OpenE2K
/
gcc
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

re PR target/23602 (1081 test failures in libjava, when configured for i486-linux) 

PR target/23602
        * toplev.c (process_options): Warn about unsupported combinations
        of unwind tables and omit-frame-pointer.
        * config/i386/i386.c (override_options): Similarly.  Enable
        accumulate-outgoing-args if not explicitly disabled.
testsuite/
        * g++.dg/eh/omit-frame-pointer.C: Remove i?86 specific options.
        * g++.dg/eh/omit-frame-pointer2.C: Likewise.
libjava/
        * configure.host (i?86-*): Set -fomit-frame-pointer in libgcj_flags,
        but not BACKTRACESPEC.
        (x86_64-*): Similarly.  Don't set -ffloat-store in 64-bit mode.

From-SVN: r105009
This commit is contained in:
Richard Henderson 2005-10-05 11:19:26 -07:00 committed by Richard Henderson
parent 10dc6e8cea
commit d3073c70c5
8 changed files with 60 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
gcc/ChangeLog
View File

@ -1,3 +1,11 @@
2005-10-05 Richard Henderson <rth@redhat.com>
PR target/23602
* toplev.c (process_options): Warn about unsupported combinations
of unwind tables and omit-frame-pointer.
* config/i386/i386.c (override_options): Similarly. Enable
accumulate-outgoing-args if not explicitly disabled.
2005-10-05 Steve Ellcey <sje@cup.hp.com>
* vect.md (vec_initv2si): Fix typo of V2SF to V2SI.

15
gcc/config/i386/i386.c
View File

@ -1653,6 +1653,21 @@ override_options (void)
&& !optimize_size)
target_flags |= MASK_ACCUMULATE_OUTGOING_ARGS;
/* ??? Unwind info is not correct around the CFG unless either a frame
pointer is present or M_A_O_A is set. Fixing this requires rewriting
unwind info generation to be aware of the CFG and propagating states
around edges. */
if ((flag_unwind_tables || flag_asynchronous_unwind_tables
|| flag_exceptions || flag_non_call_exceptions)
&& flag_omit_frame_pointer
&& !(target_flags & MASK_ACCUMULATE_OUTGOING_ARGS))
{
if (target_flags_explicit & MASK_ACCUMULATE_OUTGOING_ARGS)
warning (0, "unwind tables currently require either a frame pointer "
"or -maccumulate-outgoing-args for correctness");
target_flags |= MASK_ACCUMULATE_OUTGOING_ARGS;
}
/* Figure out what ASM_GENERATE_INTERNAL_LABEL builds as a prefix. */
{
char *p;

6
gcc/testsuite/ChangeLog
View File

@ -1,3 +1,9 @@
2005-10-05 Richard Henderson <rth@redhat.com>
PR target/23602
* g++.dg/eh/omit-frame-pointer.C: Remove i?86 specific options.
* g++.dg/eh/omit-frame-pointer2.C: Likewise.
2005-10-05 Dale Johannesen <dalej@apple.com>
* gcc.c-torture/execute/float-floor.c: New.

1
gcc/testsuite/g++.dg/eh/omit-frame-pointer.C
View File

@ -1,6 +1,5 @@
// { dg-do run }
// { dg-options -fomit-frame-pointer }
// { dg-options "-fomit-frame-pointer -mno-accumulate-outgoing-args" { target i?86-*-* } }
#include <iostream>

1
gcc/testsuite/g++.dg/eh/omit-frame-pointer2.C
View File

@ -1,6 +1,5 @@
// Reduced from PR c++/5246, PR c++/2447
// { dg-options "-O -fomit-frame-pointer" }
// { dg-options "-O -fomit-frame-pointer -mno-accumulate-outgoing-args" { target i?86-*-* } }
// { dg-do run }
void step (int)

12
gcc/toplev.c
View File

@ -1778,6 +1778,18 @@ process_options (void)
}
if (!flag_stack_protect)
warn_stack_protect = 0;
/* ??? Unwind info is not correct around the CFG unless either a frame
pointer is present or A_O_A is set. Fixing this requires rewriting
unwind info generation to be aware of the CFG and propagating states
around edges. */
if (flag_unwind_tables && !ACCUMULATE_OUTGOING_ARGS
&& flag_omit_frame_pointer)
{
warning (0, "unwind tables currently requires a frame pointer "
"for correctness");
flag_omit_frame_pointer = 0;
}
}
/* Initialize the compiler back end. */

7
libjava/ChangeLog
View File

@ -1,3 +1,10 @@
2005-10-05 Richard Henderson <rth@redhat.com>
PR target/23602
* configure.host (i?86-*): Set -fomit-frame-pointer in libgcj_flags,
but not BACKTRACESPEC.
(x86_64-*): Similarly. Don't set -ffloat-store in 64-bit mode.
2005-10-02 Andrew Pinski <pinskia@physics.uc.edu>
* testsuite/lib/jni.exp (gcj_jni_test_one): For

15
libjava/configure.host
View File

@ -95,18 +95,27 @@ case "${host}" in
;;
i686-*|i586-*|i486-*|i386-*)
sysdeps_dir=i386
libgcj_flags="${libgcj_flags} -ffloat-store"
# With -fomit-frame-pointer -maccumulate-outgoing-args (implied),
# the .text section of libgcj.so is 30k larger, and the .eh_frame
# section is 1.4M smaller.
libgcj_flags="${libgcj_flags} -ffloat-store -fomit-frame-pointer"
libgcj_interpreter=yes
libgcj_cxxflags=
libgcj_cflags=
DIVIDESPEC=-fno-use-divide-subroutine
BACKTRACESPEC=-fomit-frame-pointer
enable_hash_synchronization_default=yes
slow_pthread_self=yes
;;
x86_64-*)
sysdeps_dir=x86-64
libgcj_flags="${libgcj_flags} -ffloat-store"
# For 64-bit we always use SSE registers for arithmetic,
# which doesn't have the extra precision problems of the fpu.
# But be careful about 32-bit multilibs.
case " $CC " in
*" -m32 "*)
libgcj_flags="${libgcj_flags} -ffloat-store" ;;
esac
libgcj_flags="${libgcj_flags} -fomit-frame-pointer"
libgcj_cxxflags=
libgcj_cflags=
DIVIDESPEC=-f%{m32:no-}use-divide-subroutine