[Bug target/50099] ICE: internal compiler error: in extract_insn, at recog.c:2113 while building lttng-ust

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50099

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #5 from Greta Yorsh  2011-08-19 
18:03:48 UTC ---
I can reproduce it with GCC 4.7 r177601 with -marm but not with -mthumb. 
Here a smaller test case that seems related. 

> cat tst2.c

typedef signed char int8_t ;
typedef signed short int16_t;
typedef signed long int32_t;
typedef signed long long int64_t;

// doesn't ice if arg is unsigned
// doesn't ice if return is int32_t instead of int64_t
int64_t foo(int8_t * arg)
{
  //int64_t temp_1;
  int64_t temp_1;

  temp_1 =  arg[256];   // index must be > 255 to ice!
  return temp_1;
}

> arm-none-eabi-gcc -O1 tst2.c

tst2.c: In function 'foo':
tst2.c:15:1: error: unrecognizable insn:
(insn 22 21 23 2 (set (subreg:SI (reg:DI 138 [ MEM[(int8_t *)arg_1(D) + 256B]
]) 0)
(sign_extend:SI (mem:QI (plus:SI (reg:SI 0 r0 [ arg ])
(const_int 256 [0x100])) [0 MEM[(int8_t *)arg_1(D) +
256B]+0 S1 A8]))) tst2.c:13 -1
 (nil))
tst2.c:15:1: internal compiler error: in extract_insn, at recog.c:2115



-- Greta

[Bug middle-end/50607] [4.7 Regression] FAIL: gcc.dg/bconstp-3.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50607

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #5 from Greta Yorsh  2011-10-04 
14:32:55 UTC ---
This test also fails on arm-none-eabi for trunk r :

FAIL: gcc.dg/bconstp-3.c (test for excess errors)

It fails with trunk revision r179462 and passes with r179447.

The new tests added by revision r179462 also fail.

/gcc/testsuite/gcc.c-torture/execute/vect-shuffle-1.c
trunk/gcc/testsuite/gcc.c-torture/execute/vect-shuffle-2.c
trunk/gcc/testsuite/gcc.c-torture/execute/vect-shuffle-3.c
trunk/gcc/testsuite/gcc.c-torture/execute/vect-shuffle-4.c
trunk/gcc/testsuite/gcc.c-torture/execute/vect-shuffle-5.c


http://gcc.gnu.org/ml/gcc-cvs/2011-10/msg00055.html


arm-none-eabi-gcc gcc/testsuite/gcc.dg/bconstp-3.c -ansi -pedantic-errors
-DSTACK_SIZE=16384 -S-o bconstp-3.s 
...
Excess errors:
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:23:3:
error: initializer element is not a constant expression
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:23:3:
error: (near initialization for 'tests[4]')
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:24:3:
error: initializer element is not a constant expression
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:24:3:
error: (near initialization for 'tests[5]')
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:25:3:
error: initializer element is not a constant expression
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:25:3:
error: (near initialization for 'tests[6]')
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:27:1:
error: initializer element is not a constant expression
/arm/scratch/pdtltest/work/sandboxes/fsf/trunk/source/gcc-fsf/gcc/testsuite/gcc.dg/bconstp-3.c:27:1:
error: (near initialization for 'tests[7]')

[Bug middle-end/50607] [4.7 Regression] FAIL: gcc.dg/bconstp-3.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50607

--- Comment #9 from Greta Yorsh  2011-10-06 
10:44:34 UTC ---
There is still a problem on arm: vect-shuffle-2.c and vect-shuffle-5.c fail
with ICE at all optimization levels above -O0 (other vect-shuffle-*.c tests
pass).
ICE is a segfault in expand_vec_shuffle_expr (optabs.c:6660), here are the
details:

$ arm-eabi-gcc vect-shuffle-2.c  -O1 -c 

vect-shuffle-2.c: In function 'main':
vect-shuffle-2.c:30:8: internal compiler error: Segmentation fault

Trunk revision r179558.
Configured with: --with-cpu=cortex-a9 --with-float=softfp --with-fpu=neon 
--target=arm-eabi --enable-checking=release --disable-gdbtk --disable-werror
--disable-tui --disable-rda --disable-sid --disable-utils --disable-lto
--disable-lto --disable-werror --disable-shared --disable-nls --disable-threads
--disable-tls --enable-checking=yes --enable-languages=c,c++,fortran
--with-newlib

[Bug middle-end/50629] New: [4.7 Regression] FAIL: gcc.c-torture/execute/vect-shuffle-2.c with ICE on ARM

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50629

 Bug #: 50629
   Summary: [4.7 Regression] FAIL:
gcc.c-torture/execute/vect-shuffle-2.c with ICE on ARM
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: middle-end
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: artyom.shinkar...@gmail.com, greta.yo...@arm.com,
hjl.to...@gmail.com, h...@gcc.gnu.org,
jos...@codesourcery.com, rgue...@gcc.gnu.org
Target: arm-none-eabi


+++ This bug was initially created as a clone of Bug #50607 +++

Recently added regression tests fail on arm:
FAIL: gcc.c-torture/execute/vect-shuffle-2.c compilation,  -O1  (internal
compiler error)

The tests vect-shuffle-2.c and vect-shuffle-5.c fail with ICE at all
optimization levels above -O0 on arm-eabi. Other vect-shuffle-*.c tests pass.
ICE is a segfault in expand_vec_shuffle_expr (optabs.c:6660), here are the
details:

$ arm-eabi-gcc vect-shuffle-2.c  -O1 -c 

vect-shuffle-2.c: In function 'main':
vect-shuffle-2.c:30:8: internal compiler error: Segmentation fault

Trunk revision r179558.
Configured with: --with-cpu=cortex-a9 --with-float=softfp --with-fpu=neon 
--target=arm-eabi

[Bug middle-end/50607] [4.7 Regression] FAIL: gcc.dg/bconstp-3.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50607

--- Comment #11 from Greta Yorsh  2011-10-06 
13:40:33 UTC ---
> Please open a separate bug report, the problems are gone for i686, at least
> according to .

Cloned: PR50629

[Bug target/50635] New: ICE on invalid code: segfault in vectorize_loops

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50635

 Bug #: 50635
   Summary: ICE on invalid code: segfault in vectorize_loops
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
Target: arm-none-eabi


Created attachment 25431
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25431
small_ice.c

arm-none-eabi/gcc2/gcc/cc1 small_ice.c -O3 -quiet -mcpu=cortex-a9
-mfloat-abi=softfp -mfpu=neon 

small_ice.c: In function 'f0a':
small_ice.c:5:6: internal compiler error: Segmentation fault

Trace:

#0  flow_bb_inside_loop_p (loop=0x77436f68, bb=0x0) at
/work/local-checkouts/main/gcc-fsf/gcc/cfgloop.c:776
#1  0x009ef6ef in vect_is_simple_use (operand=0x774bfaa0,
loop_vinfo=, bb_vinfo=0x0, def_stmt=0x7fffe008,
def=0x7fffe010, dt=0x7fffe01c) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-vect-stmts.c:5595
#2  0x009f93dc in process_use (stmt=0x774b7580, use=0x774bfaa0,
loop_vinfo=0x15077a0, live_p=,
relevant=vect_used_in_scope, worklist=) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-vect-stmts.c:368
#3  0x009f9e69 in vect_mark_stmts_to_be_vectorized
(loop_vinfo=0x15077a0) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-vect-stmts.c:658
#4  0x00a0bae1 in vect_analyze_loop_2 (loop=0x77436f68) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-vect-loop.c:1504
#5  vect_analyze_loop (loop=0x77436f68) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-vect-loop.c:1650
#6  0x00a18e4e in vectorize_loops () at
/work/local-checkouts/main/gcc-fsf/gcc/tree-vectorizer.c:203
#7  0x007d7859 in execute_one_pass (pass=0x129b760) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2064
#8  0x007d7bd5 in execute_pass_list (pass=0x129b760) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2119
#9  0x007d7be7 in execute_pass_list (pass=0x129b580) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2120
#10 0x007d7be7 in execute_pass_list (pass=0x129ac80) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2120
#11 0x008deca8 in tree_rest_of_compilation (fndecl=0x77437400) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-optimize.c:420
#12 0x005c09ce in cgraph_expand_function (node=0x775a4120) at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1805
#13 0x005c3fca in cgraph_expand_all_functions () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1864
#14 cgraph_optimize () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:2141
#15 0x005c44ea in cgraph_finalize_compilation_unit () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1312
#16 0x004a6e08 in c_write_global_declarations () at
/work/local-checkouts/main/gcc-fsf/gcc/c-decl.c:9936
#17 0x0087bcca in compile_file (argc=7, argv=0x7fffe4b8) at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:581
#18 do_compile (argc=7, argv=0x7fffe4b8) at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:1925
#19 toplev_main (argc=7, argv=0x7fffe4b8) at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:2001
#20 0x77874c4d in __libc_start_main () from /lib/libc.so.6
#21 0x00496da9 in _start ()

ICE does not occur without "-mfloat-abi=softfp -mfpu=neon".
This ICE happens with trunk version of end of August 2011 and perhaps earlier.
It does not appear in gcc4.6. 
The code is invalid because uninitialized variable temp_3 is used.

[Bug target/50635] ICE on invalid code: segfault in vectorize_loops

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50635

--- Comment #1 from Greta Yorsh  2011-10-06 
18:08:25 UTC ---
In fact, it's ICE on valid - see the testcase small_ice_init.c. The command
line and backtrace are the same.

[Bug target/50635] ICE on valid: segfault in vectorize_loops

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50635

--- Comment #2 from Greta Yorsh  2011-10-06 
18:09:51 UTC ---
Created attachment 25432
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25432
small testcase for ice on valid code

[Bug rtl-optimization/53176] [4.8 Regression] gcc.target/i386/movbe-2.c and gcc.dg/lower-subreg-1.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53176

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #12 from Greta Yorsh  2012-05-03 
10:29:17 UTC ---
Fails on arm-none-eabi as well:
FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[Bug rtl-optimization/53176] [4.8 Regression] gcc.target/i386/movbe-2.c and gcc.dg/lower-subreg-1.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53176

Greta Yorsh  changed:

   What|Removed |Added

 CC||ramana.radhakrishnan at arm
   ||dot com

--- Comment #14 from Greta Yorsh  2012-05-03 
13:28:42 UTC ---
This test fails because splitting doesn't happen any more after r187015 patch.
Without splitting, better code is generated in this case. 

The test
long long test (long long a, long long b) { return a | b; }
previously compiled into
orr r2, r2, r0
orr r3, r3, r1
mov r0, r2
mov r1, r3
now compiles into a better sequence, avoiding the moves:
orr r0, r0, r2
orr r1, r1, r3

I haven't yet come across an example where the patch makes things worse on arm.
Do you have any such examples?

Thanks,
Greta

[Bug preprocessor/53463] [4.8 Regression]: system header not recognized, yielding warnings about long long preprocessor constant

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53463

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #2 from Greta Yorsh  2012-05-24 
13:08:57 UTC ---
The following tests fail on arm-none-eabi:

FAIL: gcc.dg/cpp/19990407-1.c (test for excess errors)
FAIL: gcc.dg/cpp/paste12.c (test for excess errors)
FAIL: gcc.dg/cpp/strp1.c (test for excess errors)
FAIL: gcc.dg/cpp/strp2.c (test for excess errors)
FAIL: gcc.dg/struct-ret-libc.c (test for excess errors)

For example:

/work/builds/a9-may/install/bin/arm-none-eabi-gcc
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/cpp/19990407-1.c
-mcpu=cortex-a9 -mfloat-abi=softfp -mfpu=neon -fno-diagnostics-show-caret
-pedantic-errors -o 19990407-1.exe
/work/builds/a9-may/install/arm-none-eabi/include/machine/_default_types.h:98:39:
error: use of C99 long long integer constant [-Wlong-long]

[Bug libstdc++/53475] New: Section type conflict errors in libstdc++ testsuite

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53475

 Bug #: 53475
   Summary: Section type conflict errors in libstdc++ testsuite
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: libstdc++
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: paolo.carl...@oracle.com


Failures due to "section type conflict" in libstdc++-v3 testsuite:

FAIL: 20_util/tuple/creation_functions/tie2.cc (test for excess errors)
Excess errors:
/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/include/tuple:1057:25:
error: std::ignore causes a section type conflict with std::ignore

FAIL: 25_algorithms/stable_sort/49559.cc (test for excess errors)
Excess errors:
/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/25_algorithms/stable_sort/49559.cc:37:11:
error: A causes a section type conflict with A

Known to fail: r187724
Known to work: r187546

Executing on host: /work/builds/a9-may/arm-none-eabi/gcc2/./gcc/g++
-shared-libgcc -B/work/builds/a9-may/arm-none-eabi/gcc2/./gcc -nostdinc++
-L/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/src
-L/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/src/.libs
-B/work/builds/a9-may/install/arm-none-eabi/bin/
-B/work/builds/a9-may/install/arm-none-eabi/lib/ -isystem
/work/builds/a9-may/install/arm-none-eabi/include -isystem
/work/builds/a9-may/install/arm-none-eabi/sys-include
-B/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/./libstdc++-v3/src/.libs
-D_GLIBCXX_ASSERT -fmessage-length=0 -ffunction-sections -fdata-sections -g -O2
-g -O2 -DLOCALEDIR="." -nostdinc++
-I/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/include/arm-none-eabi
-I/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/include
-I/work/local-checkouts/gcc-fsf/libstdc++-v3/libsupc++
-I/work/local-checkouts/gcc-fsf/libstdc++-v3/include/backward
-I/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/util
/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/20_util/tuple/creation_functions/tie2.cc
libstdc++_tg.o   -std=gnu++0x ./libtestc++.a-Wl,-wrap,exit
-Wl,-wrap,_exit -Wl,-wrap,main -Wl,-wrap,abort -lm   -o ./tie2.exe(timeout
= 600)
In file included from
/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/20_util/tuple/creation_functions/tie2.cc:22:0:
/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/include/tuple:1057:25:
error: std::ignore causes a section type conflict with std::ignore
   const _Swallow_assign ignore{};
 ^

Executing on host: /work/builds/a9-may/arm-none-eabi/gcc2/./gcc/g++
-shared-libgcc -B/work/builds/a9-may/arm-none-eabi/gcc2/./gcc -nostdinc++
-L/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/src
-L/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/src/.libs
-B/work/builds/a9-may/install/arm-none-eabi/bin/
-B/work/builds/a9-may/install/arm-none-eabi/lib/ -isystem
/work/builds/a9-may/install/arm-none-eabi/include -isystem
/work/builds/a9-may/install/arm-none-eabi/sys-include
-B/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/./libstdc++-v3/src/.libs
-D_GLIBCXX_ASSERT -fmessage-length=0 -ffunction-sections -fdata-sections -g -O2
-g -O2 -DLOCALEDIR="." -nostdinc++
-I/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/include/arm-none-eabi
-I/work/builds/a9-may/arm-none-eabi/gcc2/arm-none-eabi/libstdc++-v3/include
-I/work/local-checkouts/gcc-fsf/libstdc++-v3/libsupc++
-I/work/local-checkouts/gcc-fsf/libstdc++-v3/include/backward
-I/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/util
/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/25_algorithms/stable_sort/49559.cc
libstdc++_tg.o   -std=gnu++0x ./libtestc++.a-Wl,-wrap,exit
-Wl,-wrap,_exit -Wl,-wrap,main -Wl,-wrap,abort -lm   -o ./49559.exe(timeout
= 600)
/work/local-checkouts/gcc-fsf/libstdc++-v3/testsuite/25_algorithms/stable_sort/49559.cc:37:11:
error: A causes a section type conflict with A
 const int A[] = { 10 };
   ^

[Bug c++/53475] [4.8 Regression] Section type conflict errors in libstdc++ testsuite

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53475

--- Comment #3 from Greta Yorsh  2012-05-28 
11:17:00 UTC ---
Created attachment 27512
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27512
preprocessed source

[Bug c++/53475] [4.8 Regression] Section type conflict errors in libstdc++ testsuite

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53475

Greta Yorsh  changed:

   What|Removed |Added

 CC||hubicka at gcc dot gnu.org

--- Comment #4 from Greta Yorsh  2012-05-28 
11:22:46 UTC ---
The problem is caused by r187631: 
http://gcc.gnu.org/ml/gcc-patches/2012-05/msg01222.html

I am adding the author to CC and attaching the preprocessed sources.

[Bug middle-end/53476] [4.8 Regression] FAIL: gcc.dg/attr-weakref-1.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53476

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #3 from Greta Yorsh  2012-05-28 
14:56:47 UTC ---
Fails on arm-none-eabi too.

FAIL: gcc.dg/attr-weakref-1.c (test for excess errors)
Excess errors:
attr-weakref-1.c:(.text.startup+0x1c8): undefined reference to `Wv10a'
attr-weakref-1.c:(.text.startup+0x1cc): undefined reference to `Wv11a'

/work/builds/tst/install/bin/arm-none-eabi-gcc
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/attr-weakref-1.c
-fno-diagnostics-show-caret   -O2 -DSTACK_SIZE=16384 
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/attr-weakref-1a.c -o
/work/tmp/attr-weakref-1.exe
/tmp/cch8OVNm.o: In function `main':
attr-weakref-1.c:(.text.startup+0x1c8): undefined reference to `Wv10a'
attr-weakref-1.c:(.text.startup+0x1cc): undefined reference to `Wv11a'
collect2: error: ld returned 1 exit status

[Bug preprocessor/53463] [4.8 Regression]: system header not recognized, yielding warnings about long long preprocessor constant

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53463

--- Comment #5 from Greta Yorsh  2012-05-28 
15:28:16 UTC ---
(In reply to comment #3)
> It seems to me that this issue is dealt with by the patch I have proposed at
> http://gcc.gnu.org/ml/gcc-patches/2012-05/msg01389.html.  The discussion on
> that patch is still going on.
> 
> Could you please try that patch and reply to the discussion there to say if it
> address the issue on your target?
> 
> Thanks.

This patch fixes the regressions observed for arm-none-eabi.

[Bug tree-optimization/53807] New: FAIL: gcc.dg/vect/bb-slp-11.c scan-tree-dump-times slp "basic block vectorized using SLP" 1

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53807

 Bug #: 53807
   Summary: FAIL: gcc.dg/vect/bb-slp-11.c scan-tree-dump-times slp
"basic block vectorized using SLP" 1
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: tree-optimization
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: rguent...@suse.de
Target: arm-none-eabi


The test gcc.dg/vect/bb-slp-11.c fails to vectorize the basic block on
arm-none-eabi.

Caused by trunk r189012, which was fixing this:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53774

./cc1 -D__USES_INITFINI__ /work/bisect/bb-slp-11.i -quiet -dumpbase bb-slp-11.c
-mcpu=cortex-a9 -mfloat-abi=softfp -mfpu=neon -auxbase-strip ./bb-slp-11.s -O2
-fno-diagnostics-show-caret -ffast-math -ftree-vectorize -fno-vect-cost-model
-fno-common -fdump-tree-slp-details -o ./bb-slp-11.s

[Bug target/53859] ICE when calculate insn latency for armv7e-m arch in O2 level

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53859

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com,
   ||joey.ye at arm dot com

--- Comment #1 from Greta Yorsh  2012-07-06 
16:21:43 UTC ---
Caused by my patch for generating epilogue in RTL (r188742). Looking into it...

Thanks,
Greta

[Bug fortran/52149] New: FAIL: gfortran.dg/vect/vect-8.f90

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52149

 Bug #: 52149
   Summary: FAIL: gfortran.dg/vect/vect-8.f90
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: fortran
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: tkoe...@gcc.gnu.org
Target: arm-none-eabi


This new test fails on arm-none-eabi target. The test was added to trunk in
r183917, as described in PR32373.

[Bug fortran/52149] FAIL: gfortran.dg/vect/vect-8.f90

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52149

--- Comment #2 from Greta Yorsh  2012-02-07 
14:40:17 UTC ---
This patch fixes the problem. Thank you!

-- Greta

[Bug rtl-optimization/52170] New: ICE in trunc_int_for_mode, at explow.c:57 during simplify_plus_minus

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52170

 Bug #: 52170
   Summary: ICE in trunc_int_for_mode, at explow.c:57 during
simplify_plus_minus
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: rtl-optimization
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com


Created attachment 26609
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=26609
testcase that triggers the ICE

cc1 -quiet -mcpu=cortex-a9 -mfloat-abi=softfp -mfpu=neon -O3 smallest.c
smallest.c: In function 'f0a':
smallest.c:24:1: internal compiler error: in trunc_int_for_mode, at explow.c:57

57: gcc_assert (SCALAR_INT_MODE_P (mode));

(gdb) info stack
#0  fancy_abort (file=0x132ccd0 "/work/local-checkouts/gcc-fsf/gcc/explow.c",
line=57, function=0x132d130 "trunc_int_for_mode") at
/work/local-checkouts/gcc-fsf/gcc/diagnostic.c:899
#1  0x00753919 in trunc_int_for_mode (c=1, mode=V16QImode) at
/work/local-checkouts/gcc-fsf/gcc/explow.c:57
#2  0x0073c90f in gen_int_mode (c=1, mode=V16QImode) at
/work/local-checkouts/gcc-fsf/gcc/emit-rtl.c:421
#3  0x00adba40 in neg_const_int (mode=V16QImode, i=0x7765c460) at
/work/local-checkouts/gcc-fsf/gcc/simplify-rtx.c:70
#4  0x00ae96bf in simplify_plus_minus (code=MINUS, mode=V16QImode,
op0=0x775f17e0, op1=0x776010e0) at
/work/local-checkouts/gcc-fsf/gcc/simplify-rtx.c:3963
#5  0x00ae2e20 in simplify_binary_operation_1 (code=MINUS,
mode=V16QImode, op0=0x775f17e0, op1=0x776010e0, trueop0=0x775f17e0,
trueop1=0x776010e0) at
/work/local-checkouts/gcc-fsf/gcc/simplify-rtx.c:2307
#6  0x00ae1049 in simplify_binary_operation (code=MINUS,
mode=V16QImode, op0=0x775f17e0, op1=0x776010e0) at
/work/local-checkouts/gcc-fsf/gcc/simplify-rtx.c:1909
#7  0x01116fe8 in combine_simplify_rtx (x=0x775f1828,
op0_mode=V16QImode, in_dest=0, in_cond=0) at
/work/local-checkouts/gcc-fsf/gcc/combine.c:5459
#8  0x0111688a in subst (x=0x775f1828, from=0x775ff800,
to=0x775f17e0, in_dest=0, in_cond=0, unique_copy=0) at
/work/local-checkouts/gcc-fsf/gcc/combine.c:5292
#9  0x01116674 in subst (x=0x775f1840, from=0x775ff800,
to=0x775f17e0, in_dest=0, in_cond=0, unique_copy=0) at
/work/local-checkouts/gcc-fsf/gcc/combine.c:5228
#10 0x01110569 in try_combine (i3=0x77600678, i2=0x77600630,
i1=0x776005e8, i0=0x0, new_direct_jump_p=0x7fffe05c,
last_combined_insn=0x77600678) at
/work/local-checkouts/gcc-fsf/gcc/combine.c:3253
#11 0x0110b0d8 in combine_instructions (f=0x775c4a00, nregs=582) at
/work/local-checkouts/gcc-fsf/gcc/combine.c:1340
#12 0x0112be4d in rest_of_handle_combine () at
/work/local-checkouts/gcc-fsf/gcc/combine.c:13953
#13 0x00a238cf in execute_one_pass (pass=0x18e6500) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2081
#14 0x00a23abd in execute_pass_list (pass=0x18e6500) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2136
#15 0x00a23ade in execute_pass_list (pass=0x18dcb40) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2137
#16 0x00bc8d2a in tree_rest_of_compilation (fndecl=0x7759c300) at
/work/local-checkouts/gcc-fsf/gcc/tree-optimize.c:422
#17 0x0069e420 in cgraph_expand_function (node=0x775c1000) at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1829
#18 0x0069e5eb in cgraph_expand_all_functions () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1896
#19 0x0069f131 in cgraph_optimize () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:2210
#20 0x0069c16d in cgraph_finalize_compilation_unit () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1338
#21 0x004cd66d in c_write_global_declarations () at
/work/local-checkouts/gcc-fsf/gcc/c-decl.c:10030
#22 0x00b1634a in compile_file () at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:573
#23 0x00b185a8 in do_compile () at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:1938
#24 0x00b1871f in toplev_main (argc=7, argv=0x7fffe448) at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:2014
#25 0x005ce980 in main (argc=7, argv=0x7fffe448) at
/work/local-checkouts/gcc-fsf/gcc/main.c:36

Fails on trunk. 

The problem seems to be that the "mode" argument of simplify_plus_minus, which
satisfies INTEGRAL_MODE_P (mode), is passed on to trunc_int_for_mode, which
requires stronger SCALAR_INT_MODE_P (mode). The assert fails in the attached
test case, because the mode is V16QImode a vector mode not scalar mode.

[Bug fortran/52249] New: FAIL: gfortran.dg/vect/pr32380.f

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52249

 Bug #: 52249
   Summary: FAIL: gfortran.dg/vect/pr32380.f
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: minor
  Priority: P3
 Component: fortran
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com


Recently added testcase gfortran.dg/vect/pr32380.f fails on arm-none-eabi.

FAIL: gfortran.dg/vect/pr32380.f  -O  scan-tree-dump-times vect "vectorized 7
loops" 1

Vectorized 6 loops instead of 7.

[Bug testsuite/52259] New: All tm tests fail on targets without pthread support

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52259

 Bug #: 52259
   Summary: All tm tests fail on targets without pthread support
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: testsuite
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
Target: arm-none-eabi


All tm tests fail on arm-none-eabi, for example:

/work/builds/a9-feb15/arm-none-eabi/gcc2/gcc/xgcc
-B/work/builds/a9-feb15/arm-none-eabi/gcc2/gcc/
/work/local-checkouts/gcc-fsf/gcc/testsuite/c-c++-common/tm/20100127.c   
-fgnu-tm -O -fdump-tree-tmmark -DSTACK_SIZE=16384 -S-o 20100127.s   
(timeout = 60)
xgcc: error: unrecognized command line option '-pthread'
compiler exited with status 1

Caused by trunk r184174, which added -pthread to the command line of all tm
tests without checking that the target supports pthread:
http://gcc.gnu.org/ml/gcc-patches/2012-02/msg00558.html

[Bug testsuite/52259] All tm tests fail on targets without pthread support

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52259

Greta Yorsh  changed:

   What|Removed |Added

 Status|WAITING |RESOLVED
 Resolution||FIXED

--- Comment #2 from Greta Yorsh  2012-02-15 
14:10:40 UTC ---
You are right, the problem has been fixed by r184237.

Thank you,
Greta

[Bug testsuite/52296] New: FAIL: g++.dg/debug/dwarf2/pr52260.C

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52296

 Bug #: 52296
   Summary: FAIL: g++.dg/debug/dwarf2/pr52260.C
Classification: Unclassified
   Product: gcc
   Version: 4.6.2
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: testsuite
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: ja...@gcc.gnu.org
Target: arm-none-eabi


This test passes on trunk, but fails on gcc 4.6 branch (r184304) with the
following message: 

$ arm-eabi-g++
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C
-gdwarf-4 -std=c++0x  -S-o pr52260.s 
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:77:27:
error: expected unqualified-id before 'using'
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:80:15:
error: 'AD' does not name a type
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:82:23:
error: 'AZ' was not declared in this scope
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:82:24:
error: template argument 1 is invalid
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:87:2:
error: 'AZ' does not name a type
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C: In
member function 'void A::Z::foo(B)':
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:102:18:
error: no matching function for call to 'A::AA::AE::o3(A::AA::AE::AY&)'
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:102:18:
note: candidate is:
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:82:12:
note: void A::AA::AE::o3(int, bool)
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/debug/dwarf2/pr52260.C:82:12:
note:   no known conversion for argument 1 from 'A::AA::AE::AY' to 'int'

FAIL: g++.dg/debug/dwarf2/pr52260.C 

The test was introduced by this patch:
http://gcc.gnu.org/ml/gcc-patches/2012-02/msg00860.html
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52260

[Bug middle-end/52603] New: Test failures in gcc.dg/vect: vectorizing unaligned access

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52603

 Bug #: 52603
   Summary: Test failures in gcc.dg/vect: vectorizing unaligned
access
Classification: Unclassified
   Product: gcc
   Version: unknown
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: middle-end
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: rguent...@suse.de
Target: arm-none-eabi


The patch for PR52571 causes the testsuite failures  on arm in gcc.dg/vect
tests related to vectorizing unaligned access.


FAIL: gcc.dg/vect/vect-109.c scan-tree-dump-times vect "Vectorizing an
unaligned access" 3
FAIL: gcc.dg/vect/vect-13.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-17.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-18.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-19.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-2-big-array.c scan-tree-dump-times vect "Vectorizing an
unaligned acce
FAIL: gcc.dg/vect/vect-2.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-20.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-21.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-22.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-27.c scan-tree-dump-times vect "Alignment of access
forced using peeli
FAIL: gcc.dg/vect/vect-29.c scan-tree-dump-times vect "Alignment of access
forced using peeli
FAIL: gcc.dg/vect/vect-3.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-4.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-5.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-56.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 1
FAIL: gcc.dg/vect/vect-58.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-60.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 1
FAIL: gcc.dg/vect/vect-7.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-72.c scan-tree-dump-times vect "Alignment of access
forced using peeli
FAIL: gcc.dg/vect/vect-73-big-array.c scan-tree-dump-times vect "Vectorizing an
unaligned acc
FAIL: gcc.dg/vect/vect-73.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-77-global.c scan-tree-dump-times vect "Vectorizing an
unaligned access
FAIL: gcc.dg/vect/vect-78-global.c scan-tree-dump-times vect "Vectorizing an
unaligned access
FAIL: gcc.dg/vect/vect-86.c scan-tree-dump-times vect "Alignment of access
forced using peeli
FAIL: gcc.dg/vect/vect-92.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-97-big-array.c scan-tree-dump-times vect "Vectorizing an
unaligned acc
FAIL: gcc.dg/vect/vect-97.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/vect-all-big-array.c scan-tree-dump-times vect "Alignment of
access forced
FAIL: gcc.dg/vect/vect-all.c scan-tree-dump-times vect "Alignment of access
forced using peel
FAIL: gcc.dg/vect/vect-multitypes-1.c scan-tree-dump-times vect "Vectorizing an
unaligned acc
FAIL: gcc.dg/vect/vect-multitypes-4.c scan-tree-dump-times vect "Alignment of
access forced u
FAIL: gcc.dg/vect/vect-multitypes-4.c scan-tree-dump-times vect "Vectorizing an
unaligned acc
FAIL: gcc.dg/vect/vect-peel-3.c scan-tree-dump-times vect "Vectorizing an
unaligned access" 1
FAIL: gcc.dg/vect/vect-peel-4.c scan-tree-dump-times vect "Vectorizing an
unaligned access" 1
FAIL: gcc.dg/vect/vect-peel-4.c scan-tree-dump-times vect "Alignment of access
forced using p
FAIL: gcc.dg/vect/slp-25.c scan-tree-dump-times vect "Vectorizing an unaligned
access" 0
FAIL: gcc.dg/vect/slp-25.c scan-tree-dump-times vect "Alignment of access
forced using peelin

XFAIL->XPASS:
XPASS: gcc.dg/vect/vect-60.c scan-tree-dump-times vect "Vectorizing an
unaligned access" 2
XPASS: gcc.dg/vect/vect-56.c scan-tree-dump-times vect "Vectorizing an
unaligned access" 2
XPASS: gcc.dg/vect/vect-multitypes-4.c scan-tree-dump-times vect "Alignment of
access forced

Trunk r185380 
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00972.html

Configured with  --target=arm-eabi --with-cpu=cortex-a9 --with-float=softfp
--with-fpu=neon

[Bug testsuite/52614] [4.8 Regression] Test failures in gcc.dg/vect: vectorizing unaligned access

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52614

--- Comment #2 from Greta Yorsh  2012-03-20 
11:24:41 UTC ---
The tests reported in PR52603 are still failing on arm (last tested on trunk
r185474).

[Bug testsuite/52614] [4.8 Regression] Test failures in gcc.dg/vect: vectorizing unaligned access

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52614

--- Comment #4 from Greta Yorsh  2012-03-20 
16:01:02 UTC ---
These tests pass on arm-none-eabi with the option -fno-common. 

Thank you,
Greta

[Bug target/51122] New: ICE in change_address_1, at emit-rtl.c:2001

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51122

 Bug #: 51122
   Summary: ICE in change_address_1, at emit-rtl.c:2001
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: ramana.radhakrish...@arm.com
Target: arm-none-eabi


Created attachment 25818
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25818
reduced test case

ICE during IRA. 

Possibly related to a latent/unresolved bug reported here:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48256

cc1 -mfloat-abi=softfp -mfpu=neon -marm -quiet  -mcpu=cortex-a9 -O2
smith_crash_246671435.reduced.i  

csmith_crash_246671435.reduced.i: In function 'func_86':
csmith_crash_246671435.reduced.i:39:1: internal compiler error: in
change_address_1, at emit-rtl.c:2001

Backtrace:

#0  fancy_abort (file=0x12e85c8
"/work/local-checkouts/main/gcc-fsf/gcc/emit-rtl.c", line=2001,
function=0x12e9530 "change_address_1") at
/work/local-checkouts/main/gcc-fsf/gcc/diagnostic.c:899
#1  0x00733903 in change_address_1 (memref=0x776320f0, mode=HImode,
addr=0x77632510, validate=1) at
/work/local-checkouts/main/gcc-fsf/gcc/emit-rtl.c:2001
#2  0x007345a5 in adjust_address_1 (memref=0x776320f0, mode=HImode,
offset=0, validate=1, adjust=1) at
/work/local-checkouts/main/gcc-fsf/gcc/emit-rtl.c:2101
#3  0x00796903 in alter_subreg (xp=0x776324c0) at
/work/local-checkouts/main/gcc-fsf/gcc/final.c:2885
#4  0x00796623 in cleanup_subreg_operands (insn=0x7762c630) at
/work/local-checkouts/main/gcc-fsf/gcc/final.c:2832
#5  0x00a72077 in reload (first=0x775b8880, global=1) at
/work/local-checkouts/main/gcc-fsf/gcc/reload1.c:1247
#6  0x0096e4f5 in ira (f=0x0) at
/work/local-checkouts/main/gcc-fsf/gcc/ira.c:3726
#7  0x0096e691 in rest_of_handle_ira () at
/work/local-checkouts/main/gcc-fsf/gcc/ira.c:3799
#8  0x00a1360c in execute_one_pass (pass=0x1887e60) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2064
#9  0x00a137fa in execute_pass_list (pass=0x1887e60) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2119
#10 0x00a1381b in execute_pass_list (pass=0x18884a0) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2120
#11 0x00bae1ef in tree_rest_of_compilation (fndecl=0x77589900) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-optimize.c:420
#12 0x00691a67 in cgraph_expand_function (node=0x775afb40) at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1804
#13 0x00691c32 in cgraph_expand_all_functions () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1871
#14 0x00692547 in cgraph_optimize () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:2168
#15 0x0068f750 in cgraph_finalize_compilation_unit () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1312
#16 0x004cae22 in c_write_global_declarations () at
/work/local-checkouts/main/gcc-fsf/gcc/c-decl.c:9940
#17 0x00b0a47e in compile_file () at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:581
#18 0x00b0c644 in do_compile () at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:1930
#19 0x00b0c7bb in toplev_main (argc=8, argv=0x7fffe308) at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:2006
#20 0x005c48b4 in main (argc=8, argv=0x7fffe308) at
/work/local-checkouts/main/gcc-fsf/gcc/main.c:36

GCC trunk r181174 (Nov 8, 2011)
Reproducible on arm-eabi target with gcc revision from October.

Configured with: configure --target=arm-eabi --enable-checking=release
--disable-gdbtk --disable-werror --disable-tui --disable-rda --disable-sid
--disable-utils --disable-lto --with-cpu=cortex-a9 --with-float=softfp
--with-fpu=neon --disable-lto --disable-werror --disable-shared --disable-nls
--disable-threads --disable-tls --enable-checking=yes --with-newlib

I am attaching a reduce test case. 
By the way, the original test case was generated by csmith compiler-testing
tool: http://embed.cs.utah.edu/csmith/

[Bug target/51162] New: [Regression] ICE: segfault in dump_gimple_call

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51162

 Bug #: 51162
   Summary: [Regression] ICE: segfault in dump_gimple_call
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: ram...@gcc.gnu.org
Target: arm-none-eabi


After the merge with transaction memory branch (trunk r181154), segmentation
fault on arm-none-eabi in tests from gcc/testsuite/gcc.dg/vect/* that run with
-fdump-tree-vect-details, 

cc1 /work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/vect/vect-outer-6.c
-quiet -mcpu=cortex-a9 -mfloat-abi=softfp -mfpu=neon -O1 -ffast-math
-ftree-vectorize -fno-vect-cost-model -fdump-tree-vect-details -o
vect-outer-6.s

Program received signal SIGSEGV, Segmentation fault.
dump_gimple_call (buffer=0x132cfc0, gs=0x77661cc0, spc=0, flags=2) at
/work/local-checkouts/gcc-fsf/gcc/gimple-pretty-print.c:702

701: /* Dump the arguments of _ITM_beginTransaction sanely.  */
702: if (TREE_CODE (fn) == ADDR_EXPR)

(gdb) p fn
$1 = (tree) 0x0
(gdb) info stack
#0  dump_gimple_call (buffer=0x132cfc0, gs=0x77661cc0, spc=0, flags=2) at
/work/local-checkouts/gcc-fsf/gcc/gimple-pretty-print.c:702
#1  0x0071f4c5 in dump_gimple_stmt (buffer=0x132cfc0,
gs=0x77661cc0, spc=0, flags=2) at
/work/local-checkouts/gcc-fsf/gcc/gimple-pretty-print.c:1919
#2  0x00723c32 in print_gimple_stmt (file=,
g=0x77661cc0, spc=0, flags=2) at
/work/local-checkouts/gcc-fsf/gcc/gimple-pretty-print.c:100
#3  0x00a156d5 in vect_finish_stmt_generation (stmt=0x775a99b0,
vec_stmt=0x77661cc0, gsi=) at
/work/local-checkouts/gcc-fsf/gcc/tree-vect-stmts.c:1487
#4  0x00a1e7f5 in vectorizable_load (stmt=0x775a99b0, gsi=, vec_stmt=, slp_node=,
slp_node_instance=) at
/work/local-checkouts/gcc-fsf/gcc/tree-vect-stmts.c:4654
#5  0x00a239b3 in vect_transform_stmt (stmt=0x775a99b0,
gsi=0x7fffe050, strided_store=0x7fffe08f "", slp_node=0x0,
slp_node_instance=0x1) at
/work/local-checkouts/gcc-fsf/gcc/tree-vect-stmts.c:5459
#6  0x00a27bc0 in vect_transform_loop (loop_vinfo=) at /work/local-checkouts/gcc-fsf/gcc/tree-vect-loop.c:5331
#7  0x00a40fbf in vectorize_loops () at
/work/local-checkouts/gcc-fsf/gcc/tree-vectorizer.c:214
#8  0x007ec5d9 in execute_one_pass (pass=0x12e7060) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2074
#9  0x007ec945 in execute_pass_list (pass=0x12e7060) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2129
#10 0x007ec957 in execute_pass_list (pass=0x12e6e80) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2130
#11 0x007ec957 in execute_pass_list (pass=0x12e6580) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2130
#12 0x00900fa8 in tree_rest_of_compilation (fndecl=0x775c) at
/work/local-checkouts/gcc-fsf/gcc/tree-optimize.c:420
#13 0x005cc5ca in cgraph_expand_function (node=0x775c5000) at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1819
#14 0x005cfcca in cgraph_expand_all_functions () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1886
#15 cgraph_optimize () at /work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:2199
#16 0x005d0335 in cgraph_finalize_compilation_unit () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1327
#17 0x004a9980 in c_write_global_declarations () at
/work/local-checkouts/gcc-fsf/gcc/c-decl.c:10023
#18 0x00891c7a in compile_file (argc=13, argv=0x7fffe3d8) at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:581
#19 do_compile (argc=13, argv=0x7fffe3d8) at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:1931
#20 toplev_main (argc=13, argv=0x7fffe3d8) at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:2007
#21 0x77874c4d in __libc_start_main () from /lib/libc.so.6
#22 0x004997e9 in _start ()

[Bug target/51178] New: FAIL: g++.dg/lookup/builtin5.C scan-assembler _ZSt5atanhd

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51178

 Bug #: 51178
   Summary: FAIL: g++.dg/lookup/builtin5.C scan-assembler
_ZSt5atanhd
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: ram...@gcc.gnu.org
Target: arm-none-eabi


FAIL: g++.dg/lookup/builtin5.C scan-assembler _ZSt5atanhd

This test fails on arm-none-eabi for -std=c++11: the generated assembler has
the label atanh instead of _ZSt5atanhd for the function std::atanh.

This tests passes for -std=c++98.

The failure showed up in trunk after check_effective_target_c++11 was added and
-std=c++11 were enabled in the testsuite (trunk r181222).

[Bug c++/51196] New: FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51196

 Bug #: 51196
   Summary: FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: c++
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: ja...@redhat.com, paolo.carl...@oracle.com
Target: arm-none-eabi


The warnings for variable pmf are not produced as expected in this test. The
test was introduced by a patch for PR44277.

FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C  (test for warnings, line
66)
FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C  (test for warnings, line
78)
FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C  (test for warnings, line
90)
FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C  (test for warnings, line
102)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++98  (test for
warnings, line 53)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++98  (test for
warnings, line 65)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++98  (test for
warnings, line 77)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++98  (test for
warnings, line 89)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++11  (test for
warnings, line 53)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++11  (test for
warnings, line 65)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++11  (test for
warnings, line 77)
FAIL: g++.dg/warn/Wzero-as-null-pointer-constant-1.C -std=gnu++11  (test for
warnings, line 89)


Target: arm-none-eabi

Configured with: --with-cpu=cortex-a9 --with-float=softfp --with-fpu=neon
--enable-checking=release --disable-gdbtk --disable-werror --disable-tui
--disable-rda --disable-sid --disable-utils --disable-lto --disable-lto
--disable-werror --disable-shared --disable-nls --disable-threads --disable-tls
--enable-checking=yes --enable-languages=c,c++,fortran --with-newlib
Thread model: single
gcc version 4.7.0 2007 (experimental) (GCC)

[Bug c++/51196] FAIL: g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51196

--- Comment #6 from Greta Yorsh  2011-11-22 
10:56:41 UTC ---
Thanks for fixing it, Paolo. Tested as follows.

Successful cross-build of the patch version for arm-none-eabi target,
configured as before. No regression on qemu for check-g++. The tests for
g++.dg/cpp0x/Wzero-as-null-pointer-constant-1.C all pass. 

Can you see the patch committed to trunk, please? 

Thank you,
Greta

[Bug tree-optimization/51566] New: ICE: verify_gimple in testsuite/g++.dg/ipa/iinline-3.C

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51566

 Bug #: 51566
   Summary: ICE: verify_gimple in testsuite/g++.dg/ipa/iinline-3.C
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: tree-optimization
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
Target: arm-none-eabi


$ cc1plus /work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/ipa/iinline-3.C
-O3 -quiet

/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/ipa/iinline-3.C: In function
'int main(int, char**)':
/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/ipa/iinline-3.C:55:5: error:
statement marked for throw, but doesn't
D.4744_64 = String::funcOne (&S, 4);

/work/local-checkouts/gcc-fsf/gcc/testsuite/g++.dg/ipa/iinline-3.C:55:5:
internal compiler error: verify_gimple failed

Trace:
#0  internal_error (gmsgid=0x1692398 "verify_gimple failed") at
/work/local-checkouts/gcc-fsf/gcc/diagnostic.c:843
#1  0x00e5b5e0 in verify_gimple_in_cfg (fn=0x77666780) at
/work/local-checkouts/gcc-fsf/gcc/tree-cfg.c:4582
#2  0x0106a0f1 in verify_ssa (check_modified_stmt=1 '\001') at
/work/local-checkouts/gcc-fsf/gcc/tree-ssa.c:928
#3  0x00d3b219 in execute_function_todo (data=0x22) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:1725
#4  0x00d3ac3e in do_per_function (callback=0xd3b0ae
, data=0x22) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:1563
#5  0x00d3b311 in execute_todo (flags=34) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:1756
#6  0x00d3bc2c in execute_one_pass (pass=0x1c124c0) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2102
#7  0x00d3bd57 in execute_pass_list (pass=0x1c124c0) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2134
#8  0x00d3bd78 in execute_pass_list (pass=0x1c123a0) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2135
#9  0x00d3bd78 in execute_pass_list (pass=0x1c11aa0) at
/work/local-checkouts/gcc-fsf/gcc/passes.c:2135
#10 0x00edfcc2 in tree_rest_of_compilation (fndecl=0x7753d600) at
/work/local-checkouts/gcc-fsf/gcc/tree-optimize.c:421
#11 0x009b84e5 in cgraph_expand_function (node=0x77542900) at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1818
#12 0x009b86b0 in cgraph_expand_all_functions () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1885
#13 0x009b91f1 in cgraph_optimize () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:2198
#14 0x009b6232 in cgraph_finalize_compilation_unit () at
/work/local-checkouts/gcc-fsf/gcc/cgraphunit.c:1327
#15 0x00685ca8 in cp_write_global_declarations () at
/work/local-checkouts/gcc-fsf/gcc/cp/decl2.c:4050
#16 0x00e2ddee in compile_file () at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:573
#17 0x00e2ffea in do_compile () at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:1928
#18 0x00e30161 in toplev_main (argc=4, argv=0x7fffe458) at
/work/local-checkouts/gcc-fsf/gcc/toplev.c:2004
#19 0x008e8dac in main (argc=4, argv=0x7fffe458) at
/work/local-checkouts/gcc-fsf/gcc/main.c:36

Trunk r182338 configured with --target=arm-none-eabi -cpu=cortex-a9
--with-fpu=neon --with-float=softfp.

[Bug tree-optimization/51566] [4.7 Regression] ICE: verify_gimple in testsuite/g++.dg/ipa/iinline-3.C

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51566

--- Comment #2 from Greta Yorsh  2012-01-19 
10:50:09 UTC ---
I did git bisect and it shows that the ICE disappears from r182403.
Looks like the problem has been fixed by this patch:
http://gcc.gnu.org/ml/gcc-patches/2011-12/msg01247.html
It's a fix for PR51576, where the error message is similar to the error message
we get here, before ICE: "error: statement marked for throw, but doesn’t".

[Bug regression/51963] New: FAIL: g++.dg/torture/pr51344.C

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51963

 Bug #: 51963
   Summary: FAIL: g++.dg/torture/pr51344.C
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: regression
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: cestra...@gmail.com, ja...@gcc.gnu.org,
kti...@gcc.gnu.org, ram...@gcc.gnu.org
Target: arm-none-eabi


This test fails on arm-none-eabi: 

arm-none-eabi-g++ -S source/gcc-fsf/gcc/testsuite/g++.dg/torture/pr51344.C -o
pr51344.s
source/gcc-fsf/gcc/testsuite/g++.dg/torture/pr51344.C:7:58: warning: 'cdecl'
attribute directive ignored [-Wattributes]
source/gcc-fsf/gcc/testsuite/g++.dg/torture/pr51344.C:7:58: warning: 'cdecl'
attribute directive ignored [-Wattributes]


Fails on trunk/4.6/4.5.

[Bug c/51983] New: Wrong line number in an uninitialized variable warning

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51983

 Bug #: 51983
   Summary: Wrong line number in an uninitialized variable warning
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: c
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: ram...@gcc.gnu.org


For uninitialized variables of type _Complex, arm and x86 compilers report
different line numbers in warning messages:
* if only the real part of a complex variable is uninitialized, arm compiler
reports the line number where the variable is declared.
* if only the imaginary part of a complex variable is uninitialized, arm
compiler reports the line number where the variable is used.
* if only one of the parts of a complex variable is uninitialized, x86 compiler
reports the line where the other part is initialized.
* if both parts of a complex variable are uninitialized, both arm and x86
compilers (correctly) report the line where the variable is used. 

This behaviour causes a failure in gcc.dg/uninit-13.c test on arm target. 

Note that in this test, dg-warning directive expects the warning message about
uninitialized variable to report the line in which the imaginary part of the
complex variable is initialized, and not the line where the uninitialized
variable is used. That is, the test passes on x86 target, even thought the line
number in the message is not what one would (perhaps naively) expect, based on
the behaviour of the compiler for other types of uninitialized variables.

I don't know if the problem is in the frontend, middle-end, target, or
testsuite. 



$ cat -n /work/tmp/unin1.c 
 1typedef _Complex float C;
 2C foo(int *p)
 3{
 4  C f;
 5  int a = *p;
 6  *p = 42 * a;
 7  __imag__ f = 0;
 8  return f;
 9}

$ arm-none-eabi-gcc -O -Wuninitialized -S-o tmp.s  /work/tmp/unin1.c
/work/tmp/unin1.c: In function 'foo':
/work/tmp/unin1.c:4:5: warning: '__real__ f' is used uninitialized in this
function [-Wuninitialized]

$ gcc -O -Wuninitialized -S-o tmp.s  /work/tmp/unin1.c
/work/tmp/unin1.c: In function ‘foo’:
/work/tmp/unin1.c:7: warning: ‘__real__ f’ is used uninitialized in this
function


$ cat -n /work/tmp/unin2.c 
 1typedef _Complex float C;
 2C foo(int *p)
 3{
 4  C f;
 5  int a = *p;
 6  *p = 42 * a;
 7  __real__ f = 0;
 8  return f;
 9}

$ arm-none-eabi-gcc -O -Wuninitialized -S-o tmp.s  /work/tmp/unin2.c
/work/tmp/unin2.c: In function 'foo':
/work/tmp/unin2.c:8:3: warning: '__imag__ f' is used uninitialized in this
function [-Wuninitialized]

$ gcc -O -Wuninitialized -S-o tmp.s  /work/tmp/unin2.c
/work/tmp/unin2.c: In function ‘foo’:
/work/tmp/unin2.c:7: warning: ‘__imag__ f’ is used uninitialized in this
function

$ cat -n /work/tmp/unin3.c 
 1typedef _Complex float C;
 2C foo(int *p)
 3{
 4  C f;
 5  int a = *p;
 6  *p = 42 * a;
 7
 8  return f;
 9}
$ arm-none-eabi-gcc -O -Wuninitialized -S-o tmp.s  /work/tmp/unin3.c
/work/tmp/unin3.c: In function 'foo':
/work/tmp/unin3.c:8:3: warning: 'f' is used uninitialized in this function
[-Wuninitialized]

$ gcc -O -Wuninitialized -S-o tmp.s  /work/tmp/unin3.c
/work/tmp/unin3.c: In function ‘foo’:
/work/tmp/unin3.c:8: warning: ‘f’ is used uninitialized in this function

[Bug regression/51996] New: ICE in extract_insn gcc.dg/pr48335-5.c

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51996

 Bug #: 51996
   Summary: ICE in extract_insn gcc.dg/pr48335-5.c
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: regression
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: rguent...@suse.de
Target: arm-none-eabi


./xgcc -mcpu=cortex-a9 -mfpu=neon -mfloat-abi=softfp
-B/work/bisect/build-systest-183470/gcc -O2 -fno-tree-sra -DSTACK_SIZE=16384
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/pr48335-5.c -S -o
/work/tmp/pr48335-5.s
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/pr48335-5.c: In function
'f1':
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/pr48335-5.c:17:1: error:
unrecognizable insn:
(insn 11 10 12 3 (set (reg:DI 141)
(unspec:DI [
(mem/s/c:DI (plus:SI (reg/f:SI 129 virtual-stack-vars)
(const_int -8 [0xfff8])) [2 S8 A32])
] UNSPEC_MISALIGNED_ACCESS))
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/pr48335-5.c:16 -1
 (nil))
/work/local-checkouts/gcc-fsf/gcc/testsuite/gcc.dg/pr48335-5.c:17:1: internal
compiler error: in extract_insn, at recog.c:2123

The failure introduced by r183470, described here:
http://gcc.gnu.org/ml/gcc-patches/2012-01/msg01138.html

[Bug regression/52020] New: ICE in immed_double_const, at emit-rtl.c:544

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52020

 Bug #: 52020
   Summary: ICE in immed_double_const, at emit-rtl.c:544
Classification: Unclassified
   Product: gcc
   Version: 4.7.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: regression
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: greta.yo...@arm.com
CC: m...@suse.de, ram...@gcc.gnu.org
Target: arm-none-eabi


Regression PASS->FAIL: gcc.target/arm/pr48183.c (internal compiler error) on
arm-none-eabi. 

Revision that introduced the problem: r183566, patch described here:
http://gcc.gnu.org/ml/gcc-patches/2012-01/msg01387.html

./cc1 -quiet
/work/local-checkouts/main/gcc-fsf/gcc/testsuite/gcc.target/arm/pr48183.c -O 
-o /work/tmp/pr48183.s -mcpu=cortex-a9 -mfloat-abi=softfp -mfpu=neon -iprefix
/work/builds/a15-jan/install/lib/gcc/arm-none-eabi/4.7.0/ -g

/work/local-checkouts/main/gcc-fsf/gcc/testsuite/gcc.target/arm/pr48183.c: In
function 'move_16bit_to_32bit':
/work/local-checkouts/main/gcc-fsf/gcc/testsuite/gcc.target/arm/pr48183.c:10:6:
internal compiler error: in immed_double_const, at emit-rtl.c:544

(gdb) info stack
#0  fancy_abort (file=0xe15150
"/work/local-checkouts/main/gcc-fsf/gcc/emit-rtl.c", line=544,
function=0xe15ff0 "immed_double_const") at
/work/local-checkouts/main/gcc-fsf/gcc/diagnostic.c:898
#1  0x00648f5c in immed_double_const (i0=0, i1=,
mode=OImode) at /work/local-checkouts/main/gcc-fsf/gcc/emit-rtl.c:544
#2  0x0066d8cb in expand_expr_real_1 (exp=0x770d8fa0, target=0x0,
tmode=, modifier=, alt_rtl=0x0) at
/work/local-checkouts/main/gcc-fsf/gcc/expr.c:9118
#3  0x005aa46d in expand_expr (exp=0x770d8fa0) at
/work/local-checkouts/main/gcc-fsf/gcc/expr.h:442
#4  expand_debug_expr (exp=0x770d8fa0) at
/work/local-checkouts/main/gcc-fsf/gcc/cfgexpand.c:2699
#5  0x005b13ca in expand_debug_locations () at
/work/local-checkouts/main/gcc-fsf/gcc/cfgexpand.c:3668
#6  gimple_expand_cfg () at
/work/local-checkouts/main/gcc-fsf/gcc/cfgexpand.c:4528
#7  0x007f0d29 in execute_one_pass (pass=0x130aa80) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2081
#8  0x007f1095 in execute_pass_list (pass=0x130aa80) at
/work/local-checkouts/main/gcc-fsf/gcc/passes.c:2136
#9  0x00907238 in tree_rest_of_compilation (fndecl=0x76c59100) at
/work/local-checkouts/main/gcc-fsf/gcc/tree-optimize.c:420
#10 0x005cec5a in cgraph_expand_function (node=0x76c565a0) at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1819
#11 0x005d249a in cgraph_expand_all_functions () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1886
#12 cgraph_optimize () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:2200
#13 0x005d29f5 in cgraph_finalize_compilation_unit () at
/work/local-checkouts/main/gcc-fsf/gcc/cgraphunit.c:1328
#14 0x004aca88 in c_write_global_declarations () at
/work/local-checkouts/main/gcc-fsf/gcc/c-decl.c:10030
#15 0x00896f87 in compile_file () at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:573
#16 do_compile () at /work/local-checkouts/main/gcc-fsf/gcc/toplev.c:1938
#17 0x00897664 in toplev_main (argc=16, argv=0x7fffe348) at
/work/local-checkouts/main/gcc-fsf/gcc/toplev.c:2014
#18 0x77874c4d in __libc_start_main () from /lib/libc.so.6
#19 0x0049a7a9 in _start ()

[Bug c/49362] Arm Neon intrinsic types not correctly interpreted by compiler.

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49362

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #3 from Greta Yorsh  2011-06-14 
12:59:11 UTC ---
It looks like the problem you described has already been fixed.

When the example is compiled with gcc from trunk (gcc version 4.7.0 with -O2),
vld1q variant has 15 instructions and vld2q variant has 13 instructions (see
below).
The version of gcc you use is 4.4.1. The issue hasn't been fixed in the latest
gcc release 4.6, but the fix should be included in the next release and
probably won't be backported to 4.5 and 4.6 releases.


Disassembly of section .text:

 :
   0:f460438f vld2.32{d20-d23}, [r0]
   4:f461038f vld2.32{d16-d19}, [r1]
   8:f34481f0 veorq12, q10, q8
   c:f34601f2 veorq8, q11, q9
  10:f3f02568 vcnt.8q9, q12
  14:f3f00560 vcnt.8q8, q8
  18:f24208e0 vadd.i8q8, q9, q8
  1c:f3f002e0 vpaddl.u8q8, q8
  20:f3f402e0 vpaddl.u16q8, q8
  24:f26121b1 vorrd18, d17, d17
  28:f2620bb0 vpadd.i32d16, d18, d16
  2c:f2600bb0 vpadd.i32d16, d16, d16
  30:ee100b90 vmov.32r0, d16[0]
  34:e12fff1e bxlr

0038 :
  38:f4602a8d vld1.32{d18-d19}, [r0]!
  3c:f4610a8d vld1.32{d16-d17}, [r1]!
  40:f34221f0 veorq9, q9, q8
  44:f4604a8f vld1.32{d20-d21}, [r0]
  48:f3f02562 vcnt.8q9, q9
  4c:f4610a8f vld1.32{d16-d17}, [r1]
  50:f34401f0 veorq8, q10, q8
  54:f3f00560 vcnt.8q8, q8
  58:f24208e0 vadd.i8q8, q9, q8
  5c:f3f002e0 vpaddl.u8q8, q8
  60:f3f402e0 vpaddl.u16q8, q8
  64:f26121b1 vorrd18, d17, d17
  68:f2620bb0 vpadd.i32d16, d18, d16
  6c:f2600bb0 vpadd.i32d16, d16, d16
  70:ee100b90 vmov.32r0, d16[0]
  74:e12fff1e bxlr

[Bug target/50022] [4.7 regression] "incorrect condition in IT block" when building mozilla code base for ARM

[Greta.Yorsh at arm dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50022

Greta Yorsh  changed:

   What|Removed |Added

 CC||Greta.Yorsh at arm dot com

--- Comment #2 from Greta Yorsh  2011-08-09 
15:55:46 UTC ---
Here is a small test case:

long long glob;
void bar(int val)
{
  glob = (val >= 0 && val <= 2147483647 ? val : 2147483647) * 100L; 
}

I reproduced the bug with the flags
arm-none-eabi-gcc -c  -O1 -mcpu=cortex-a8  -mthumb t1.c
but not with -O0  and not with -mcpu=cortex-a9.
By the way, changing the large numbers to smaller or removing conditions makes
the error message go away. 

The error seems to be generating K+1 instructions instead of K in the IT block.


bar:
cmpr0, #0
ige
movwger3, #16960
movtger3, 15
mulger0, r3, r0
movger2, r0

itege
asrger3, r2, #31
adrltr3, .L4
ldrdltr2, [r3]

movwr1, #:lower16:glob
movtr1, #:upper16:glob
strdr2, [r1]
bxlr


-- Greta