class-tests-1.h: Ditto. * objc/execute/protocol-isEqual-1.m: Ditto. * objc/execute/protocol-isEqual-3.m: Ditto. * objc/execute/_cmd.m: Ditto. * objc/execute/function-message-1.m: Ditto. * objc/execute/bf-20.m: Ditto. * objc/execute/bf-common.h: Ditto. * objc/execute/np-2.m: Ditto. * objc/execute/class-10.m: Ditto. * objc/execute/class-12.m: Ditto. * objc/execute/enumeration-2.m: Ditto. * objc/execute/class-14.m: Ditto. * objc/execute/encode-1.m: Ditto. * objc/execute/formal_protocol-1.m: Ditto. * objc/execute/formal_protocol-3.m: Ditto. * objc/execute/accessing_ivars.m: Ditto. * objc/execute/bycopy-2.m: Ditto. * objc/execute/class-2.m: Ditto. * objc/execute/bf-10.m: Ditto. * objc/execute/formal_protocol-7.m: Ditto. * objc/execute/root_methods.m: Ditto. * objc/execute/class-4.m: Ditto. * objc/execute/bf-12.m: Ditto. * objc/execute/class-6.m: Ditto. * objc/execute/bf-14.m: Ditto. * objc/execute/nested-func-1.m: Ditto. * objc/execute/class-8.m: Ditto. * objc/execute/private.m: Ditto. * objc/execute/bf-16.m: Ditto. * objc/execute/bf-18.m: Ditto. * objc/execute/load-3.m: Ditto. * objc/execute/compatibility_alias.m: Ditto. * objc/execute/bf-1.m: Ditto. * objc/execute/no_clash.m: Ditto. * objc/execute/bf-3.m: Ditto. * objc/execute/string2.m: Ditto. * objc/execute/bf-5.m: Ditto. * objc/execute/string4.m: Ditto. * objc/execute/bf-7.m: Ditto. * objc/execute/object_is_meta_class.m: Ditto. * objc/execute/bf-9.m: Ditto. * objc/execute/bf-21.m: Ditto. * objc/execute/cascading-1.m: Ditto. * objc/execute/trivial.m: Ditto. * objc/execute/np-1.m: Ditto. * objc/compile/trivial.m: Ditto. * objc/execute/class_self-2.m: Include stdlib.h. * objc/execute/forward-1.x: Do not XFAIL for 32bit powerpc-darwin. * objc.dg/desig-init-1.m: Use shared wrapper headers (Object1.h
[Bug testsuite/43512] [4.5 regression] Many objc test failures
--- Comment #6 from janis at gcc dot gnu dot org 2010-03-25 22:25 --- Subject: Bug 43512 Author: janis Date: Thu Mar 25 22:25:05 2010 New Revision: 157731 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157731 Log: 2010-03-25 Iain Sandoe iain.san...@sandoe-acoustics.co.uk PR objc/35165 PR testsuite/43512 * objc-obj-c++-shared: New directory. * objc-obj-c++-shared/Object1-implementation.h: New file. * objc-obj-c++-shared/Object1.h: New file. * objc-obj-c++-shared/Protocol1.h: New file. * objc-obj-c++-shared/next-abi.h: New file. * objc-obj-c++-shared/next-mapping.h: New file. * objc/execute/next_mapping.h: Delete. * objc.dg/special/special.exp: For all targets run the tests with -fgnu-runtime, for darwin targets also run the tests with -fnext-runtime. * objc.dg/dg.exp: Ditto. * obj-c++.dg/dg.exp: Ditto. * objc/execute/forward-1.m: Use shared wrapper headers (Object1.h, Protocol1.h) and next-mapping.h as required. Amend testcase to include use of updated NeXT interface. * objc/execute/formal_protocol-5.m: Ditto. * objc/execute/protocol-isEqual-2.m: Ditto. * objc/execute/protocol-isEqual-4.m: Ditto. * objc/execute/class-11.m: Use shared wrapper headers (Object1.h, Protocol1.h) and next-mapping.h as required. * objc/execute/object_is_class.m: Ditto. * objc/execute/enumeration-1.m: Ditto. * objc/execute/class-13.m: Ditto. * objc/execute/formal_protocol-2.m: Ditto. * objc/execute/formal_protocol-4.m: Ditto. * objc/execute/class-1.m: Ditto. * objc/execute/bycopy-1.m: Ditto. * objc/execute/formal_protocol-6.m: Ditto. * objc/execute/bycopy-3.m: Ditto. * objc/execute/class-3.m: Ditto. * objc/execute/bf-11.m: Ditto. * objc/execute/class-5.m: Ditto. * objc/execute/bf-13.m: Ditto. * objc/execute/class-7.m: Ditto. * objc/execute/bf-15.m: Ditto. * objc/execute/class-9.m: Ditto. * objc/execute/bf-17.m: Ditto. * objc/execute/bf-19.m: Ditto. * objc/execute/IMP.m: Ditto. * objc/execute/exceptions/catchall-1.m: Ditto. * objc/execute/exceptions/trivial.m: Ditto. * objc/execute/exceptions/finally-1.m: Ditto. * objc/execute/exceptions/local-variables-1.m: Ditto. * objc/execute/exceptions/foward-1.m: Ditto. * objc/execute/bf-2.m: Ditto. * objc/execute/string1.m: Ditto. * objc/execute/bf-4.m: Ditto. * objc/execute/informal_protocol.m: Ditto. * objc/execute/string3.m: Ditto. * objc/execute/bf-6.m: Ditto. * objc/execute/bf-8.m: Ditto. * objc/execute/class-tests-1.h: Ditto. * objc/execute/protocol-isEqual-1.m: Ditto. * objc/execute/protocol-isEqual-3.m: Ditto. * objc/execute/_cmd.m: Ditto. * objc/execute/function-message-1.m: Ditto. * objc/execute/bf-20.m: Ditto. * objc/execute/bf-common.h: Ditto. * objc/execute/np-2.m: Ditto. * objc/execute/class-10.m: Ditto. * objc/execute/class-12.m: Ditto. * objc/execute/enumeration-2.m: Ditto. * objc/execute/class-14.m: Ditto. * objc/execute/encode-1.m: Ditto. * objc/execute/formal_protocol-1.m: Ditto. * objc/execute/formal_protocol-3.m: Ditto. * objc/execute/accessing_ivars.m: Ditto. * objc/execute/bycopy-2.m: Ditto. * objc/execute/class-2.m: Ditto. * objc/execute/bf-10.m: Ditto. * objc/execute/formal_protocol-7.m: Ditto. * objc/execute/root_methods.m: Ditto. * objc/execute/class-4.m: Ditto. * objc/execute/bf-12.m: Ditto. * objc/execute/class-6.m: Ditto. * objc/execute/bf-14.m: Ditto. * objc/execute/nested-func-1.m: Ditto. * objc/execute/class-8.m: Ditto. * objc/execute/private.m: Ditto. * objc/execute/bf-16.m: Ditto. * objc/execute/bf-18.m: Ditto. * objc/execute/load-3.m: Ditto. * objc/execute/compatibility_alias.m: Ditto. * objc/execute/bf-1.m: Ditto. * objc/execute/no_clash.m: Ditto. * objc/execute/bf-3.m: Ditto. * objc/execute/string2.m: Ditto. * objc/execute/bf-5.m: Ditto. * objc/execute/string4.m: Ditto. * objc/execute/bf-7.m: Ditto. * objc/execute/object_is_meta_class.m: Ditto. * objc/execute/bf-9.m: Ditto. * objc/execute/bf-21.m: Ditto. * objc/execute/cascading-1.m: Ditto. * objc/execute/trivial.m: Ditto. * objc/execute/np-1.m: Ditto. * objc/compile/trivial.m: Ditto. * objc/execute/class_self-2.m: Include stdlib.h. * objc/execute/forward-1.x: Do not XFAIL for 32bit powerpc-darwin. * objc.dg/desig-init-1.m: Use shared wrapper headers (Object1.h
[Bug testsuite/42348] Syntax of dg-skip-if in two obj-c++ tests
--- Comment #10 from janis at gcc dot gnu dot org 2010-03-24 22:51 --- Subject: Bug 42348 Author: janis Date: Wed Mar 24 22:51:08 2010 New Revision: 157715 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157715 Log: 2010-03-24 Iain Sandoe iain.san...@sandoe-acoustics.co.uk PR testsuite/42348 * lib/target-supports.exp: Add support for ObjC/ObjC++ tools in standard tests. (check_effective_target_objc2): New proc. (check_effective_target_next_runtime): New proc. * lib/objc.exp: Determine which runtime is in force and support it. * lib/obj-c++.exp: Ditto. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/lib/obj-c++.exp trunk/gcc/testsuite/lib/objc.exp trunk/gcc/testsuite/lib/target-supports.exp -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42348
[Bug testsuite/41609] Torture tests do not check trivial.{m,mm} for each run case.
--- Comment #3 from janis at gcc dot gnu dot org 2010-03-24 22:53 --- Subject: Bug 41609 Author: janis Date: Wed Mar 24 22:52:47 2010 New Revision: 157716 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157716 Log: 2010-03-24 Iain Sandoe iain.san...@sandoe-acoustics.co.uk PR testsuite/41609 * lib/objc-torture.exp (objc-set-runtime-options): New. * objc/execute/execute.exp: Check runtime options on each pass. * objc/execute/exceptions/exceptions.exp: Ditto. * objc/compile/compile.exp: Ditto. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/lib/objc-torture.exp trunk/gcc/testsuite/objc/compile/compile.exp trunk/gcc/testsuite/objc/execute/exceptions/exceptions.exp trunk/gcc/testsuite/objc/execute/execute.exp -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41609
[Bug c/43374] ICE with __builtin_isinf() and _Decimal argument
--- Comment #4 from janis at gcc dot gnu dot org 2010-03-18 17:27 ---
The tests also fail on powerpc64-linux, although the first one gets the same
error with and without optimization.
elm3c105% cat 43374-1.c
int func(_Decimal32 v) {
return __builtin_isinf(v);
}
elm3c105% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c 43374-1.c
43374-1.c: In function func:
43374-1.c:2:25: internal compiler error: in prepare_float_lib_cmp, at
optabs.c:4391
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
elm3c105% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c -O 43374-1.c
43374-1.c: In function func:
43374-1.c:1:5: internal compiler error: in prepare_float_lib_cmp, at
optabs.c:4391
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
elm3c105% cat 43374-2.c
_Decimal64 func() {
return 9e384dd + 9e384dd;
}
elm3c105% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c 43374-2.c
43374-2.c:3:1: internal compiler error: in decimal_to_decnumber, at dfp.c:115
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
elm3c105% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c -O 43374-2.c
43374-2.c:3:1: internal compiler error: in decimal_to_decnumber, at dfp.c:115
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
The failures are the same for -m32 and -m64.
--
janis at gcc dot gnu dot org changed:
What|Removed |Added
CC||bergner at gcc dot gnu dot
||org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43374
[Bug testsuite/43363] [4.5 Regression] FAIL: g++.dg/ext/altivec-17.C
--- Comment #1 from janis at gcc dot gnu dot org 2010-03-15 21:27 --- Subject: Bug 43363 Author: janis Date: Mon Mar 15 21:27:23 2010 New Revision: 157467 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=157467 Log: PR testsuite/43363 * g++.dg/ext/altivec-17.C: Handle changes to error message. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/g++.dg/ext/altivec-17.C -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43363
[Bug tree-optimization/43349] New: very long compile time for gamess with -floop-interchange
Benchmark test 416.gamess from SPEC CPU2006 takes a very long time to compile with current trunk on powerpc64-linux (both -m32 and -m64) with -O2 -floop-interchange. The slowdown started between r157243 on 20100305 and r157293 on 20100308. Here's a minimized testcase that takes several minutes now, but much less than a second before the slowdown: SUBROUTINE BUG(A,B,X,Y,Z,N) IMPLICIT NONE DOUBLE PRECISION A(*),B(*),X(*),Y(*),Z(*) INTEGER N,J,K K = 0 DO J = 1,N K = K+1 X(K) = B(J+N*7) Y(K) = B(J+N*8) Z(K) = B(J+N*2) + A(J+N*2) K = K+1 X(K) = B(J+N*3) + A(J+N*3) Y(K) = B(J+N*9) + A(J) Z(K) = B(J+N*15) K = K+1 X(K) = B(J+N*4) + A(J+N*4) Y(K) = B(J+N*15) Z(K) = B(J+N*10) + A(J) K = K+1 X(K) = B(J+N*11) + A(J+N) Y(K) = B(J+N*5) + A(J+N*5) Z(K) = B(J+N*16) K = K+1 X(K) = B(J+N*16) Y(K) = B(J+N*6) + A(J+N*6) Z(K) = B(J+N*12) + A(J+N) K = K+1 X(K) = B(J+N*13) + A(J+N*2) Y(K) = B(J+N*17) Z(K) = B(J+N*7) + A(J+N*7) ENDDO RETURN END -- Summary: very long compile time for gamess with -floop- interchange Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: compile-time-hog Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43349
[Bug tree-optimization/43349] [graphite] very long compile time for gamess with -floop-interchange
--- Comment #3 from janis at gcc dot gnu dot org 2010-03-12 23:12 --- I can't find that parameter in trunk, just in the graphite branch. The branch is fine, it's trunk that's taking a long time. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43349
[Bug tree-optimization/43349] [graphite] very long compile time for gamess with -floop-interchange
--- Comment #5 from janis at gcc dot gnu dot org 2010-03-12 23:25 --- In response to comment #4, it looks like that change would fix the problem in trunk. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43349
[Bug target/43249] unsigned int arg with no prototype gets full 64-bit reg
--- Comment #1 from janis at gcc dot gnu dot org 2010-03-10 18:35 --- GCC 3.4 also has this bug. -- janis at gcc dot gnu dot org changed: What|Removed |Added CC||bergner at gcc dot gnu dot ||org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43249
[Bug target/43249] New: unsigned int arg with no prototype gets full 64-bit reg
For a function with no prototype that defines an argument as unsigned
int, GCC for powerpc64-linux with -m64 and optimization uses the
entire 64-bit argument register rather than masking off the lower
32 bits. This problem was discoverd with 176.gcc using options
-m64 -O2 -ftree-vectorize -maltivec -fdata-sections, which changed
the order of 32-element arrays so that accessing element 63 instead of
element 31 started getting garbage that no longer happened to work.
Here's a simple testcase that works with -O0 and fails with -O1.
It's useful to compare this with a a version whose argument is
unsigned short. For the GCC testsuite, just remove references to
printf.
---
extern int printf (const char *, ...);
int
__attribute__((noinline))
floor_log2_wide (x)
unsigned int x;
{
int log = -1;
while (x != 0)
log++,
x = 1;
return log;
}
int
main ()
{
int m;
unsigned int n = 4294967259U;
m = floor_log2_wide ((int)n);
printf (n = %u, m = %d\n, n, m);
if (m != 31)
__builtin_abort ();
return 0;
}
---
elm3c105% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -m64 -O0 bug.c ./a.out
n = 4294967259, m = 31
elm3c105% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -m64 -O1 bug.c ./a.out
n = 4294967259, m = 63
Aborted
I've tried compilers back to GCC 4.1 and they all have the same problem;
it's probably not a regression.
--
Summary: unsigned int arg with no prototype gets full 64-bit reg
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: wrong-code
Severity: normal
Priority: P3
Component: target
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43249
[Bug c++/2288] Variable declared in for-loop-header is in wrong scope
--- Comment #9 from janis at gcc dot gnu dot org 2010-02-23 19:07 --- Is there a reason this hasn't been fixed? If not, I'll try. -- janis at gcc dot gnu dot org changed: What|Removed |Added CC||janis at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=2288
[Bug tree-optimization/42640] wrong code for -ftree-loop-distribution in 175.vpr
--- Comment #3 from janis at gcc dot gnu dot org 2010-02-23 19:27 --- The minimized testcase and 175.vpr still fail on powerpc64-linux with -O2 -ftree-loop-distribution for both -m32 and -m64. With the confirmation in comment #2 this report ought to be moved from UNCONFIRMED to NEW. This bug is not a regression from 4.4 but it is a wrong-code bug and ought to be considered for fixing in 4.5. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42640
[Bug tree-optimization/42326] [4.5 Regression][graphite] segfault in tree-data-ref.c with Graphite building 200.sixtrack
--- Comment #8 from janis at gcc dot gnu dot org 2010-02-23 21:17 --- With today's trunk both testcases fail with -m32 but not -m64. sixtrack builds, but not applu. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42326
[Bug c++/2288] Variable declared in for-loop-header is in wrong scope
--- Comment #11 from janis at gcc dot gnu dot org 2010-02-23 21:54 --- In response to comment #10, there are several duplicates to similar bug PR18770 and demand for a fix from within my company. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=2288
[Bug tree-optimization/43097] New: ICE in rename_map_elt_info for Graphite, vect in 173.applu
GCC trunk gets an internal compiler error when building SPEC CPU2000 test 173.applu on powerpc64-linux with options -O2 -fgraphite-identity -ftree-vectorize for either -m32 or -m64. Here's a minimized testcase: subroutine foo (ldmx,ldmy,nx,ny,v) implicit real*8 (a-h, o-z) dimension v(5,ldmx,ldmy,*) dimension tmat(5,5) k = 2 do j = 2, ny-1 do i = 2, nx-1 do ip = 1, 4 do m = ip+1, 5 v(m,i,j,k) = v(m,i,j,k) * m end do end do do m = 5, 1, -1 do l = m+1, 5 v(m,i,j,k) = v(l,i,j,k) end do v(m,i,j,k) = m end do end do end do return end elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gfortran -c -O2 -fgraphite-identity -ftree-vectorize bug.f bug.f: In function foo: bug.f:1:0: internal compiler error: tree check: expected ssa_name, have var_decl in rename_map_elt_info, at sese.c:81 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The benchmark and testcase compile cleanly with GCC 4.4.2. The failure begins with this patch: http://gcc.gnu.org/viewcvs?view=revrev=156711 r156711 | spop | 2010-02-11 19:42:38 + (Thu, 11 Feb 2010) -- Summary: ICE in rename_map_elt_info for Graphite, vect in 173.applu Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43097
[Bug target/42431] [4.5 Regression] wrong code for 200.sixtrack with vectorization and -fdata-sections
--- Comment #5 from janis at gcc dot gnu dot org 2010-02-12 20:35 --- Created an attachment (id=19853) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19853action=view) minimized executable testcase Compile with -m64 -O2 -maltivec -ftree-vectorize -fdata-sections. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42431
[Bug target/42431] [4.5 Regression] wrong code for 200.sixtrack with vectorization and -fdata-sections
--- Comment #6 from janis at gcc dot gnu dot org 2010-02-12 20:36 --- In the attached testcase wrong code is generated for the loop in subroutine sub3 that sets 12 of the elements of array k to zero. While minimizing the testcase I saw different kinds of bad code for that array, but with this version it's: vxor 31,31,31 ld 9,@toc(2) ld 29,@toc(2) li 0,16 stvx 31,0,29 add 9,9,9 stvx 31,0,9 stvx 31,9.0 Needless to say, doubling the array address has bad consequences. It would make sense if the instruction were instead add 9,29,0 I'm planning to stare at dump files for awhile but won't complain if someone else figures it out first. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42431
[Bug tree-optimization/43023] New: missing SSA_NAME def for -ftree-loop-distribution in 459.GemsFDTD
GCC trunk gets an internal compiler error when building SPEC CPU2006 test 459.GemsFDTD on powerpc64-linux with -O2 -ftree-loop-distribution for either -m32 or -m64, as demonstrated by this minimized testcase: - MODULE NFT_mod implicit none integer :: Nangle real:: Z0 real, dimension(:,:), allocatable :: Angle real, dimension(:), allocatable :: exth, ezth, hxth, hyth, hyphi CONTAINS SUBROUTINE NFT_Init() real :: th, fi integer :: n do n = 1,Nangle th = Angle(n,1) fi = Angle(n,2) exth(n) = cos(fi)*cos(th) ezth(n) = -sin(th) hxth(n) = -sin(fi) hyth(n) = cos(fi) hyphi(n) = -sin(fi) end do END SUBROUTINE NFT_Init END MODULE NFT_mod - elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gfortran -c -O2 -ftree-loop-distribution bug.f90 bug.f90: In function nft_init: bug.f90:11:0: error: missing definition for SSA_NAME: D.772_61 in statement: # .MEM_98 = VDEF .MEM_83 (*pretmp.25_71)[D.772_61] = D.775_97; bug.f90:11:0: internal compiler error: verify_ssa failed Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The testcase and GemsFTDT compile correctly with GCC 4.4.2. The failure starts with this patch: http://gcc.gnu.org/viewcvs?view=revrev=151761 r151761 | matz | 2009-09-16 16:12:18 + (Wed, 16 Sep 2009) -- Summary: missing SSA_NAME def for -ftree-loop-distribution in 459.GemsFDTD Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43023
[Bug middle-end/42344] [4.5 Regression] ICE in rs6000.md with ipa-sra for 252.eon
--- Comment #7 from janis at gcc dot gnu dot org 2010-02-10 22:48 --- Alan, do you have an update on this? Does you patch fix just the ICE or also the runtime segfault described in comment #2? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42344
[Bug middle-end/42344] [4.5 Regression] ICE in rs6000.md with ipa-sra for 252.eon
--- Comment #9 from janis at gcc dot gnu dot org 2010-02-11 00:50 --- I tried the patch and it fixes both of the problems with eon. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42344
[Bug tree-optimization/43026] New: ICE in sese.c with -fgraphite-identity in 447.dealII
GCC trunk gets an internal compiler error when building SPEC CPU2006 test 447.dealII on powerpc64-linux with -O2 -fgraphite-identity or -O2 -floop-parallelize-all for either -m32 or -m64, as demonstrated by a minimized testcase that I'll attach to this PR. Perhaps someone else will be able to minimize it more. elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/g++ -c -O2 -fgraphite-identity bug.cc bug.cc: In member function void Basedim::getf2d(vectorTensor2, dim ) const [with int dim = 3]: bug.cc:38:6: internal compiler error: in expand_scalar_variables_expr, at sese.c:907 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The testcase and dealII compile correctly with GCC 4.4.2 using -O2 -fgraphite-identity. The failure starts with this patch: http://gcc.gnu.org/viewcvs?view=revrev=155739 r155739 | spop | 2010-01-08 16:07:18 + (Fri, 08 Jan 2010) -- Summary: ICE in sese.c with -fgraphite-identity in 447.dealII Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43026
[Bug tree-optimization/43026] ICE in sese.c with -fgraphite-identity in 447.dealII
--- Comment #1 from janis at gcc dot gnu dot org 2010-02-11 01:02 --- Created an attachment (id=19837) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19837action=view) minimized testcase -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43026
[Bug tree-optimization/43012] New: wrong code for -floop-strip-mine in 453.povray
SPEC CPU2006 test 453.povray is miscompiled by GCC mainline on powerpc64-linux
for -O2 -floop-strip-line with either -m32 or -m64. Here's the function that
is miscompiled; I'll attach a complete executable testcase:
void
pre_init_tokenizer ()
{
int i;
for (i = 0; i LAST_TOKEN; i++)
{
Table[i] = i;
if (i FLOAT_FUNCT_TOKEN)
Table[i] = FLOAT_FUNCT_TOKEN;
else
{
if (i VECTOR_FUNCT_TOKEN)
Table[i] = VECTOR_FUNCT_TOKEN;
else
{
if (i COLOUR_KEY_TOKEN)
Table[i] = COLOUR_KEY_TOKEN;
}
}
}
}
Array elements 51-63 and 89-101 are not assigned values when the code is
compiled with -O2 -floop-strip-mine.
The testcase (and 453.povray) work with GCC 4.4.2 with those options, so this
is a regression in GCC 4.5. The failure starts with this patch:
http://gcc.gnu.org/viewcvs?view=revrev=154637
r154637 | spop | 2009-11-25 05:30:35 + (Wed, 25 Nov 2009)
--
Summary: wrong code for -floop-strip-mine in 453.povray
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: wrong-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43012
[Bug tree-optimization/43012] wrong code for -floop-strip-mine in 453.povray
--- Comment #1 from janis at gcc dot gnu dot org 2010-02-09 21:29 --- Created an attachment (id=19832) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19832action=view) minimized executable testcase -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43012
[Bug target/42431] [4.5 Regression] wrong code for 200.sixtrack with vectorization and -fdata-sections
--- Comment #4 from janis at gcc dot gnu dot org 2010-02-09 22:38 --- Peter and Mike, can one of you please look at this bug? It looks like something that you could figure out without a minimized testcase, but if it helps I'll come up with one. I can also do a regression hunt if that would help, but I'd have to set things up on a new machine. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42431
[Bug target/42416] ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
--- Comment #6 from janis at gcc dot gnu dot org 2010-02-08 21:13 --- Fixed by the patch in comment #5. -- janis at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug c/39037] FLOAT_CONST_DECIMAL64 pragma not supported
--- Comment #4 from janis at gcc dot gnu dot org 2010-01-13 23:15 --- This is fixed for GCC 4.5.0. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39037
[Bug c/39034] Decimal floating-point math done wrong
--- Comment #6 from janis at gcc dot gnu dot org 2010-01-13 23:18 --- This was fixed in the GCC 4.3 branch on 2009-02-25. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39034
[Bug target/42416] ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
--- Comment #5 from janis at gcc dot gnu dot org 2010-01-11 20:29 --- Subject: Bug 42416 Author: janis Date: Mon Jan 11 20:29:40 2010 New Revision: 155816 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=155816 Log: PR target/42416 * config/rs6000/rs6000.c (rs6000_override_options): On targets that support VSX, warn for -mno-altivec if vsx is not disabled, and disable vsx. * gcc.target/powerpc/warn-1.c: New test. * gcc.target/powerpc/warn-2.c: New test. Added: trunk/gcc/testsuite/gcc.target/powerpc/warn-1.c trunk/gcc/testsuite/gcc.target/powerpc/warn-2.c Modified: trunk/gcc/ChangeLog trunk/gcc/config/rs6000/rs6000.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug middle-end/42245] ICE in verify_backedges for 197.parser with sel-sched
--- Comment #8 from janis at gcc dot gnu dot org 2010-01-11 22:10 --- I bootstrapped on powerpc64-linux c,c++,fortran with the earlier big selective-scheduling patch plus the patch in comment #7 of this PR and the patch in comment #4 of PR42246. The resulting GCC gets no failures for SPEC CPU2000 using several combinations of selective-scheduling options. Thanks for being so responsive! -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42245
[Bug lto/42696] New: lto1 error with -fsigned-char and char comparison
SPEC CPU2006 tests 464.h264ref and 482.sphinx3 fail to build on powerpc64-linux
with -m64 -lto -fsigned-char. For some reason we've added the option
-fsigned-char to the SPEC config files for those two tests; without it both
tests fail at runtime for -m64, but it doesn't affect the results for -m32.
The failure to build with -flto might be due to some problem with -fsigned-char
on powerpc64, rather than a problem with LTO.
Here's a minimized testcase:
-
extern char *a;
extern int f;
void
foo (void)
{
f = (*a != '-');
}
-
And here's what happens:
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -m64 -flto -fsigned-char
bug.c -c
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -m64 -flto -fsigned-char
bug.o -shared -o x.so
In function foo:
lto1: error: type mismatch in indirect reference
unnamed-signed:8
unnamed-unsigned:8
D.1228_2 = *a.0_1;
lto1: internal compiler error: verify_stmts failed
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
lto-wrapper: /home/janis/tools/gcc-trunk-anonsvn/bin/gcc returned 1 exit status
collect2: lto-wrapper returned 1 exit status
If the type of 'a' is signed char then everything works fine.
--
Summary: lto1 error with -fsigned-char and char comparison
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: lto
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42696
[Bug target/42417] ICE in change_address_1 for 173.applu with -mvsx
--- Comment #2 from janis at gcc dot gnu dot org 2010-01-08 18:01 --- *** This bug has been marked as a duplicate of 42416 *** -- janis at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||DUPLICATE http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42417
[Bug target/42416] ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
--- Comment #3 from janis at gcc dot gnu dot org 2010-01-08 18:01 --- *** Bug 42417 has been marked as a duplicate of this bug. *** -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug target/42416] ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
-- janis at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|unassigned at gcc dot gnu |janis at gcc dot gnu dot org |dot org | Status|UNCONFIRMED |ASSIGNED Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2010-01-08 18:01:42 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug target/42419] ICE in reload_cse_simplify_operands for 254.gap with -mcpu=power7
--- Comment #4 from janis at gcc dot gnu dot org 2010-01-08 18:02 --- *** This bug has been marked as a duplicate of 42416 *** -- janis at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||DUPLICATE http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42419
[Bug target/42416] ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
--- Comment #4 from janis at gcc dot gnu dot org 2010-01-08 18:02 --- *** Bug 42419 has been marked as a duplicate of this bug. *** -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug tree-optimization/42248] compat test struct-by-value-17 fails execution with -O1 -fschedule-insns
--- Comment #2 from janis at gcc dot gnu dot org 2010-01-07 21:47 ---
The testcase still fails with current mailing when compiled with -m64 -O2 or
-m64 -O1 -fschedule-insns. The testcase is missing the piece that makes it
useful as an executable test:
--
int
main ()
{
init (g1s, (_Complex double)1);
check (g1s, (_Complex double)1);
return 0;
}
--
The tests in gcc.dg/compat are not normally compiled with -O2 so it doesn't
show up as a failure in test reports.
--
janis at gcc dot gnu dot org changed:
What|Removed |Added
CC||bergner at gcc dot gnu dot
||org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42248
[Bug tree-optimization/42326] [4.5 Regression] segfault in tree-data-ref.c with Graphite building 200.sixtrack
--- Comment #7 from janis at gcc dot gnu dot org 2010-01-07 22:07 --- I still see both failures with -m32 on powerpc64-linux but not with -m64. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42326
[Bug tree-optimization/42637] New: wrong code for -floop-interchange -ftree-loop-distribution
As mentioned in PR 42334, 200.sixtrack from SPEC CPU2000 started getting wrong answers on powerpc64-linux with the Graphite merge at r140301 when compiled with -O2 -floop-interchange -ftree-loop-distribution. The loop that is miscompiled is: real*8 rt(6,6),r(6,6),rtt(6,6) do i=1,6 do j=1,6 do ia=1,6 rtt(i,ia)=rt(i,j)*r(j,ia)+rtt(i,ia) end do end do end do I'll attach a complete executable testcase that demonstrates the failure. This testcase, and 200.sixtrack, pass when compiled with these options with GCC 4.4.2 and with trunk r150248 (the last trunk revision before r150301 that builds on powerpc64-linux). -- Summary: wrong code for -floop-interchange -ftree-loop- distribution Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42637
[Bug tree-optimization/42637] wrong code for -floop-interchange -ftree-loop-distribution
--- Comment #1 from janis at gcc dot gnu dot org 2010-01-06 18:27 --- Created an attachment (id=19488) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19488action=view) executable testcase -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42637
[Bug middle-end/42245] ICE in verify_backedges for 197.parser with sel-sched
--- Comment #6 from janis at gcc dot gnu dot org 2010-01-06 18:44 ---
With the patch at http://gcc.gnu.org/ml/gcc-patches/2009-12/msg01209.html the
testcase in the submitter's description no longer fails, but there is other
code in 197.parser that gets the same ICE with the same options. Here's a
minimized testcase for that:
int
strictly_smaller_name (char *s, char *t)
{
int ss, tt;
while ((*s != '\0') || (*t != '\0'))
{
if (*s == '\0')
ss = '*';
else
ss = *s++;
if (*t != '\0')
tt = *t;
if (ss == tt)
return 0;
}
}
This failure occurs with and without the patch cited above.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42245
[Bug rtl-optimization/42246] ICE in init_seqno for 186.crafty with sel-sched
--- Comment #2 from janis at gcc dot gnu dot org 2010-01-06 18:48 --- The patch at http://gcc.gnu.org/ml/gcc-patches/2009-12/msg01209.html fixes the testcase in the submitter's description, but there is code in CPU2000 test 200.sixtrack that triggers the same ICE with the same set of options, as shown by this minimized testcase: subroutine distance(x,clo) implicit real*8 (a-h,o-z) dimension x(2,6),x1(2,6),clo(6) do 60 i=1,2 do 20 j=1,6 x(i,j)=clo(j) 20 continue do 40 iq=1,6 x1(i,iq)=0.0d0 40 continue do 50 j=1,6 x(i,j)=x1(i,j) 50 continue 60 continue return end This testcase gets the ICE with and without the patch cited above. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42246
[Bug rtl-optimization/42246] ICE in init_seqno for 186.crafty with sel-sched
--- Comment #3 from janis at gcc dot gnu dot org 2010-01-06 18:52 --- I should have mentioned in comment #2 that the ICE for sixtrack only happens with -m32, not -m64. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42246
[Bug tree-optimization/42640] New: wrong code for -ftree-loop-distribution in 175.vpr
Benchmark test 175.vpr from SPEC CPU2000 gets incorrect results on
powerpc64-linux when compiled with -O2 -ftree-loop-distribution, as shown by
a minimized testcase that I'll attach to this PR. The inner loop is:
inode = s_node;
for (iloop = 1; iloop = 2; iloop++)
{
rr_node[inode].a = i;
rr_node[inode].b = j;
rr_node[inode].c = ipad;
inode = p_node;
}
Expected output:
elm3b149% /home/janis/tools/gcc-trunk-anonsvn-32/bin/gcc -O2 bug.c ./a.out
rr_node[0] = 1 2 0
rr_node[1] = 1 2 1
rr_node[2] = 1 2 0
rr_node[3] = 1 2 1
Bad output:
elm3b149% /home/janis/tools/gcc-trunk-anonsvn-32/bin/gcc -O2
-ftree-loop-distribution bug.c ./a.out
rr_node[0] = 1 2 0
rr_node[1] = 1 2 1
rr_node[2] = 0 0 0
rr_node[3] = 0 0 1
The test also fails with GCC 4.4.2.
--
Summary: wrong code for -ftree-loop-distribution in 175.vpr
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: wrong-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42640
[Bug tree-optimization/42640] wrong code for -ftree-loop-distribution in 175.vpr
--- Comment #1 from janis at gcc dot gnu dot org 2010-01-06 23:11 --- Created an attachment (id=19493) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19493action=view) minimized executable testcase -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42640
[Bug tree-optimization/42644] New: wrong code for -fgraphite-identify in 183.equake
Test 183.equake from SPEC CPU2000 gets wrong results on powerpc64-linux (-m32 or -m64) when compiled with either -O2 -fgraphite-identity or -O2 -floop-parallelize-all. I'll attach an executable testcase that demonstrates the failure. Test 178.galgel fails with the same options but I haven't investigated it. The failure starts with the merge from the Graphite branch for r150301. With options -O2 -fgraphite-identity the minimized testcase and the two CPU2000 tests pass with r150248, the last revision before that merge that builds on powerpc64-linux. They also pass with GCC 4.4.2 with -O2 -fgraphite-identity. -- Summary: wrong code for -fgraphite-identify in 183.equake Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42644
[Bug tree-optimization/42644] wrong code for -fgraphite-identify in 183.equake
--- Comment #1 from janis at gcc dot gnu dot org 2010-01-07 01:43 --- Created an attachment (id=19495) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19495action=view) minimized executable testcase -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42644
[Bug target/42431] [4.5 Regression] wrong code for 200.sixtrack with vectorization and -fdata-sections
--- Comment #3 from janis at gcc dot gnu dot org 2010-01-04 18:06 --- No testcase yet, I was hoping to get bergner or meissner to look at it since they have access to the benchmark. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42431
[Bug target/42427] [4.5 Regression] invalid assembly code for 301.apsi for -fnon-call-exceptions
--- Comment #3 from janis at gcc dot gnu dot org 2010-01-04 22:12 --- I get the same error with mainline built today. I'm using a compiler that defaults to -m32; it works fine with -m64. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42427
[Bug target/42416] ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
--- Comment #2 from janis at gcc dot gnu dot org 2010-01-05 00:53 --- Currently the compiler explicitly allows -mvsx -mno-altivec. If those options are used together, should vsx and altivec both be turned on, or both turned off? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug target/42419] ICE in reload_cse_simplify_operands for 254.gap with -mcpu=power7
--- Comment #3 from janis at gcc dot gnu dot org 2009-12-22 18:51 --- The bug (which I didn't know from the documentation) is that -mvsx -mno-altivec is not a valid combination but is not reported as an error. Instead the compiler accepts it and then generates invalid code. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42419
[Bug target/41605] Static linking of libgcc/libgfortran/libstdc++ can cause inconsistent symbol resolution.
--- Comment #7 from janis at gcc dot gnu dot org 2009-12-22 22:11 --- Subject: Bug 41605 Author: janis Date: Tue Dec 22 22:11:27 2009 New Revision: 155408 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=155408 Log: 2009-12-22 Iain Sandoe iain.san...@sandoe-acoustics.co.uk gcc/ PR target/41605 * lib/g++.exp: Provide -B options to allow for link spec %s substitutions for static libraries. libstdc++-v3/ PR target/41605 * testsuite/libstdc++-abi/abi.exp: Provide -B options to allow for link spec %s substitutions for static libraries. * testsuite/lib/libstdc++.exp: Ditto libgomp/ PR target/41605 * testsuite/lib/libgomp.exp: Provide -B options to allow for link spec %s substitutions for static libraries. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/lib/g++.exp trunk/libgomp/ChangeLog trunk/libgomp/testsuite/lib/libgomp.exp trunk/libstdc++-v3/ChangeLog trunk/libstdc++-v3/testsuite/lib/libstdc++.exp trunk/libstdc++-v3/testsuite/libstdc++-abi/abi.exp -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41605
[Bug target/42427] New: invalid assembly code for 301.apsi for -fnon-call-exceptions
GCC trunk gets a ICE when building SPEC CPU2000 test 301.apsi with -O2
-fnon-call-exceptions -fpeel-loops -fexceptions, as demonstrated by this
minimized testcase:
SUBROUTINE TEST(HELP,WM,NZ)
IMPLICIT REAL*8 (A-H, O-Z)
REAL*8 HELP(NZ)
COMPLEX*16 WM(NZ)
DO K=1,NZ
ZNEW=ABS(WM(K))
ZOLD=HELP(K)
HELP(K)=ZNEW
ENDDO
RETURN
END
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gfortran -c -O2 -fexceptions
-fnon-call-exceptions -fpeel-loops bug.f
/tmp/ccZKTZkM.s: Assembler messages:
/tmp/ccZKTZkM.s:38: Error: syntax error; found `,' but expected `('
/tmp/ccZKTZkM.s:38: Error: junk at end of line: `,3
The assembly code here is:
lwz 4,4(9)
lwz 3,9,3 line 38
lwz 5,8(9)
lwz 6,12(9)
bl cabs
The failure begins with this patch:
http://gcc.gnu.org/viewcvs?view=revrev=154688
r154688 | bernds | 2009-11-26 21:41:42 + (Thu, 26 Nov 2009)
I ran into this by running CPU2000 with a number of sets of options generated
by a tool called allpairs.
--
Summary: invalid assembly code for 301.apsi for -fnon-call-
exceptions
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: wrong-code
Severity: normal
Priority: P3
Component: target
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42427
[Bug target/42431] New: wrong code for 200.sixtrack with vectorization and -fdata-sections
When compiled with current trunk and options -m64 -O2 -ftree-vectorize -maltivec -fdata-sections, CPU2000 test 200.sixtrack gets wrong results and segfaults before finishing. I've reproduced the failure on POWER6 and POWER7 hardware. The source file that is miscompiled is midbloc6.f. If I add this line: if (icount1.eq.187) write (*,*) ' howdy' to any two of the DO loops associated with labels 20, 40, 50, 110, 200, 210, 220, 260, and 280 then the test runs to completion and gets the expected output. The test passes with these options with GCC 4.4.0. I don't currently have a reg hunt setup on a machine with AltiVec but can do that it if would be useful. -- Summary: wrong code for 200.sixtrack with vectorization and - fdata-sections Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: target AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42431
[Bug target/42431] wrong code for 200.sixtrack with vectorization and -fdata-sections
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-18 23:03 --- Oh yeah, 176.gcc fails with the same options. With test input it runs for a few minutes instead of a couple of seconds, and the results are wrong. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42431
[Bug target/42416] New: ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
GCC trunk gets a ICE when building SPEC CPU2000 test 177.mesa with -O2 -mvsx
-mno-altivec -ftree-vectorize, as demonstrated by this minimized testcase:
void
gl_xform_normals_3fv (unsigned int n, float v[][3], const float m[16],
float u[][3], unsigned char normalize)
{
unsigned int i;
float m0 = m[0], m4 = m[4], m8 = m[8];
float m1 = m[1], m5 = m[5], m9 = m[9];
float m2 = m[2], m6 = m[6], m10 = m[10];
for (i = 0; i n; i++)
{
float ux = u[i][0], uy = u[i][1], uz = u[i][2];
v[i][0] = ux * m0 + uy * m1 + uz * m2;
v[i][1] = ux * m4 + uy * m5 + uz * m6;
v[i][2] = ux * m8 + uy * m9 + uz * m10;
}
}
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -O2 -mvsx -mno-altivec
-ftree-vectorize -c bug.c
bug.c: In function gl_xform_normals_3fv:
bug.c:2:1: internal compiler error: in rs6000_builtin_vec_perm, at
config/rs6000/rs6000.c:3102
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
VSX support is new in GCC 4.5 so this is not a regression from earlier
releases.
--
Summary: ICE in rs6000_bulitin_vec_perm for 177.mesa with -mvsx
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: target
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42416
[Bug target/42417] New: ICE in change_address_1 for 173.applu with -mvsx
GCC trunk gets a ICE when building SPEC CPU2000 test 173.applu and several others with -O2 -mvsx -mno-altivec -ftree-vectorize, as demonstrated by this minimized testcase: subroutine ssor implicit real*8 (a-h,o-z) parameter (iar = 60) common/cgcon/ nx, ny, nz common/cvar/ u(5,iar,iar,iar), rsd(5,iar,iar,iar) common/ctscon/ itmax do istep = 1, itmax call jacld do k = 2, nz-1 do j = 2, ny-1 do i = 2, nx-1 do m = 1, 5 u( m, i, j, k ) = u( m, i, j, k ) $+ tmp * rsd( m, i, j, k ) end do end do end do end do end do return end elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gfortran -O2 -mvsx -mno-altivec -ftree-vectorize -c bug.f bug.f: In function ssor: bug.f:21:0: internal compiler error: in change_address_1, at emit-rtl.c:1912 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. VSX support is new in GCC 4.5 so this is not a regression from earlier releases. -- Summary: ICE in change_address_1 for 173.applu with -mvsx Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: target AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42417
[Bug target/42419] New: ICE in reload_cse_simplify_operands for 254.gap with -mcpu=power7
GCC trunk gets a ICE when building SPEC CPU2000 test 254.gap with -m64 -O2
-mcpu=power7 -mno-altivec -ftree-vectorize, as demonstrated by a rather large
minimized testcase that I'll attach to this PR.
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c -m64 -O2 -mcpu=power7
-mno-altivec -ftree-vectorize bug.c
bug.c: In function SyFgets:
bug.c:124:1: error: insn does not satisfy its constraints:
(insn 787 422 425 43 bug.c:87 (set (reg:TI 11 11)
(const_double 2314885530818453536 [0x2020202020202020]
2314885530818453536 [0x2020202020202020] 0 [0x0])) 375 {*movti_ppc64} (nil))
bug.c:124:1: internal compiler error: in reload_cse_simplify_operands, at
postreload.c:396
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
POWER7 support is new in GCC 4.5 so this is not a regression from earlier
releases.
--
Summary: ICE in reload_cse_simplify_operands for 254.gap with -
mcpu=power7
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: target
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: poewrpc64-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42419
[Bug target/42419] ICE in reload_cse_simplify_operands for 254.gap with -mcpu=power7
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-18 00:15 --- Created an attachment (id=19341) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19341action=view) Minimized testcase. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42419
[Bug rtl-optimization/42388] New: ICE in move_bb_info with sel-sched and modulo-sched for 176.gcc
GCC trunk gets a ICE when building SPEC CPU2000 test 176.gcc with -O2 -fmodulo-sched -fselective-scheduling as demonstrated by a minimized testcase that I'll attach to this PR. elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -O2 -fmodulo-sched -fselective-scheduling -c bug.c bug.c: In function unroll_loop: bug.c:63:1: internal compiler error: in move_bb_info, at sel-sched-ir.c:4968 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The ICE begins with this patch: http://gcc.gnu.org/viewcvs?view=revrev=153491 r153491 | rguenth | 2009-10-23 09:34:46 + (Fri, 23 Oct 2009) -- Summary: ICE in move_bb_info with sel-sched and modulo-sched for 176.gcc Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: rtl-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42388
[Bug rtl-optimization/42388] ICE in move_bb_info with sel-sched and modulo-sched for 176.gcc
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-16 00:09 --- Created an attachment (id=19316) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19316action=view) Minimized testcase. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42388
[Bug rtl-optimization/42388] ICE in move_bb_info with sel-sched and modulo-sched for 176.gcc
--- Comment #2 from janis at gcc dot gnu dot org 2009-12-16 00:16 --- With GCC 4.4.2 using the same options, the minimized testcase compiles cleanly and 176.gcc compiles and executes successfully. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42388
[Bug rtl-optimization/42389] New: ICE in advance_state_on_fence with sel-schd for 175.vpr
GCC trunk gets a ICE when building SPEC CPU2000 test 175.vpr with -O2 -fselective-scheduling -fsel-sched-pipelining -fsched-pressure as demonstrated by a minimized testcase that I'll attach to this PR. elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c -O2 -fselective-scheduling -fsel-sched-pipelining -fsched-pressure bug.c bug.c: In function try_route: bug.c:83:1: internal compiler error: in advance_state_on_fence, at sel-sched.c:5264 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The ICE begins with this patch: http://gcc.gnu.org/viewcvs?view=revrev=152291 r152291 | pthaugen | 2009-09-29 18:58:56 + (Tue, 29 Sep 2009) Option -fsched-pressure is new in GCC 4.5 so this is not a regression from 4.4. -- Summary: ICE in advance_state_on_fence with sel-schd for 175.vpr Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: rtl-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42389
[Bug rtl-optimization/42389] ICE in advance_state_on_fence with sel-schd for 175.vpr
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-16 00:19 --- Created an attachment (id=19317) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19317action=view) Minimized testcase. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42389
[Bug rtl-optimization/42294] [4.5 Regression] ICE in code_motion_path_driver for 416.gamess
--- Comment #6 from janis at gcc dot gnu dot org 2009-12-12 00:27 --- The testcase with the options shown in the submitter's description compiles cleanly with the current 4.4 branch and with GCC 4.4.1. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42294
[Bug middle-end/42344] [4.5 Regression] ICE in rs6000.md with ipa-sra for 252.eon
--- Comment #2 from janis at gcc dot gnu dot org 2009-12-10 17:25 ---
The ICE happens with either -m32 or -m64, but after this same patch eon
segfaults in a destructor when compiled with -m32 -O2 -fPIC when built with
GCC with secureplt enabled. valgrind reports some other invalid accesses
before that segfault. valgrind also complains about mismatched
free()/delete/delete[] in eon, which can be fixed by patching ggString.h with
@@ -89,7 +89,7 @@ private:
~StringRep();
void *operator new(size_t, unsigned long len);
void *operator new(size_t s) { return new char[s]; }
-//void operator delete(void *p) { delete p; }
+void operator delete(void *p) { delete [] (char *)p; }
int len;
int refs;
char str[1];
Peter, can you get someone to look at this further? I've got a setup to build
and run eon outside of the SPEC framework.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42344
[Bug tree-optimization/42344] New: ICE in rs6000.md with ipa-sra for 252.eon
GCC trunk gets a ICE when building SPEC CPU2000 test 252.eon with -O2 -fPIC
-fno-inline, but not with -fno-ipa-sra, on powerpc-linux with secureplt
enabled, as demonstrated by this minimized testcase:
class mrSurfaceList
{
public:
int *operator[] (int i) const { }
};
int *
makeGrid (mrSurfaceList * l)
{
return (*l)[0];
}
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/g++ -c -O2 -fPIC -fno-inline
bug.cc
bug.cc: In function int* makeGrid(mrSurfaceList*):
bug.cc:10:1: internal compiler error: in output_507, at
config/rs6000/rs6000.md:12527
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
The ICE begins with this patch, which adds IPA SRA:
http://gcc.gnu.org/viewcvs?view=revrev=151800
r151800 | jamborm | 2009-09-17 11:35:38 + (Thu, 17 Sep 2009)
--
Summary: ICE in rs6000.md with ipa-sra for 252.eon
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42344
[Bug tree-optimization/42334] New: segfault in graphite-poly.h for 200.sixtrack
GCC trunk gets a segmentation fault when building SPEC CPU2000 test 200.sixtrack with -floop-interchange -ftree-loop-distribution on powerpc-linux, as demonstrated by this minimized testcase: subroutine blockdis(bl1eg,bl2eg) implicit real*8 (a-h,o-z) parameter(nblo=300) common/str /mblo common/str2 /mel(nblo) dimension h(nblo,2,6),g(nblo,2,6) dimension bl1eg(nblo,2,6),bl2eg(nblo,2,6) do k=1,mblo jm=mel(k) do l=1,2 do m=1,6 bl1eg(k,l,m)=h(jm,l,m) bl2eg(k,l,m)=g(jm,l,m) enddo enddo enddo return end n function blockdis: bug.f:1:0: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions.See http://gcc.gnu.org/bugs.html for instructions. The segfault is in VEC_lst_p_base_iterate at graphite-poly.h:623. sixtrack passed with the now-failing options until r150248. With the merge from the Graphite branch at r150301 that test started failing at runtime. (In between those two revisions trunk didn't build with Graphite for powerpc-linux). sixtrack failed at runtime with r154631, and GCC started getting the segmentation fault at r154632. -- Summary: segfault in graphite-poly.h for 200.sixtrack Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42334
[Bug tree-optimization/42326] New: segfault in tree-data-ref.c with Graphite building 200.sixtrack
GCC trunk gets an internal compiler error when building SPEC CPU2000 test 200.sixtrack with -O2 -floop-parallelize-all -fprefetch-loops-arrays on powerpc-linux, as demonstrated by this minimized testcase: subroutine phasad(t,i,ium) implicit none real t(5,4) integer i,l,ll,ium do l=1,2 ll=2*l do i=1,ium t(i,ll-1)=t(i,ll-1)+t(i,ll) enddo enddo return end elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gfortran -c -O2 -floop-parallelize-all -fprefetch-loop-arrays bug.f bug.f: In function phasad: bug.f:1:0: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The failure has existed since the large Graphite merge that introduced -fgraphite-force-parallel which was later renamed to -floop-parallelize-all. The segmentation fault occurs in initialize_matrix_A at tree-data-ref.c:1914 when A[index] has the value (lambda_vector) 0xa. -- Summary: segfault in tree-data-ref.c with Graphite building 200.sixtrack Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42326
[Bug tree-optimization/42326] segfault in tree-data-ref.c with Graphite building 200.sixtrack
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-08 00:31 --- This testcase minimized from 173.applu fails in the same place when compiled with -m32 -O2 floop-strip-mine -fprefetch-loop-arrays on powerpc64-linux. subroutine blts ( ldmx, ldmy, v, tmp1, i, j, k) implicit none integer ldmx, ldmy, i, j, k, ip, m, l real*8 tmp, tmp1, v( 5, ldmx, ldmy, *), tmat(5,5) do ip = 1, 4 do m = ip+1, 5 tmp = tmp1 * tmat( m, ip ) do l = ip+1, 5 tmat( m, l ) = tmat( m, l ) - tmat( ip, l ) end do v( m, i, j, k ) = tmp end do end do return end -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42326
[Bug tree-optimization/42284] New: failing tree check in graphite-sese-to-poly.c for 164.gzip
GCC trunk gets an internal compiler error when building SPEC CPU2000 test 164.gzip with -O2 -floop-interchange on powerpc-linux, as demonstrated by a minimized testcase that I'll attach to this PR. elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -c -O2 -floop-interchange bug.c bug.c: In function inflate_fixed: bug.c:56:1: internal compiler error: tree check: expected ssa_name, have integer_cst in detect_commutative_reduction, at graphite-sese-to-poly.c:2654 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. A Graphite-enabled trunk build is broken during the 19-hour range from r150250 to 150300. The test passes with the build just before that range and fails with r150301, the first revision that builds after a merge from the Graphite branch. -- Summary: failing tree check in graphite-sese-to-poly.c for 164.gzip Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42284
[Bug tree-optimization/42284] failing tree check in graphite-sese-to-poly.c for 164.gzip
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-04 20:39 --- Created an attachment (id=19227) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19227action=view) Minimized testcase. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42284
[Bug tree-optimization/42285] New: ICE in Graphite's scan_tree_for_params for 416.gamess
GCC trunk gets an internal compiler error when building SPEC CPU2006 test 416.gamess with -O2 -floop-interchange on powerpc-linux, as demonstrated by this minimized testcase: SUBROUTINE EFGRDM(NCF,NFRG,G,RTRMS,GM,IOPT,K1) IMPLICIT DOUBLE PRECISION (A-H,O-Z) DIMENSION G(*),RTRMS(*),GM(*) DUM = 0 DO I=1,NFRG DO J=1,3 IF (IOPT.EQ.0) THEN GM(K1)=G(K1) END IF END DO DO J=1,3 JDX=NCF*9+IOPT*9*NFRG DO M=1,3 DUM=DUM+RTRMS(JDX+M) END DO GM(K1)=DUM END DO END DO RETURN END elm3b187% /opt/gcc-nightly/trunk/bin/gfortran -c -O2 -floop-interchange bug.f bug.f: In function efgrdm: bug.f:1:0: internal compiler error: in scan_tree_for_params, at graphite-sese-to-poly.c:815 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. A Graphite-enabled trunk build is broken during the 19-hour range from r150250 to 150300. The test passes with the build just before that range and fails with r150301, the first revision that builds after a merge from the Graphite branch. -- Summary: ICE in Graphite's scan_tree_for_params for 416.gamess Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42285
[Bug tree-optimization/42294] New: ICE in code_motion_path_driver for 416.gamess
GCC trunk gets an internal compiler error when building SPEC CPU2000 test 416.gamess with -O2 -fselective-scheduling2 -fsel-sched-pipelining -funroll-all-loops on powerpc64-linux, as demonstrated by a minimized testcase that I'll attach to this PR. elm3b187% /opt/gcc-nightly/trunk/bin/gfortran -c -O2 -fselective-scheduling2 -fsel-sched-pipelining -funroll-all-loops bug.f bug.f: In function orien: bug.f:35:0: internal compiler error: in code_motion_path_driver, at sel-sched.c:6532 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html for instructions. The failure starts with this very large patch (Expand from SSA): http://gcc.gnu.org/viewcvs?view=revrev=146817 r146817 | matz | 2009-04-26 19:35:04 + (Sun, 26 Apr 2009) -- Summary: ICE in code_motion_path_driver for 416.gamess Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: ice-on-valid-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42294
[Bug rtl-optimization/42294] ICE in code_motion_path_driver for 416.gamess
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-05 01:06 --- Created an attachment (id=19233) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19233action=view) Minimized testcase. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42294
[Bug rtl-optimization/42294] ICE in code_motion_path_driver for 416.gamess
--- Comment #2 from janis at gcc dot gnu dot org 2009-12-05 01:07 --- The ICE occurs with a compiler built for powerpc64-linux but only for -m32 (the default for my compiler), not for -m64. It does not occur for a powerpc-linux compiler. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42294
[Bug tree-optimization/42250] [4.5 Regression] segfault in ipa-type-escape.c for several cpu2000 tests
--- Comment #3 from janis at gcc dot gnu dot org 2009-12-02 18:10 --- The option -fipa-type-escape is not documented on its own, just with -fipa-struct-reorg which says it's needed with that option. Shouldn't broken options at least be documented as experimental? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42250
[Bug debug/42244] New: var-tracking ICE for 300.twolf
GCC trunk gets an internal compiler error when building SPEC CPU2000 test
300.twolf with -O2 -g -ffloat-store -fmodulo-sched on powerpc-linux, as
demonstrated by this minimized testcase:
extern int numerator, denom;
double
compute_and_combination (double states)
{
for (; numerator denom; numerator--)
states *= (double) numerator;
return states;
}
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -O2 -g -ffloat-store
-fmodulo-sched -c bug.c
bug.c: In function compute_and_combination:
bug.c:9:1: internal compiler error: in create_ddg_dep_no_link, at ddg.c:215
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
The failure started when var-tracking-assignments was enabled by default if
var-tracking was enabled; the test case passes with the additional option
-fno-var-tracking.
--
Summary: var-tracking ICE for 300.twolf
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: debug
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42244
[Bug tree-optimization/42245] New: ICE in verify_backedges for 197.parser with sel-sched
GCC trunk gets an internal compiler error when building SPEC CPU2000 test
197.parser with -O2 -fselective-scheduling -fsel-sched-pipelining on
powerpc-linux, as demonstrated by this minimized testcase:
---
extern int N_words;
typedef struct DIS_node_struct DIS_node;
typedef struct CON_list_struct CON_list;
struct DIS_node_struct
{
CON_list *cl;
};
void
build_DIS_CON_tree (void)
{
int w;
DIS_node *dnroot, *dn;
CON_list *child, *xchild;
for (w = 0; w N_words; w++)
{
if (dnroot == ((void *) 0))
{
dnroot = dn;
for (child = dn-cl; child != ((void *) 0); child = xchild)
{
}
}
}
}
---
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -O2
-fselective-scheduling -fsel-sched-pipelining -c bug.c
bug.c: In function build_DIS_CON_tree:
bug.c:26:1: internal compiler error: in verify_backedges, at
sel-sched-ir.c:3498
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
The failure starts with this very large patch (Expand from SSA):
http://gcc.gnu.org/viewcvs?view=revrev=146817
r146817 | matz | 2009-04-26 19:35:04 + (Sun, 26 Apr 2009)
--
Summary: ICE in verify_backedges for 197.parser with sel-sched
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42245
[Bug tree-optimization/42246] New: ICE in init_seqno for 186.crafty with sel-sched
GCC trunk gets an internal compiler error when building SPEC CPU2000 test
186.crafty with -O2 -fselective-scheduling -fsel-sched-pipelining
-fsel-sched-pipelining-outer-loops on powerpc-linux, as demonstrated by this
minimized testcase:
-
typedef enum
{
empty = 0, pawn = 1, knight = 2, king = 3, bishop = 5, rook = 6, queen = 7
}
PIECE;
extern int p_values[15];
extern int *last[65];
int
Quiesce (int alpha, int beta, int wtm, int ply)
{
register int initial_alpha, value, delta;
register int *goodmv, *movep, moves = 0, *sortv, temp;
for (movep = last[ply - 1]; movep last[ply]; movep++)
if (p_values[(((*movep) 15) 7) + 7] +
p_values[(((*movep) 18) 7) + 7] = delta)
{
register int done;
register int *end = last[ply - 1] + moves - 1;
do
{
done = 1;
movep = last[ply - 1];
for (; movep end; movep++, sortv++)
if (*sortv *(sortv + 1))
{
*(movep + 1) = temp;
done = 0;
}
}
while (!done);
}
}
-
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -O2
-fselective-scheduling -fsel-sched-pipelining
-fsel-sched-pipelining-outer-loops -c bug.c
bug.c: In function Quiesce:
bug.c:32:1: internal compiler error: in init_seqno, at sel-sched.c:6722
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
The failure started with this large patch:
http://gcc.gnu.org/viewcvs?view=revrev=145494
r145494 | rguenth | 2009-04-03 10:24:28 + (Fri, 03 Apr 2009)
--
Summary: ICE in init_seqno for 186.crafty with sel-sched
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42246
[Bug tree-optimization/42248] New: compat test struct-by-value-17 fails execution with -O1 -fschedule-insns
-- Summary: compat test struct-by-value-17 fails execution with -O1 -fschedule-insns Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42248
[Bug tree-optimization/42248] compat test struct-by-value-17 fails execution with -O1 -fschedule-insns
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-02 00:07 ---
Oops.
GCC trunk fails execution for five tests from gcc.dg/compat when compiled with
-m64 -O1 -fschedule-insns on powerpc64-linux. The tests are:
gcc.dg/compat/struct-by-value-16_main.c
gcc.dg/compat/struct-by-value-16a_main.c
gcc.dg/compat/struct-by-value-17_main.c
gcc.dg/compat/struct-by-value-17a_main.c
gcc.dg/compat/struct-by-value-18_main.c
This testcase is minimized from struct-by-value-17:
---
typedef struct {
_Complex double a;
_Complex double b;
_Complex double c;
_Complex double d;
_Complex double e;
_Complex double f;
_Complex double g;
_Complex double h;
_Complex double i;
_Complex double j;
} Scf10;
Scf10 g1s;
void
check (Scf10 x, _Complex double y)
{
if (x.a != y) __builtin_abort ();
if (x.b != y) __builtin_abort ();
if (x.c != y) __builtin_abort ();
if (x.d != y) __builtin_abort ();
if (x.e != y) __builtin_abort ();
if (x.f != y) __builtin_abort ();
if (x.g != y) __builtin_abort ();
if (x.h != y) __builtin_abort ();
if (x.i != y) __builtin_abort ();
if (x.j != y) __builtin_abort ();
}
void
init (Scf10 *p, _Complex double y)
{
p-a = y;
p-b = y;
p-c = y;
p-d = y;
p-e = y;
p-f = y;
p-g = y;
p-h = y;
p-i = y;
p-j = y;
}
---
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -m64 -O1 -fschedule-insns
-static bug.c
elm3b149% ./a.out
Aborted
The failure starts with this patch from Richard Guenther and Andrey
Belevantsev:
http://gcc.gnu.org/viewcvs?view=revrev=149624
r149624 | rguenth | 2009-07-14 09:59:18 + (Tue, 14 Jul 2009)
I ran into this due to a bug in lto.exp leaking compiler option lists to the
compat tests.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42248
[Bug tree-optimization/42249] New: unrecognizable insn for 254.gap with sel-sched
GCC trunk and 4.4 branch get internal compiler errors when building SPEC
CPU2000 test 254.gap with -O2 -fselective-scheduling -fsel-sched-pipelining
on powerpc-linux, as demonstrated by this minimized testcase:
---
typedef struct TypHeader
{
unsigned long size;
struct TypHeader **ptr;
} *TypHandle;
extern TypHandle NewBag (unsigned int type, unsigned long size);
typedef unsigned long TypPoint32;
TypHandle
QuoPP (TypHandle hdL, TypHandle hdR)
{
TypHandle hdC;
unsigned long degC;
TypPoint32 *ptC;
unsigned long degL;
TypPoint32 *ptL;
unsigned long degR;
TypPoint32 *ptR;
unsigned long p;
degL = ((hdL)-size) / sizeof (TypPoint32);
degR = ((hdR)-size) / sizeof (TypPoint32);
degC = degL degR ? degR : degL;
hdC = NewBag (9, (unsigned long) (degC * sizeof (TypPoint32)));
ptC = (TypPoint32 *) ((TypHandle *) ((hdC)-ptr));
for (p = 0; p degC; p++)
ptC[(((p) (degR)) ? (ptR)[(p)] : (p))] =
((p) (degL)) ? (ptL)[(p)] : (p)))
(degR)) ? (ptR)[p) (degL)) ? (ptL)[(p)] : (p)))]
: p) (degL)) ? (ptL)[(p)] : (p;
}
---
Trunk:
elm3b149% /home/janis/tools/gcc-trunk-anonsvn/bin/gcc -O2
-fselective-scheduling -fsel-sched-pipelining -c bug.c
bug.c: In function QuoPP:
bug.c:33:1: error: unrecognizable insn:
(insn 176 57 98 11 bug.c:31 (set (reg:CCUNS 199)
(reg:CCUNS 200)) -1 (nil))
bug.c:33:1: internal compiler error: in extract_insn, at recog.c:2097
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
4.4 branch:
elm3b149% /home/janis/tools/gcc-4.4-anonsvn/bin/gcc -O2 -fselective-scheduling
-fsel-sched-pipelining -c bug.c
bug.c: In function QuoPP:
bug.c:33: error: unrecognizable insn:
(insn 167 91 89 13 bug.c:31 (set (reg:CCUNS 179)
(reg:CCUNS 180)) -1 (nil))
bug.c:33: internal compiler error: in get_attr_type, at
config/rs6000/rs6000.md:13160
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructions.
--
Summary: unrecognizable insn for 254.gap with sel-sched
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42249
[Bug tree-optimization/42250] New: segfault in ipa-type-escape.c for several cpu2000 tests
GCC trunk gets a segmentation fault when building SPEC CPU2000 tests 164.gzip,
175.vpr, 176.gcc, 197.parser, 252.eon, 253.perlbmk, and 177.mesa with -O2
-fipa-type-escape on powerpc-linux. That option is meant to be used with
-fipa-struct-reorg, which is meant to be used with -combine and
-fwhole-program, and the same segfault happens with that whole set of options.
The segfault always happens in the same place, verified by testing with this
patch:
-
Index: gcc/ipa-type-escape.c
===
--- gcc/ipa-type-escape.c (revision 154833)
+++ gcc/ipa-type-escape.c (working copy)
@@ -1663,6 +1663,7 @@ analyze_function (struct cgraph_node *fn
struct function *this_cfun = DECL_STRUCT_FUNCTION (decl);
basic_block this_block;
+gcc_assert (this_cfun != NULL);
FOR_EACH_BB_FN (this_block, this_cfun)
{
gimple_stmt_iterator gsi;
-
The minimized test case is kind of large so I'll attach it.
The failure starts with this patch:
http://gcc.gnu.org/viewcvs?view=revrev=147294
r147294 | hubicka | 2009-05-08 19:19:51 + (Fri, 08 May 2009)
--
Summary: segfault in ipa-type-escape.c for several cpu2000 tests
Product: gcc
Version: 4.5.0
Status: UNCONFIRMED
Keywords: ice-on-valid-code
Severity: normal
Priority: P3
Component: tree-optimization
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: janis at gcc dot gnu dot org
GCC target triplet: powerpc-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42250
[Bug tree-optimization/42250] segfault in ipa-type-escape.c for several cpu2000 tests
--- Comment #1 from janis at gcc dot gnu dot org 2009-12-02 01:19 --- Created an attachment (id=19205) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19205action=view) minimized testcase -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42250
[Bug testsuite/42212] [4.5 Regression] ERROR: tcl error sourcing /Users/regress/tbox/svn-gcc/gcc/testsuite/gcc.target/powerpc/powerpc.exp.
--- Comment #2 from janis at gcc dot gnu dot org 2009-11-30 19:15 --- Subject: Bug 42212 Author: janis Date: Mon Nov 30 19:14:58 2009 New Revision: 154837 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154837 Log: PR testsuite/42212 * gcc.target/powerpc/regnames-1.c: Add missing brace dg-do. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.target/powerpc/regnames-1.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42212
[Bug testsuite/41878] [4.5 Regression] ERROR: gfortran.dg/vect/vect-2.f90
--- Comment #1 from janis at gcc dot gnu dot org 2009-11-02 19:32 --- Subject: Bug 41878 Author: janis Date: Mon Nov 2 19:32:05 2009 New Revision: 153821 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=153821 Log: PR testsuite/41878 * gfortran.dg/vect/vect-2.f90: Fix effective-target expression, add spaces in some expressions. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gfortran.dg/vect/vect-2.f90 -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41878
[Bug testsuite/41878] [4.5 Regression] ERROR: gfortran.dg/vect/vect-2.f90
--- Comment #2 from janis at gcc dot gnu dot org 2009-11-02 20:33 --- HJ, I couldn't reproduce the failure so please verify that the patch fixes the error. -- janis at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|unassigned at gcc dot gnu |janis at gcc dot gnu dot org |dot org | Status|UNCONFIRMED |ASSIGNED Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2009-11-02 20:33:22 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41878
[Bug middle-end/41812] New: test 20071030-1.c fails execution on powerpc64
Test gcc.c-torture/execute/20071030-2.c has been failing execution for -m64 with option lists including -O2 and -O3 since this patch was added to trunk: http://gcc.gnu.org/viewcvs?view=revrev=149011 r149011 | bonzini | 2009-06-27 14:50:00 + (Sat, 27 Jun 2009) That patch was added shortly before another one that broke bootstrap, as reported in PR 40597, and this test failure was then lost in the noise. I'll attach a shorter testcase that demonstrates the problem. -- Summary: test 20071030-1.c fails execution on powerpc64 Product: gcc Version: 4.5.0 Status: UNCONFIRMED Keywords: wrong-code Severity: normal Priority: P3 Component: middle-end AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: janis at gcc dot gnu dot org GCC target triplet: powerpc64-unknown-linux-gnu http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41812
[Bug middle-end/41812] test 20071030-1.c fails execution on powerpc64
--- Comment #1 from janis at gcc dot gnu dot org 2009-10-23 23:50 --- Created an attachment (id=18884) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18884action=view) minimized testcase Minimized testcase. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41812
[Bug c/41049] conversion from integer to decimal float loses trailing zeros
--- Comment #5 from janis at gcc dot gnu dot org 2009-10-22 18:24 --- Subject: Bug 41049 Author: janis Date: Thu Oct 22 18:24:31 2009 New Revision: 153468 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=153468 Log: Backport from 4.3 branch: 2009-09-21 Janis Johnson janis...@us.ibm.com PR c/41049 * real.c decimal_from_integer, decimal_integer_string): New. (real_from_integer): Use them as special case for decimal float. * config/dfp-bit.c (_si_to_sd, _usi_to_sd): Use default rounding. (_di_to_sd, _di_to_dd, _di_to_td, _udi_to_sd, _udi_to_dd, _udi_to_td): Do not append zero after the decimal point in string to convert. * gcc.dg/dfp/pr41049.c: New test. Added: branches/ibm/gcc-4_3-branch/gcc/testsuite/gcc.dg/dfp/pr41049.c Modified: branches/ibm/gcc-4_3-branch/gcc/ChangeLog.ibm branches/ibm/gcc-4_3-branch/gcc/config/dfp-bit.c branches/ibm/gcc-4_3-branch/gcc/real.c branches/ibm/gcc-4_3-branch/gcc/testsuite/ChangeLog.ibm -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41049
[Bug testsuite/41659] FAIL: gcc.dg/lto/20090914-2 c_lto_20090914-2_0.o assemble, -O0 -fwhopr
--- Comment #2 from janis at gcc dot gnu dot org 2009-10-12 16:06 --- It doesn't have the desired or expected effect, just the documented effect. From the GCC internals manual section 6.4.6 Support for testing link-time optimizations: Unlike dg-do, dg-lto-do does not support an optional `target' or `xfail' list. Use dg-skip-if, dg-xfail-if, or dg-xfail-run-if. That's rather obscure. For now I'll make it an error and then later maybe I'll have time to support target and xfail for dg-lto-do. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41659
[Bug testsuite/41659] FAIL: gcc.dg/lto/20090914-2 c_lto_20090914-2_0.o assemble, -O0 -fwhopr
--- Comment #3 from janis at gcc dot gnu dot org 2009-10-12 17:13 --- Subject: Bug 41659 Author: janis Date: Mon Oct 12 17:13:41 2009 New Revision: 152671 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=152671 Log: PR testsuite/41659 * gcc.dg/lto/20090914-2.c: Use dg-skip-if to skip test. * lib/lto.exp (lto-get-options-main): Report error for using target or xfail with dg-lto-dg. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.dg/lto/20090914-2_0.c trunk/gcc/testsuite/lib/lto.exp -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41659
[Bug c/41049] conversion from integer to decimal float loses trailing zeros
--- Comment #4 from janis at gcc dot gnu dot org 2009-09-22 16:16 --- Fixed in trunk, 4.4 branch, 4.3 branch. -- janis at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41049
[Bug c/41049] conversion from integer to decimal float loses trailing zeros
--- Comment #3 from janis at gcc dot gnu dot org 2009-09-21 16:22 --- Subject: Bug 41049 Author: janis Date: Mon Sep 21 16:22:43 2009 New Revision: 151934 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=151934 Log: PR c/41049 * real.c decimal_from_integer, decimal_integer_string): New. (real_from_integer): Use them as special case for decimal float. * config/dfp-bit.c (_si_to_sd, _usi_to_sd): Use default rounding. (_di_to_sd, _di_to_dd, _di_to_td, _udi_to_sd, _udi_to_dd, _udi_to_td): Do not append zero after the decimal point in string to convert. * gcc.dg/dfp/pr41049.c: New test. Added: branches/gcc-4_3-branch/gcc/testsuite/gcc.dg/dfp/pr41049.c Modified: branches/gcc-4_3-branch/gcc/ChangeLog branches/gcc-4_3-branch/gcc/config/dfp-bit.c branches/gcc-4_3-branch/gcc/real.c branches/gcc-4_3-branch/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41049
