[Patch, Fortran, committed] PR51972 - Extended checking of gfortran.dg/class_48.f90
The patch for PR 51972 fixed some issues with intrinsic assignment, which involved polymorphic components. However, I couldn't check the correctness in some cases as I run ICEs. Those were fixed by Mikael's recent patch (cf. PR 51754 et alia). The committed (Rev. ) and attached patch adds the missing checks. On the way, I run into another issue (now: PR 52102), which I worked around. However, as MALLOC_PERTURB_ or valgrind shows, the issue also affects the current version. Thus, in the committed patch, only test3() has proper checking, for test4() I have added a commented out test case, which without MALLOC_PERTURB_ should work. Committed as Rev. 183904. Tobias Index: gcc/testsuite/gfortran.dg/class_48.f90 === --- gcc/testsuite/gfortran.dg/class_48.f90 (Revision 183903) +++ gcc/testsuite/gfortran.dg/class_48.f90 (Arbeitskopie) @@ -70,14 +70,16 @@ type(t2) :: one, two - one = two + allocate (two%a(2)) + two%a(1)%x = 4 + two%a(2)%x = 6 if (allocated (one%a)) call abort () - - allocate (two%a(2), source=[t(4), t(6)]) one = two if (.not.allocated (one%a)) call abort () -! FIXME: Check value + if ((one%a(1)%x /= 4)) call abort () + if ((one%a(2)%x /= 6)) call abort () + deallocate (two%a) one = two if (allocated (one%a)) call abort () @@ -94,12 +96,43 @@ type(t2) :: one, two - one = two if (allocated (one%a)) call abort () - -! allocate (two%a(2)) ! ICE: SEGFAULT + if (allocated (two%a)) call abort () +! +! FIXME: Fails due to PR 51754 +! +! NOTE: Might be only visible with MALLOC_PERTURB_ or with valgrind +! +! allocate (two%a(2)) +! if (allocated (two%a(1)%x)) call abort () +! if (allocated (two%a(2)%x)) call abort () +! allocate (two%a(1)%x(3), source=[1,2,3]) +! allocate (two%a(2)%x(5), source=[5,6,7,8,9]) ! one = two ! if (.not. allocated (one%a)) call abort () +! if (.not. allocated (one%a(1)%x)) call abort () +! if (.not. allocated (one%a(2)%x)) call abort () +! +! if (size(one%a) /= 2) call abort() +! if (size(one%a(1)%x) /= 3) call abort() +! if (size(one%a(2)%x) /= 5) call abort() +! if (any (one%a(1)%x /= [1,2,3])) call abort () +! if (any (one%a(2)%x /= [5,6,7,8,9])) call abort () +! +! deallocate (two%a(1)%x) +! one = two +! if (.not. allocated (one%a)) call abort () +! if (allocated (one%a(1)%x)) call abort () +! if (.not. allocated (one%a(2)%x)) call abort () +! +! if (size(one%a) /= 2) call abort() +! if (size(one%a(2)%x) /= 5) call abort() +! if (any (one%a(2)%x /= [5,6,7,8,9])) call abort () +! +! deallocate (two%a) + one = two + if (allocated (one%a)) call abort () + if (allocated (two%a)) call abort () end subroutine test4 Index: gcc/testsuite/ChangeLog === --- gcc/testsuite/ChangeLog (Revision 183903) +++ gcc/testsuite/ChangeLog (Arbeitskopie) @@ -1,3 +1,8 @@ +2012-02-05 Tobias Burnus + + PR fortran/51972 + * gfortran.dg/class_48.f90: Add some further checks. + 2012-02-05 Ira Rosen PR tree-optimization/52091
[patch, fortran] Fix PR 48847, bogus "unused parameter" warning
Hello world,
the attached, rather obvious patch fixed a bogus "unused parameter"
warning with procedure dummy arguments and warns about these if
they occur.
Regression-tested. OK for trunk? Or rather wait until 4.8?
Thomas
2012-02-05 Thomas König
PR fortran/48847
* trans-decl.c: Warn about unused dummy procedure arguments
if -Wunused-dummy-argument is specified. Suppress middle-end
warnings about procedure arguments.
2012-02-05 Thomas König
PR fortran/48847
* gfortran.dg/warn_unused_dummy_argument_3.f90: New test.
Index: trans-decl.c
===
--- trans-decl.c (Revision 183873)
+++ trans-decl.c (Arbeitskopie)
@@ -4683,6 +4683,22 @@ generate_local_decl (gfc_symbol * sym)
&& sym->ts.type == BT_CHARACTER && sym->ts.is_c_interop
&& sym->ns->proc_name != NULL && sym->ns->proc_name->attr.is_bind_c)
gfc_conv_scalar_char_value (sym, NULL, NULL);
+
+ /* Unused procedure passed as dummy argument. */
+ if (sym->attr.flavor == FL_PROCEDURE)
+ {
+ if (!sym->attr.referenced)
+ {
+ if (gfc_option.warn_unused_dummy_argument)
+ gfc_warning ("Unused dummy argument '%s' at %L", sym->name,
+ &sym->declared_at);
+ }
+
+ /* Silence bogus "unused parameter" warnings from the
+ middle end. */
+ if (sym->backend_decl != NULL_TREE)
+ TREE_NO_WARNING(sym->backend_decl) = 1;
+ }
}
/* Make sure we convert the types of the derived types from iso_c_binding
! { dg-do compile }
! { dg-options "-Wunused-dummy-argument -Wunused-parameter" }
! PR 48847 - we used to generate a warning for g(), and none for h()
program main
contains
function f(g,h)
interface
real function g()
end function g
end interface
interface
real function h() ! { dg-warning "Unused dummy argument" }
end function h
end interface
real :: f
f = g()
end function f
end program main
New Swedish PO file for 'gcc' (version 4.7-b20120128)
Hello, gentle maintainer. This is a message from the Translation Project robot. A revised PO file for textual domain 'gcc' has been submitted by the Swedish team of translators. The file is available at: http://translationproject.org/latest/gcc/sv.po (This file, 'gcc-4.7-b20120128.sv.po', has just now been sent to you in a separate email.) All other PO files for your package are available in: http://translationproject.org/latest/gcc/ Please consider including all of these in your next release, whether official or a pretest. Whenever you have a new distribution with a new version number ready, containing a newer POT file, please send the URL of that distribution tarball to the address below. The tarball may be just a pretest or a snapshot, it does not even have to compile. It is just used by the translators when they need some extra translation context. The following HTML page has been updated: http://translationproject.org/domain/gcc.html If any question arises, please contact the translation coordinator. Thank you for all your work, The Translation Project robot, in the name of your translation coordinator.
[v3] libstdc++/52119
Hi, the library bits seem straightforward enough to go in now. Of course the substantive issue seems C++ front end diagnostics, post 4.7.0 material. Sanity checked x86_64-linux. Thanks, Paolo. /// 2012-02-05 Jeffrey Yasskin Paolo Carlini PR libstdc++/52119 * include/std/limits (__glibcxx_min): Fix to aboid undefined behavior. Index: include/std/limits === --- include/std/limits (revision 183904) +++ include/std/limits (working copy) @@ -1,7 +1,7 @@ // The template and inlines for the numeric_limits classes. -*- C++ -*- // Copyright (C) 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, -// 2008, 2009, 2010, 2011 Free Software Foundation, Inc. +// 2008, 2009, 2010, 2011, 2012 Free Software Foundation, Inc. // // This file is part of the GNU ISO C++ Library. This library is free // software; you can redistribute it and/or modify it under the @@ -129,7 +129,7 @@ #define __glibcxx_signed(T)((T)(-1) < 0) #define __glibcxx_min(T) \ - (__glibcxx_signed (T) ? (T)1 << __glibcxx_digits (T) : (T)0) + (__glibcxx_signed (T) ? -__glibcxx_max (T) - 1 : (T)0) #define __glibcxx_max(T) \ (__glibcxx_signed (T) ? \
[patch] [4.8] Support pattern recognition in SLP
Hi,
This patch adds a support of pattern recognition in basic block SLP.
Bootstrapped and tested on powerpc64-suse-linux.
Ira
ChangeLog:
* tree-vectorizer.h (vect_pattern_recog): Add new argument.
* tree-vect-loop.c (vect_analyze_loop_2): Update call to
vect_pattern_recog.
* tree-vect-patterns.c (widened_name_p): Pass basic block
info to vect_is_simple_use.
(vect_recog_dot_prod_pattern): Fail for basic blocks.
(vect_recog_widen_sum_pattern): Likewise.
(vect_handle_widen_op_by_const): Support basic blocks.
(vect_operation_fits_smaller_type,
vect_recog_over_widening_pattern): Likewise.
(vect_recog_vector_vector_shift_pattern): Support basic blocks.
Update call to vect_is_simple_use.
(vect_recog_mixed_size_cond_pattern): Support basic blocks.
Add printing.
(check_bool_pattern): Add an argument, update call to
vect_is_simple_use and the recursive calls.
(vect_recog_bool_pattern): Update relevant function calls.
Add printing.
(vect_mark_pattern_stmts): Update calls to new_stmt_vec_info.
(vect_pattern_recog_1): Check for reduction only in loops.
(vect_pattern_recog): Add new argument. Support basic blocks.
* tree-vect-stmts.c (vectorizable_conversion): Pass basic block
info to vect_is_simple_use_1.
* tree-vect-slp.c (vect_get_and_check_slp_defs): Support basic
blocks.
(vect_slp_analyze_bb_1): Call vect_pattern_recog.
testsuite/ChangeLog:
* gcc.dg/vect/bb-slp-pattern-1.c: New test.
* gcc.dg/vect/bb-slp-pattern-2.c: New test.
(See attached file: bb-patterns.txt)Index: testsuite/gcc.dg/vect/bb-slp-pattern-1.c
===
--- testsuite/gcc.dg/vect/bb-slp-pattern-1.c(revision 0)
+++ testsuite/gcc.dg/vect/bb-slp-pattern-1.c(revision 0)
@@ -0,0 +1,54 @@
+/* { dg-require-effective-target vect_int } */
+
+#include
+#include "tree-vect.h"
+
+#define N 8
+
+unsigned short X[N];
+unsigned short Y[N];
+unsigned int result[N];
+
+/* unsigned short->unsigned int widening-mult. */
+__attribute__ ((noinline, noclone)) void
+foo (void)
+{
+ result[0] = (unsigned int) (X[0] * Y[0]);
+ result[1] = (unsigned int) (X[1] * Y[1]);
+ result[2] = (unsigned int) (X[2] * Y[2]);
+ result[3] = (unsigned int) (X[3] * Y[3]);
+ result[4] = (unsigned int) (X[4] * Y[4]);
+ result[5] = (unsigned int) (X[5] * Y[5]);
+ result[6] = (unsigned int) (X[6] * Y[6]);
+ result[7] = (unsigned int) (X[7] * Y[7]);
+}
+
+int main (void)
+{
+ int i, tmp;
+
+ check_vect ();
+
+ for (i = 0; i < N; i++)
+{
+ X[i] = i;
+ Y[i] = 64-i;
+}
+
+ foo ();
+
+ for (i = 0; i < N; i++)
+{
+ __asm__ volatile ("");
+ tmp = X[i] * Y[i];
+ if (result[i] != tmp)
+abort ();
+}
+
+ return 0;
+}
+
+/* { dg-final { scan-tree-dump-times "vectorizing stmts using SLP" 1 "slp" {
target { vect_widen_mult_hi_to_si || vect_unpack } } } } */
+/* { dg-final { scan-tree-dump-times "vect_recog_widen_mult_pattern: detected"
8 "slp" { target vect_widen_mult_hi_to_si_pattern } } } */
+/* { dg-final { scan-tree-dump-times "pattern recognized" 8 "slp" { target
vect_widen_mult_hi_to_si_pattern } } } */
+/* { dg-final { cleanup-tree-dump "slp" } } */
Index: testsuite/gcc.dg/vect/bb-slp-pattern-2.c
===
--- testsuite/gcc.dg/vect/bb-slp-pattern-2.c(revision 0)
+++ testsuite/gcc.dg/vect/bb-slp-pattern-2.c(revision 0)
@@ -0,0 +1,52 @@
+/* { dg-require-effective-target vect_condition } */
+
+#include "tree-vect.h"
+
+#define N 128
+
+__attribute__((noinline, noclone)) void
+foo (short * __restrict__ a, int * __restrict__ b, int stride)
+{
+ int i;
+
+ for (i = 0; i < N/stride; i++, a += stride, b += stride)
+ {
+ a[0] = b[0] ? 1 : 7;
+ a[1] = b[1] ? 2 : 0;
+ a[2] = b[2] ? 3 : 0;
+ a[3] = b[3] ? 4 : 0;
+ a[4] = b[4] ? 5 : 0;
+ a[5] = b[5] ? 6 : 0;
+ a[6] = b[6] ? 7 : 0;
+ a[7] = b[7] ? 8 : 0;
+ }
+}
+
+short a[N];
+int b[N];
+int main ()
+{
+ int i;
+
+ check_vect ();
+
+ for (i = 0; i < N; i++)
+{
+ a[i] = i;
+ b[i] = -i;
+}
+
+ foo (a, b, 8);
+
+ for (i = 1; i < N; i++)
+if (a[i] != i%8 + 1)
+ abort ();
+
+ if (a[0] != 7)
+abort ();
+
+ return 0;
+}
+
+/* { dg-final { scan-tree-dump-times "basic block vectorized using SLP" 1
"slp" { target { vect_element_align && vect_pack_trunc } } } } */
+/* { dg-final { cleanup-tree-dump "slp" } } */
Index: tree-vectorizer.h
===
--- tree-vectorizer.h (revision 183902)
+++ tree-vectorizer.h (working copy)
@@ -933,7 +933,7 @@ extern void vect_slp_transform_bb (basic_block);
in the future. */
typedef gimple (* vect_recog_func_ptr) (VEC (gimple, heap) **, tree *, tree *);
#define NUM_PATTERNS 10
Re: [patch] should not define bool, true or false as macros for C++
On 4 February 2012 23:35, Gerald Pfeifer wrote:
> For what it's worth, I strongly suggest that you only define those when
> __cpluplus is pre-C++11.
>
> There is simply too much software out there which will run into this
Really? Why would any C++ code assume "bool" is defined as a macro?
It's been a keyword in C++ for longer than C99 has defined it as a macro.
> and being aggressive in breaking (admittedly non-standard confirming
> programs) gives GCC a bad reputation and is not nice to our users to
> begin with.
Fair enough, this revised patch still defines bool, true and false for
C++98 mode, but not for C++11 mode.
gcc/
* ginclude/stdbool.h (true, false, bool): Do not define for C++11.
libstdc++/
* testsuite/18_support/headers/cstdbool/macros.cc: New.
Tested x86_64-linux, OK for trunk?
diff --git a/gcc/ginclude/stdbool.h b/gcc/ginclude/stdbool.h
index 4ed911f..4645654 100644
--- a/gcc/ginclude/stdbool.h
+++ b/gcc/ginclude/stdbool.h
@@ -1,4 +1,4 @@
-/* Copyright (C) 1998, 1999, 2000, 2009 Free Software Foundation, Inc.
+/* Copyright (C) 1998, 1999, 2000, 2009, 2012 Free Software Foundation, Inc.
This file is part of GCC.
@@ -36,11 +36,15 @@ see the files COPYING3 and COPYING.RUNTIME respectively.
If not, see
#else /* __cplusplus */
-/* Supporting in C++ is a GCC extension. */
+/* Supporting _Bool in C++ is a GCC extension. */
#define _Bool bool
+
+#if __cplusplus < 201103L
+/* Defining these macros in C++98 is a GCC extension. */
#define bool bool
#define false false
#define true true
+#endif
#endif /* __cplusplus */
diff --git a/libstdc++-v3/testsuite/18_support/headers/cstdbool/macros.cc
b/libstdc++-v3/testsuite/18_support/headers/cstdbool/macros.cc
new file mode 100644
index 000..58631d8
--- /dev/null
+++ b/libstdc++-v3/testsuite/18_support/headers/cstdbool/macros.cc
@@ -0,0 +1,38 @@
+// { dg-do compile }
+// { dg-options "-std=gnu++11" }
+
+// Copyright (C) 2012 Free Software Foundation, Inc.
+//
+// This file is part of the GNU ISO C++ Library. This library is free
+// software; you can redistribute it and/or modify it under the
+// terms of the GNU General Public License as published by the
+// Free Software Foundation; either version 3, or (at your option)
+// any later version.
+
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU General Public License for more details.
+
+// You should have received a copy of the GNU General Public License along
+// with this library; see the file COPYING3. If not see
+//
Re: [v3] adjust weak_ptr testcase
On 23 December 2011 17:33, Jonathan Wakely wrote:
> This modifies the test to PASS when the expected type of exception is
> caught, instead of being XFAIL due to uncaught exception.
>
> Tested x86_64-linux, committed to trunk.
>
> * testsuite/tr1/2_general_utilities/shared_ptr/cons/
> weak_ptr_expired.cc: Modify to PASS instead of XFAIL.
And the same for the 20_util equivalent of the same test, this avoids
aborting which should make it run slightly faster (and means Fedora's
ABRT doesn't flag up a crash when running the testsuite)
2012-02-05 Jonathan Wakely
* testsuite/20_util/shared_ptr/cons/weak_ptr_expired.cc: Modify to
PASS instead of XFAIL.
Tested x86_64-linux and committed to trunk.
diff --git a/libstdc++-v3/testsuite/20_util/shared_ptr/cons/weak_ptr_expired.cc
b/libstdc++-v3/testsuite/20_util/shared_ptr/cons/weak_ptr_expired.cc
index cff9b3b..d2bf508 100644
--- a/libstdc++-v3/testsuite/20_util/shared_ptr/cons/weak_ptr_expired.cc
+++ b/libstdc++-v3/testsuite/20_util/shared_ptr/cons/weak_ptr_expired.cc
@@ -1,7 +1,6 @@
// { dg-options "-std=gnu++0x" }
-// { dg-do run { xfail *-*-* } }
-// Copyright (C) 2005, 2006, 2007, 2009 Free Software Foundation
+// Copyright (C) 2005, 2006, 2007, 2009, 2012 Free Software Foundation
//
// This file is part of the GNU ISO C++ Library. This library is free
// software; you can redistribute it and/or modify it under the
@@ -31,7 +30,7 @@ struct A { };
int
test01()
{
- bool test __attribute__((unused)) = true;
+ bool test = false;
std::shared_ptr a1(new A);
std::weak_ptr wa(a1);
@@ -44,12 +43,9 @@ test01()
catch (const std::bad_weak_ptr&)
{
// Expected.
- __throw_exception_again;
- }
- catch (...)
- {
-// Failed.
+test = true;
}
+ VERIFY( test );
return 0;
}
[MIPS, committed] Tidy atomic-related failures for mips-elf
Plain MIPS I testing on mips-elf has quite a few atomic-related failures,
due to the lack of LL and SC. Fixed as below.
Tested on mips-elf, mips64-linux-gnu, and a host of others. Applied.
Richard
gcc/testsuite/
* lib/target-supports.exp (check_effective_target_mips_llsc): New.
(check_effective_target_sync_int_long): Use it.
(check_effective_target_sync_char_short): Likewise.
* gcc.target/mips/atomic-memory-1.c: Restrict error check to mips_llsc.
Index: gcc/testsuite/lib/target-supports.exp
===
--- gcc/testsuite/lib/target-supports.exp 2012-02-05 14:34:48.0
+
+++ gcc/testsuite/lib/target-supports.exp 2012-02-05 14:51:03.0
+
@@ -890,6 +890,26 @@ proc check_effective_target_mips_newabi_
} "-mabi=64"]
}
+# Return true if the target is a MIPS target that has access
+# to the LL and SC instructions.
+
+proc check_effective_target_mips_llsc { } {
+if { ![istarget mips*-*-*] } {
+ return 0
+}
+# Assume that these instructions are always implemented for
+# non-elf* targets, via emulation if necessary.
+if { ![istarget *-*-elf*] } {
+ return 1
+}
+# Otherwise assume LL/SC support for everything but MIPS I.
+return [check_no_compiler_messages mips_llsc assembly {
+ #if __mips == 1
+ #error FOO
+ #endif
+}]
+}
+
# Return 1 if the current multilib does not generate PIC by default.
proc check_effective_target_nonpic { } {
@@ -3770,7 +3790,7 @@ proc check_effective_target_sync_int_lon
|| [istarget powerpc*-*-*]
|| [istarget sparc64-*-*]
|| [istarget sparcv9-*-*]
-|| [istarget mips*-*-*] } {
+|| [check_effective_target_mips_llsc] } {
set et_sync_int_long_saved 1
}
}
@@ -3800,7 +3820,7 @@ proc check_effective_target_sync_char_sh
|| [istarget powerpc*-*-*]
|| [istarget sparc64-*-*]
|| [istarget sparcv9-*-*]
-|| [istarget mips*-*-*] } {
+|| [check_effective_target_mips_llsc] } {
set et_sync_char_short_saved 1
}
}
Index: gcc/testsuite/gcc.target/mips/atomic-memory-1.c
===
--- gcc/testsuite/gcc.target/mips/atomic-memory-1.c 2012-02-05
14:34:48.0 +
+++ gcc/testsuite/gcc.target/mips/atomic-memory-1.c 2012-02-05
14:50:33.0 +
@@ -1,6 +1,6 @@
/* { dg-do run } */
-/* { dg-message "note: '__sync_nand_and_fetch' changed semantics in GCC 4.4"
"" { target *-*-* } 0 } */
+/* { dg-message "note: '__sync_nand_and_fetch' changed semantics in GCC 4.4"
"" { target mips_llsc } 0 } */
extern void abort (void);
extern void exit (int);
[MIPS, committed] XFAIL gcc.dg/pr48774.c for MIPS REL targets
gcc.dg/pr48774.c fails on MIPS REL targets because of an orphaned
HI16 relocation; see PR 52125 for details. This isn't a regression,
so I'm just going to XFAIL it for now.
Tested on mips64-linux-gnu and various *-elf targets. Applied.
Richard
gcc/testsuite/
PR target/52125
* lib/target-supports.exp (check_effective_target_mips_rel): New.
* gcc.dg/pr48774.c: Skip on MIPS REL targets.
Index: gcc/testsuite/lib/target-supports.exp
===
--- gcc/testsuite/lib/target-supports.exp 2012-02-05 14:53:05.0
+
+++ gcc/testsuite/lib/target-supports.exp 2012-02-05 14:53:19.0
+
@@ -910,6 +910,20 @@ proc check_effective_target_mips_llsc {
}]
}
+# Return true if the target is a MIPS target that uses in-place relocations.
+
+proc check_effective_target_mips_rel { } {
+if { ![istarget mips*-*-*] } {
+ return 0
+}
+return [check_no_compiler_messages mips_rel object {
+ #if (defined _ABIN32 && _MIPS_SIM == _ABIN32) \
+ || (defined _ABI64 && _MIPS_SIM == _ABI64)
+ #error FOO
+ #endif
+}]
+}
+
# Return 1 if the current multilib does not generate PIC by default.
proc check_effective_target_nonpic { } {
Index: gcc/testsuite/gcc.dg/pr48774.c
===
--- gcc/testsuite/gcc.dg/pr48774.c 2012-02-05 14:30:22.0 +
+++ gcc/testsuite/gcc.dg/pr48774.c 2012-02-05 14:53:46.0 +
@@ -1,5 +1,6 @@
/* PR target/48774 */
/* { dg-do run } */
+/* { dg-skip-if "PR 52125" { mips_rel } { "*" } { "" } } */
/* { dg-options "-O2 -funroll-loops" } */
extern void abort (void);
[MIPS, committed] Expect MIPS16 to use memcpy for short strings
Although I changed non-MIPS16 code to allow larger inline copies:
http://gcc.gnu.org/ml/gcc-patches/2012-01/msg00053.html
MIPS16 code is still for now expected to use memcpy. This patch
adjusts the testsuite accordingly.
Tested on mips64-linux-gnu and various ELF targets. Applied.
Richard
gcc/testsuite/
* gcc.dg/tree-prof/stringop-2.c (main): Add a nomips16 attribute
on MIPS targets.
* gfortran.dg/pr45636.f90: XFAIL for MIPS16 targets.
Index: gcc/testsuite/gcc.dg/tree-prof/stringop-2.c
===
--- gcc/testsuite/gcc.dg/tree-prof/stringop-2.c 2012-02-04 10:57:10.0
+
+++ gcc/testsuite/gcc.dg/tree-prof/stringop-2.c 2012-02-05 14:23:07.0
+
@@ -3,6 +3,10 @@
int b[1000];
int size=1;
int max=1;
+#ifdef __mips
+/* We allow short memcpy()s for MIPS16. */
+int __attribute__((nomips16))
+#endif
main()
{
int i;
Index: gcc/testsuite/gfortran.dg/pr45636.f90
===
--- gcc/testsuite/gfortran.dg/pr45636.f90 2012-02-05 14:23:36.0
+
+++ gcc/testsuite/gfortran.dg/pr45636.f90 2012-02-05 14:26:41.0
+
@@ -10,5 +10,5 @@ program main
b = y
call sub(a, b)
end program main
-! { dg-final { scan-tree-dump-times "memset" 0 "forwprop2" } }
+! { dg-final { scan-tree-dump-times "memset" 0 "forwprop2" { xfail { mips*-*-*
&& { ! nomips16 } } } } }
! { dg-final { cleanup-tree-dump "forwprop2" } }
Re: [PATCH] MIPS16 TLS support for GCC
Richard Sandiford writes:
> "Maciej W. Rozycki" writes:
>>> If there's still some concern that __mips16_rdhwr might not have
>>> the right ABI, then maybe we should simply emit a link-once function
>>> in each object that needs it. We could then switch to another function
>>> (and another API) without having to keep the old one in libgcc.a for
>>> compatibility. It would also avoid the -shared-libgcc thing.
>>>
>>> Admittedly that's just an off-the-top-of-my-head idea. :-)
>>> What do you think?
>>
>> Actually I had that idea of a link-once function too, but it turned out
>> quite complicated to do without rewriting some generic parts of GCC as it
>> is currently not prepared to emit link-once functions outside C++
>> compilations. It's been a while and I did lots of other stuff meanwhile,
>> so please excuse me if I got anything wrong here.
>
> Hmm, OK, I wouldn't have expected that. But if you've tried making
> __mips16_rdhwr link-once and had a bad experience with it, then yeah,
> let's go with the hidden libgcc function. It's just a shame that we're
> having to force static linking of libgcc for this one case.
>
> I'll take the relevant parts from Chung-Lin's patch and test them
> over the weekend.
Here's what I committed after testing on mips64-linux-gnu, both with
and without PLTs.
Richard
libgcc/
2012-02-05 Chung-Lin Tang
* config.host (mips64*-*-linux*, mipsisa64*-*-linux*, mips*-*-linux*):
Add t-slibgcc-libgcc to tmake_file.
* config/mips/libgcc-mips16.ver: Revert previous patch.
* config/mips/mips16.S (__mips16_rdhwr): Hide.
Index: libgcc/config.host
===
--- libgcc/config.host 2012-02-04 10:27:18.0 +
+++ libgcc/config.host 2012-02-04 11:34:03.0 +
@@ -746,12 +746,12 @@ mips*-*-netbsd*) # NetBSD/mips, either
;;
mips64*-*-linux* | mipsisa64*-*-linux*)
extra_parts="$extra_parts crtfastmath.o"
- tmake_file="${tmake_file} t-crtfm mips/t-mips16 mips/t-tpbit"
+ tmake_file="${tmake_file} t-crtfm mips/t-mips16 mips/t-tpbit
t-slibgcc-libgcc"
md_unwind_header=mips/linux-unwind.h
;;
mips*-*-linux*)# Linux MIPS, either endian.
extra_parts="$extra_parts crtfastmath.o"
- tmake_file="${tmake_file} t-crtfm mips/t-mips16"
+ tmake_file="${tmake_file} t-crtfm mips/t-mips16 t-slibgcc-libgcc"
md_unwind_header=mips/linux-unwind.h
;;
mips*-*-openbsd*)
Index: libgcc/config/mips/libgcc-mips16.ver
===
--- libgcc/config/mips/libgcc-mips16.ver2012-02-04 11:12:01.0
+
+++ libgcc/config/mips/libgcc-mips16.ver2012-02-04 11:34:03.0
+
@@ -84,7 +84,3 @@ GCC_4.4.0 {
__mips16_call_stub_dc_9
__mips16_call_stub_dc_10
}
-
-GCC_4.7.0 {
- __mips16_rdhwr
-}
Index: libgcc/config/mips/mips16.S
===
--- libgcc/config/mips/mips16.S 2012-02-04 10:27:18.0 +
+++ libgcc/config/mips/mips16.S 2012-02-04 11:34:03.0 +
@@ -712,6 +712,9 @@ CALL_STUB_RET (__mips16_call_stub_dc_10,
#ifdef L_m16rdhwr
STARTFN (__mips16_rdhwr)
+ /* Forced always hidden, because the PLT resolver function would
+ not preserve all necessary registers. */
+ .hidden __mips16_rdhwr
.setpush
.setmips32r2
.setnoreorder
[MIPS, committed] Fix lengths of MIPS16 call patterns
gcc.dg/fixed-point/convert.c and libgomp.fortran/vla3.f90 were failing
on MIPS16 for mips64-linux-gnu with PLTs enabled because we were
underestimating the length of a direct JAL.
Tested on mips64-linux-gnu and various ELF targets. Applied.
Richard
gcc/
* config/mips/mips.md (sibcall_internal, sibcall_value_internal)
(sibcall_value_multiple_internal, call_split, call_internal_direct)
(call_direct_split, call_value_split, call_value_internal_direct)
(call_value_direct_split, call_value_multiple_split): Use jal and
jal_macro attributes.
Index: gcc/config/mips/mips.md
===
--- gcc/config/mips/mips.md 2012-02-04 20:15:02.0 +
+++ gcc/config/mips/mips.md 2012-02-04 20:17:58.0 +
@@ -6119,7 +6119,8 @@ (define_insn "sibcall_internal"
(match_operand 1 "" ""))]
"TARGET_SIBCALLS && SIBLING_CALL_P (insn)"
{ return MIPS_CALL ("j", operands, 0, 1); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "indirect,direct")
+ (set_attr "jal_macro" "no")])
(define_expand "sibcall_value"
[(parallel [(set (match_operand 0 "")
@@ -6139,7 +6140,8 @@ (define_insn "sibcall_value_internal"
(match_operand 2 "" "")))]
"TARGET_SIBCALLS && SIBLING_CALL_P (insn)"
{ return MIPS_CALL ("j", operands, 1, 2); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "indirect,direct")
+ (set_attr "jal_macro" "no")])
(define_insn "sibcall_value_multiple_internal"
[(set (match_operand 0 "register_operand" "")
@@ -6150,7 +6152,8 @@ (define_insn "sibcall_value_multiple_int
(match_dup 2)))]
"TARGET_SIBCALLS && SIBLING_CALL_P (insn)"
{ return MIPS_CALL ("j", operands, 1, 2); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "indirect,direct")
+ (set_attr "jal_macro" "no")])
(define_expand "call"
[(parallel [(call (match_operand 0 "")
@@ -6214,13 +6217,14 @@ (define_insn_and_split "call_internal"
[(set_attr "jal" "indirect,direct")])
(define_insn "call_split"
- [(call (mem:SI (match_operand 0 "call_insn_operand" "cS"))
+ [(call (mem:SI (match_operand 0 "call_insn_operand" "c,S"))
(match_operand 1 "" ""))
(clobber (reg:SI RETURN_ADDR_REGNUM))
(clobber (reg:SI 28))]
"TARGET_SPLIT_CALLS"
{ return MIPS_CALL ("jal", operands, 0, 1); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "indirect,direct")
+ (set_attr "jal_macro" "no")])
;; A pattern for calls that must be made directly. It is used for
;; MIPS16 calls that the linker may need to redirect to a hard-float
@@ -6240,7 +6244,7 @@ (define_insn_and_split "call_internal_di
gen_call_direct_split (operands[0], operands[1]));
DONE;
}
- [(set_attr "type" "call")])
+ [(set_attr "jal" "direct")])
(define_insn "call_direct_split"
[(call (mem:SI (match_operand 0 "const_call_insn_operand"))
@@ -6250,7 +6254,8 @@ (define_insn "call_direct_split"
(clobber (reg:SI 28))]
"TARGET_SPLIT_CALLS"
{ return MIPS_CALL ("jal", operands, 0, -1); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "direct")
+ (set_attr "jal_macro" "no")])
(define_expand "call_value"
[(parallel [(set (match_operand 0 "")
@@ -6284,13 +6289,14 @@ (define_insn_and_split "call_value_inter
(define_insn "call_value_split"
[(set (match_operand 0 "register_operand" "")
-(call (mem:SI (match_operand 1 "call_insn_operand" "cS"))
+(call (mem:SI (match_operand 1 "call_insn_operand" "c,S"))
(match_operand 2 "" "")))
(clobber (reg:SI RETURN_ADDR_REGNUM))
(clobber (reg:SI 28))]
"TARGET_SPLIT_CALLS"
{ return MIPS_CALL ("jal", operands, 1, 2); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "indirect,direct")
+ (set_attr "jal_macro" "no")])
;; See call_internal_direct.
(define_insn_and_split "call_value_internal_direct"
@@ -6309,7 +6315,7 @@ (define_insn_and_split "call_value_inter
operands[2]));
DONE;
}
- [(set_attr "type" "call")])
+ [(set_attr "jal" "direct")])
(define_insn "call_value_direct_split"
[(set (match_operand 0 "register_operand")
@@ -6320,7 +6326,8 @@ (define_insn "call_value_direct_split"
(clobber (reg:SI 28))]
"TARGET_SPLIT_CALLS"
{ return MIPS_CALL ("jal", operands, 1, -1); }
- [(set_attr "type" "call")])
+ [(set_attr "jal" "direct")
+ (set_attr "jal_macro" "no")])
;; See comment for call_internal.
(define_insn_and_split "call_value_multiple_internal"
@@ -6345,7 +6352,7 @@ (define_insn_and_split "call_value_multi
(define_insn "call_value_multiple_split"
[(set (match_operand 0 "register_operand" "")
-(call (mem:SI (match_operand 1 "call_insn_operand" "cS"))
+(call (mem:SI (match_operand 1 "call_insn_operand" "c,S"))
(match_operand 2 "" "")))
(set (match_operand 3 "register_operand" "")
(call (mem:SI (match_dup 1))
@@ -6354,7 +6361,8 @@ (define_insn "call_
Documenting the MIPS changes in 4.7
I've committed this patch to describe the MIPS changes in GCC 4.7. Corrections, comments, and help with wordsmithing are all welcome. Thanks, Richard Index: htdocs/gcc-4.7/changes.html === RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-4.7/changes.html,v retrieving revision 1.74 diff -u -p -r1.74 changes.html --- htdocs/gcc-4.7/changes.html 16 Jan 2012 08:39:01 - 1.74 +++ htdocs/gcc-4.7/changes.html 5 Feb 2012 15:10:10 - @@ -51,6 +51,9 @@ Support has been removed for the NetWare x86 configuration obsoleted in GCC 4.6. +It is no longer possible to use the "l" +constraint in MIPS16 asm statements. + More information on porting to GCC 4.7 from previous versions of GCC can be found in the http://gcc.gnu.org/gcc-4.7/porting_to.html";>porting @@ -571,9 +574,29 @@ well. ... - + +GCC now supports thread-local storage (TLS) for MIPS16. +This support requires GNU binutils 2.22 or later. + +GCC can now generate code specifically for the Cavium Octeon+ +and Octeon2 processors. The associated command-line options are +-march=octeon+ and -march=octeon2 +respectively. Both options require GNU binutils 2.22 or later. + +GCC can now work around certain 24k errata, under the control +of the command-line option -mfix-24k. +These workarounds require GNU binutils 2.20 or later. + +32-bit MIPS GNU/Linux targets such as mips-linux-gnu +can now build n32 and n64 multilibs. The result is effectively +a 64-bit GNU/Linux toolchain that generates 32-bit code by default. +Use the configure-time option --enable-targets=all +to select these extra multilibs. + +Passing -fno-delayed-branch now also stops the +assembler from automatically filling delay slots. +
[PATCH 4.8, i386]: Enable post-reload compare optimization pass (PR28685)
Hello!
Attached patch enables post-reload compare optimization pass for x86 targets.
As shown in following test, patch removes redundant compare:
--cut here--
int test(int a, int b)
{
int lt = a < b;
int eq = a == b;
if (lt)
return 1;
return eq;
}
--cut here--
gcc -O2 on x86_64, a cmove target:
test:
xorl%edx, %edx
cmpl%esi, %edi
movl$1, %eax
sete%dl
cmovge %edx, %eax
ret
gcc-O2 -fno-compare-elim:
test:
xorl%edx, %edx
cmpl%esi, %edi
movl$1, %eax
sete%dl
cmpl%esi, %edi
cmovge %edx, %eax
ret
2012-02-05 Uros Bizjak
PR target/28685
* config/i386/i386.c (TARGET_FLAGS_REGNUM): New.
testsuite/ChangeLog:
PR target/28685
* gcc.target/i386/pr28685.c: New test.
The patch was tested on x86_64-pc-linux-gnu {,-m32}. The patch will be
committed to 4.8 once stage 1 opens.
Uros.
Index: config/i386/i386.c
===
--- config/i386/i386.c (revision 183904)
+++ config/i386/i386.c (working copy)
@@ -38685,6 +38685,9 @@
#undef TARGET_EXPAND_TO_RTL_HOOK
#define TARGET_EXPAND_TO_RTL_HOOK ix86_maybe_switch_abi
+#undef TARGET_FLAGS_REGNUM
+#define TARGET_FLAGS_REGNUM FLAGS_REG
+
#undef TARGET_LEGITIMATE_ADDRESS_P
#define TARGET_LEGITIMATE_ADDRESS_P ix86_legitimate_address_p
Index: testsuite/gcc.target/i386/pr28685.c
===
--- testsuite/gcc.target/i386/pr28685.c (revision 0)
+++ testsuite/gcc.target/i386/pr28685.c (revision 0)
@@ -0,0 +1,12 @@
+/* { dg-options "-O2" } */
+
+int test(int a, int b)
+{
+ int lt = a < b;
+ int eq = a == b;
+ if (lt)
+return 1;
+ return eq;
+}
+
+/* { dg-final { scan-assembler-times "cmp" 1 } } */
[wwwdocs] Add missing Steering Committee member
It has been brought to my attention that we have been missing a prominent member of the Steering Committee on our web page. Fixed thusly. Gerald Index: steering.html === RCS file: /cvs/gcc/wwwdocs/htdocs/steering.html,v retrieving revision 1.32 diff -u -3 -p -r1.32 steering.html --- steering.html 20 Nov 2011 11:57:38 - 1.32 +++ steering.html 5 Feb 2012 17:09:29 - @@ -40,6 +40,12 @@ place to reach them is the gcc
Re: [wwwdocs] Add missing Steering Committee member
On 02/05/2012 06:15 PM, Gerald Pfeifer wrote: It has been brought to my attention that we have been missing a prominent member of the Steering Committee on our web page. Fixed thusly. Gerald Index: steering.html === RCS file: /cvs/gcc/wwwdocs/htdocs/steering.html,v retrieving revision 1.32 diff -u -3 -p -r1.32 steering.html --- steering.html 20 Nov 2011 11:57:38 - 1.32 +++ steering.html 5 Feb 2012 17:09:29 - @@ -40,6 +40,12 @@ place to reach them is the gcc s/Stallmann/Stallman Paolo.
Re: [PATCH 4.8, i386]: Enable post-reload compare optimization pass (PR28685)
On Sun, Feb 5, 2012 at 4:27 PM, Uros Bizjak wrote: > 2012-02-05 Uros Bizjak > > PR target/28685 > * config/i386/i386.c (TARGET_FLAGS_REGNUM): New. > Hmm, how is this (apparently new in 2011) TARGET_FLAGS_REGNUM different from the older targetm.fixed_condition_code_regs? Ciao! Steven
Re: [wwwdocs] Add missing Steering Committee member
On Sun, 5 Feb 2012, Paolo Carlini wrote: >> +Richard Stallmann (Free Software Foundation) > s/Stallmann/Stallman Oops, of course! Too much German recently. :-o Thanks, Paolo. Fixed right away. Gerald
Re: [PATCH 4.8, i386]: Enable post-reload compare optimization pass (PR28685)
On Sun, Feb 5, 2012 at 6:27 PM, Steven Bosscher wrote: >> PR target/28685 >> * config/i386/i386.c (TARGET_FLAGS_REGNUM): New. >> > > Hmm, how is this (apparently new in 2011) TARGET_FLAGS_REGNUM > different from the older targetm.fixed_condition_code_regs? This is how backend enables the pass, please see gate_compare_elim_after_reload function in gcc/compare-elim.c. FWIW, the value could be anything != INVALID_REGNUM. Uros.
Re: [patch, fortran] Fix PR 48847, bogus "unused parameter" warning
On Sun, Feb 05, 2012 at 11:12:01AM +0100, Thomas Koenig wrote: > Hello world, > > the attached, rather obvious patch fixed a bogus "unused parameter" > warning with procedure dummy arguments and warns about these if > they occur. > > Regression-tested. OK for trunk? Or rather wait until 4.8? > You need a space after TREE_NO_WARNING. I think that the patch can be applied now given its simplicity. -- Steve
[v3] libstdc++/51956 improve pretty printers for shared_ptr and weak_ptr
PR libstdc++/51956
* python/libstdcxx/v6/printers.py (StdPointerPrinter): Rename to...
(SharedPointerPrinter): This. Also show weak count.
* testsuite/libstdc++-prettyprinters/shared_ptr.cc: New.
tested x86_64-linux, committed to trunk.
diff --git a/libstdc++-v3/python/libstdcxx/v6/printers.py
b/libstdc++-v3/python/libstdcxx/v6/printers.py
index 4f34733..f47da61 100644
--- a/libstdc++-v3/python/libstdcxx/v6/printers.py
+++ b/libstdc++-v3/python/libstdcxx/v6/printers.py
@@ -45,19 +45,24 @@ def find_type(orig, name):
raise ValueError, "Cannot find type %s::%s" % (str(orig), name)
typ = field.type
-class StdPointerPrinter:
-"Print a smart pointer of some kind"
+class SharedPointerPrinter:
+"Print a shared_ptr or weak_ptr"
def __init__ (self, typename, val):
self.typename = typename
self.val = val
def to_string (self):
-if self.val['_M_refcount']['_M_pi'] == 0:
-return '%s (empty) %s' % (self.typename, self.val['_M_ptr'])
-return '%s (count %d) %s' % (self.typename,
-
self.val['_M_refcount']['_M_pi']['_M_use_count'],
- self.val['_M_ptr'])
+state = 'empty'
+refcounts = self.val['_M_refcount']['_M_pi']
+if refcounts != 0:
+usecount = refcounts['_M_use_count']
+weakcount = refcounts['_M_weak_count']
+if usecount == 0:
+state = 'expired, weak %d' % weakcount
+else:
+state = 'count %d, weak %d' % (usecount, weakcount - 1)
+return '%s (%s) %s' % (self.typename, state, self.val['_M_ptr'])
class UniquePointerPrinter:
"Print a unique_ptr"
@@ -862,8 +867,8 @@ def build_libstdcxx_dictionary ():
# These are the TR1 and C++0x printers.
# For array - the default GDB pretty-printer seems reasonable.
-libstdcxx_printer.add_version('std::', 'shared_ptr', StdPointerPrinter)
-libstdcxx_printer.add_version('std::', 'weak_ptr', StdPointerPrinter)
+libstdcxx_printer.add_version('std::', 'shared_ptr', SharedPointerPrinter)
+libstdcxx_printer.add_version('std::', 'weak_ptr', SharedPointerPrinter)
libstdcxx_printer.add_container('std::', 'unordered_map',
Tr1UnorderedMapPrinter)
libstdcxx_printer.add_container('std::', 'unordered_set',
@@ -875,8 +880,8 @@ def build_libstdcxx_dictionary ():
libstdcxx_printer.add_container('std::', 'forward_list',
StdForwardListPrinter)
-libstdcxx_printer.add_version('std::tr1::', 'shared_ptr',
StdPointerPrinter)
-libstdcxx_printer.add_version('std::tr1::', 'weak_ptr', StdPointerPrinter)
+libstdcxx_printer.add_version('std::tr1::', 'shared_ptr',
SharedPointerPrinter)
+libstdcxx_printer.add_version('std::tr1::', 'weak_ptr',
SharedPointerPrinter)
libstdcxx_printer.add_version('std::tr1::', 'unordered_map',
Tr1UnorderedMapPrinter)
libstdcxx_printer.add_version('std::tr1::', 'unordered_set',
diff --git a/libstdc++-v3/testsuite/libstdc++-prettyprinters/shared_ptr.cc
b/libstdc++-v3/testsuite/libstdc++-prettyprinters/shared_ptr.cc
new file mode 100644
index 000..9328c5f9
--- /dev/null
+++ b/libstdc++-v3/testsuite/libstdc++-prettyprinters/shared_ptr.cc
@@ -0,0 +1,83 @@
+// { dg-do run }
+// { dg-options "-std=gnu++11 -g" }
+
+// Copyright (C) 2012 Free Software Foundation, Inc.
+//
+// This file is part of the GNU ISO C++ Library. This library is free
+// software; you can redistribute it and/or modify it under the
+// terms of the GNU General Public License as published by the
+// Free Software Foundation; either version 3, or (at your option)
+// any later version.
+
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU General Public License for more details.
+
+// You should have received a copy of the GNU General Public License along
+// with this library; see the file COPYING3. If not see
+//
Re: Documenting the MIPS changes in 4.7
On Sun, Feb 5, 2012 at 7:13 AM, Richard Sandiford wrote: > I've committed this patch to describe the MIPS changes in GCC 4.7. > Corrections, comments, and help with wordsmithing are all welcome. > > Thanks, > Richard > > > Index: htdocs/gcc-4.7/changes.html > === > RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-4.7/changes.html,v > retrieving revision 1.74 > diff -u -p -r1.74 changes.html > --- htdocs/gcc-4.7/changes.html 16 Jan 2012 08:39:01 - 1.74 > +++ htdocs/gcc-4.7/changes.html 5 Feb 2012 15:10:10 - > @@ -51,6 +51,9 @@ > Support has been removed for the NetWare x86 configuration > obsoleted in GCC 4.6. > > + It is no longer possible to use the "l" > + constraint in MIPS16 asm statements. > + > More information on porting to GCC 4.7 from previous versions > of GCC can be found in > the http://gcc.gnu.org/gcc-4.7/porting_to.html";>porting > @@ -571,9 +574,29 @@ well. > ... > > > - > + > + GCC now supports thread-local storage (TLS) for MIPS16. > + This support requires GNU binutils 2.22 or later. > + > + GCC can now generate code specifically for the Cavium Octeon+ > + and Octeon2 processors. The associated command-line options are > + -march=octeon+ and -march=octeon2 > + respectively. Both options require GNU binutils 2.22 or later. This part looks good from my point of view. Thanks, Anrew Pinski > + > + GCC can now work around certain 24k errata, under the control > + of the command-line option -mfix-24k. > + These workarounds require GNU binutils 2.20 or later. > + > + 32-bit MIPS GNU/Linux targets such as mips-linux-gnu > + can now build n32 and n64 multilibs. The result is effectively > + a 64-bit GNU/Linux toolchain that generates 32-bit code by default. > + Use the configure-time option --enable-targets=all > + to select these extra multilibs. > + > + Passing -fno-delayed-branch now also stops the > + assembler from automatically filling delay slots. > + > >
Re: [Patch, fortran] PR52102 - [OOP] Wrong result with ALLOCATE of CLASS components with array constructor SOURCE-expr
Committed as revision 183915 after green light from Tobias on #gfortran. 2012-02-05 Paul Thomas * trans-array.c (gfc_array_allocate): Zero memory for all class array allocations. * trans-stmt.c (gfc_trans_allocate): Ditto for class scalars. PR fortran/52102 * trans-stmt.c (gfc_trans_allocate): Before correcting a class array reference, ensure that 'dataref' points to the _data component that is followed by the array reference.. 2012-02-05 Paul Thomas PR fortran/52102 * gfortran.dg/class_48.f90 : Add test of allocate class array component with source in subroutine test3. Remove commenting out in subroutine test4, since branching on unitialized variable is now fixed (no PR for this last.). Paul
Re: [patch, fortran] Fix PR 48847, bogus "unused parameter" warning
Hi Steve, You need a space after TREE_NO_WARNING. I think that the patch can be applied now given its simplicity. Committed as rev. 183916 with the space. Thanks a lot for the review! Thomas
Gthreads patch to disable static initializer macros
PRs libstdc++/51296 and libstdc++/51906 are both caused by problems
with the Pthreads static initializer macros such as
PTHREAD_MUTEX_INITIALIZER.
On Tru64 PTHREAD_MUTEX_INITIALIZER and PTHREAD_COND_INITIALIZER can
only be used for statically-initialized variables, as allowed by POSIX
currently, but http://austingroupbugs.net/view.php?id=70#c127 removes
that limitation for the next POSIX standard. C++11 needs to use those
macros for variables with automatic and dynamic scope, because the
init functions can't be used in constexpr constructors.
On Mac OS X 10.7 the PTHREAD_RECURSIVE_MUTEX_INITIALIZER is buggy.
This has shown up now because C++11 threads weren't enabled on darwin
before 4.7
My suggestion is to modify gthr-posix.h with the attached patch, so
that target maintainers can define e.g.
_GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC to force use of the init
function instead of the __GTHREAD_RECURSIVE_MUTEX_INIT initializer.
The patch includes a change to do that for darwin.
This has been testing on darwin, if Rainer tests it successfully on
Tru64 will the gthr-posix.h change be acceptable?
diff --git a/libgcc/gthr-posix.h b/libgcc/gthr-posix.h
index 46054f6..45b15a8 100644
--- a/libgcc/gthr-posix.h
+++ b/libgcc/gthr-posix.h
@@ -74,6 +74,20 @@ typedef struct timespec __gthread_time_t;
#define __GTHREAD_COND_INIT PTHREAD_COND_INITIALIZER
#define __GTHREAD_TIME_INIT {0,0}
+#ifdef _GTHREAD_USE_MUTEX_INIT_FUNC
+# undef __GTHREAD_MUTEX_INIT
+# define __GTHREAD_MUTEX_INIT_FUNCTION __gthread_mutex_init_function
+#endif
+#ifdef _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC
+# undef __GTHREAD_RECURSIVE_MUTEX_INIT
+# undef __GTHREAD_RECURSIVE_MUTEX_INIT_FUNCTION
+# define __GTHREAD_RECURSIVE_MUTEX_INIT_FUNCTION
__gthread_recursive_mutex_init_function
+#endif
+#ifdef _GTHREAD_USE_COND_INIT_FUNC
+# undef __GTHREAD_COND_INIT
+# define __GTHREAD_COND_INIT_FUNCTION __gthread_cond_init_function
+#endif
+
#if SUPPORTS_WEAK && GTHREAD_USE_WEAK
# ifndef __gthrw_pragma
# define __gthrw_pragma(pragma)
@@ -730,6 +744,15 @@ __gthread_setspecific (__gthread_key_t __key, const void
*__ptr)
return __gthrw_(pthread_setspecific) (__key, __ptr);
}
+#ifdef _GTHREAD_USE_MUTEX_INIT_FUNC
+static inline void
+__gthread_mutex_init_function (__gthread_mutex_t *__mutex)
+{
+ if (__gthread_active_p ())
+__gthrw_(pthread_mutex_init) (__mutex, NULL);
+}
+#endif
+
static inline int
__gthread_mutex_destroy (__gthread_mutex_t *__mutex)
{
@@ -778,7 +801,8 @@ __gthread_mutex_unlock (__gthread_mutex_t *__mutex)
return 0;
}
-#ifndef PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP
+#if !defined( PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP) \
+ || defined(_GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC)
static inline int
__gthread_recursive_mutex_init_function (__gthread_recursive_mutex_t *__mutex)
{
@@ -828,6 +852,15 @@ __gthread_recursive_mutex_unlock
(__gthread_recursive_mutex_t *__mutex)
return __gthread_mutex_unlock (__mutex);
}
+#ifdef _GTHREAD_USE_COND_INIT_FUNC
+static inline void
+__gthread_cond_init_function (__gthread_cond_t *__cond)
+{
+ if (__gthread_active_p ())
+__gthrw_(pthread_cond_init) (__cond, NULL);
+}
+#endif
+
static inline int
__gthread_cond_broadcast (__gthread_cond_t *__cond)
{
diff --git a/libstdc++-v3/config/os/bsd/darwin/os_defines.h
b/libstdc++-v3/config/os/bsd/darwin/os_defines.h
index ccefeaf..421478d 100644
--- a/libstdc++-v3/config/os/bsd/darwin/os_defines.h
+++ b/libstdc++-v3/config/os/bsd/darwin/os_defines.h
@@ -39,4 +39,7 @@
// -flat_namespace to work around the way that it doesn't.
#define _GLIBCXX_WEAK_DEFINITION __attribute__ ((weak))
+// Static initializer macro is buggy in darwin, see libstdc++/51906
+#define _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC
+
#endif
Re: [wwwdocs] Add missing Steering Committee member
On Sun, 5 Feb 2012, Gerald Pfeifer wrote: > It has been brought to my attention that we have been missing > a prominent member of the Steering Committee on our web page. Should the list distinguish which members are or are not GNU maintainers of GCC? fencepost:/gd/gnuorg/maintainers lists only the other twelve people as maintainers for GCC. -- Joseph S. Myers jos...@codesourcery.com
Added test case for PR 32373, missed vectorization with equivalence
Hello world,
I added the attached test case from the PR after verifying that all
loops mentioned in the PR are now vectorized. The PR is now closed.
2012-02-05 Thomas König
PR fortran/32373
* gfortran.dg/vect/vect-8.f90: New test case.
! PR fortran/32373 - eight of these loops were not vectorized.
! { dg-do compile }
! { dg-require-effective-target vect_float }
module lfk_prec
integer, parameter :: dp=kind(1.d0)
end module lfk_prec
!***
SUBROUTINE kernel(tk)
!***
! *
!KERNEL executes 24 samples of Fortran computation *
! TK(1) - total cpu time to execute only the 24 kernels. *
! TK(2) - total Flops executed by the 24 Kernels *
!***
! *
! L. L. N. L. F O R T R A N K E R N E L S: M F L O P S *
! *
! These kernels measure Fortran numerical computation rates for a *
! spectrum of CPU-limited computational structures. Mathematical *
! through-put is measured in units of millions of floating-point *
! operations executed per Second, called Mega-Flops/Sec. *
! *
! This program measures a realistic CPU performance range for the *
! Fortran programming system on a given day. The CPU performance *
! rates depend strongly on the maturity of the Fortran compiler's *
! ability to translate Fortran code into efficient machine code. *
! [ The CPU hardware capability apart from compiler maturity (or *
! availability), could be measured (or simulated) by programming the *
! kernels in assembly or machine code directly. These measurements *
! can also serve as a framework for tracking the maturation of the *
! Fortran compiler during system development.] *
! *
! Fonzi's Law: There is not now and there never will be a language *
! in which it is the least bit difficult to write *
! bad programs. *
!F.H.MCMAHON 1972 *
!***
! l1 := param-dimension governs the size of most 1-d arrays
! l2 := param-dimension governs the size of most 2-d arrays
! Loop := multiple pass control to execute kernel long enough to ti
!me.
! n := DO loop control for each kernel. Controls are set in subr.
! SIZES
! **
use lfk_prec
implicit double precision (a-h,o-z)
!IBM IMPLICIT REAL*8 (A-H,O-Z)
REAL(kind=dp), INTENT(inout):: tk
INTEGER :: test !!,AND
COMMON/alpha/mk,ik,im,ml,il,mruns,nruns,jr,iovec,npfs(8,3,47)
COMMON/beta/tic,times(8,3,47),see(5,3,8,3),terrs(8,3,47),csums(8,3 &
,47),fopn(8,3,47),dos(8,3,47)
COMMON/spaces/ion,j5,k2,k3,loop1,laps,loop,m,kr,lp,n13h,ibuf,nx,l, &
npass,nfail,n,n1,n2,n13,n213,n813,n14,n16,n416,n21,nt1,nt2,last,idebug &
,mpy,loop2,mucho,mpylim,intbuf(16)
COMMON/spacer/a11,a12,a13,a21,a22,a23,a31,a32,a33,ar,br,c0,cr,di,dk &
,dm22,dm23,dm24,dm25,dm26,dm27,dm28,dn,e3,e6,expmax,flx,q,qa,r,ri &
,s,scale,sig,stb5,t,xnc,xnei,xnm
COMMON/space0/time(47),csum(47),ww(47),wt(47),ticks,fr(9),terr1(47 &
),sumw(7),start,skale(47),bias(47),ws(95),total(47),flopn(47),iq(7 &
),npf,npfs1(47)
COMMON/spacei/wtp(3),mul(3),ispan(47,3),ipass(47,3)
! **
INTEGER :: e,f,zone
COMMON/ispace/e(96),f(96),ix(1001),ir(1001),zone(300)
COMMON/space1/u(1001),v(1001),w(1001),x(1001),y(1001),z(1001),g(1001) &
,du1(101),du2(101),du3(101),grd(1001),dex(1001),xi(1001),ex(1001) &
,ex1(1001),dex1(1001),vx(1001),xx(1001),rx(1001),rh(2048),vsp(101) &
,vstp(101),vxne(101),vxnd(101),ve3(101),vlr(101),vlin(101),b5(101) &
,plan(300),d(300),sa(101),sb(101)
COMMON/space2/p(4,512),px(25,101),cx(25,101),vy(101,25),vh(101,7), &
vf(101,7),vg(101,7),vs(101,7),za(101,7),zp(101,7),zq(101,7),zr(101 &
,7),zm(101,7),zb(101,7),zu(101,7),zv(101,7),zz(101,7),b(64,64),c(64,64) &
,h(64,64),u1(5,101,2),u2(5,101,2),u3(5,101,2)
! **
dimension zx(1023),xz(447,3),tk(6),mtmp(1)
EQUIVALENCE(zx(1),z(1)),(xz(1,1),x(1))
double precision temp
logical ltmp
! **
! STANDARD PRO
New German PO file for 'gcc' (version 4.7-b20120128)
Hello, gentle maintainer. This is a message from the Translation Project robot. A revised PO file for textual domain 'gcc' has been submitted by the German team of translators. The file is available at: http://translationproject.org/latest/gcc/de.po (This file, 'gcc-4.7-b20120128.de.po', has just now been sent to you in a separate email.) All other PO files for your package are available in: http://translationproject.org/latest/gcc/ Please consider including all of these in your next release, whether official or a pretest. Whenever you have a new distribution with a new version number ready, containing a newer POT file, please send the URL of that distribution tarball to the address below. The tarball may be just a pretest or a snapshot, it does not even have to compile. It is just used by the translators when they need some extra translation context. The following HTML page has been updated: http://translationproject.org/domain/gcc.html If any question arises, please contact the translation coordinator. Thank you for all your work, The Translation Project robot, in the name of your translation coordinator.
Re: Gthreads patch to disable static initializer macros
On Sun, Feb 05, 2012 at 08:26:22PM +, Jonathan Wakely wrote:
> PRs libstdc++/51296 and libstdc++/51906 are both caused by problems
> with the Pthreads static initializer macros such as
> PTHREAD_MUTEX_INITIALIZER.
>
> On Tru64 PTHREAD_MUTEX_INITIALIZER and PTHREAD_COND_INITIALIZER can
> only be used for statically-initialized variables, as allowed by POSIX
> currently, but http://austingroupbugs.net/view.php?id=70#c127 removes
> that limitation for the next POSIX standard. C++11 needs to use those
> macros for variables with automatic and dynamic scope, because the
> init functions can't be used in constexpr constructors.
>
> On Mac OS X 10.7 the PTHREAD_RECURSIVE_MUTEX_INITIALIZER is buggy.
> This has shown up now because C++11 threads weren't enabled on darwin
> before 4.7
>
> My suggestion is to modify gthr-posix.h with the attached patch, so
> that target maintainers can define e.g.
> _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC to force use of the init
> function instead of the __GTHREAD_RECURSIVE_MUTEX_INIT initializer.
> The patch includes a change to do that for darwin.
>
> This has been testing on darwin, if Rainer tests it successfully on
> Tru64 will the gthr-posix.h change be acceptable?
There are no unexpected libstdc++ testsuite regressions on
x86_64-apple-darwin11 with this
patch applied to current gcc trunk.
> diff --git a/libgcc/gthr-posix.h b/libgcc/gthr-posix.h
> index 46054f6..45b15a8 100644
> --- a/libgcc/gthr-posix.h
> +++ b/libgcc/gthr-posix.h
> @@ -74,6 +74,20 @@ typedef struct timespec __gthread_time_t;
> #define __GTHREAD_COND_INIT PTHREAD_COND_INITIALIZER
> #define __GTHREAD_TIME_INIT {0,0}
>
> +#ifdef _GTHREAD_USE_MUTEX_INIT_FUNC
> +# undef __GTHREAD_MUTEX_INIT
> +# define __GTHREAD_MUTEX_INIT_FUNCTION __gthread_mutex_init_function
> +#endif
> +#ifdef _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC
> +# undef __GTHREAD_RECURSIVE_MUTEX_INIT
> +# undef __GTHREAD_RECURSIVE_MUTEX_INIT_FUNCTION
> +# define __GTHREAD_RECURSIVE_MUTEX_INIT_FUNCTION
> __gthread_recursive_mutex_init_function
> +#endif
> +#ifdef _GTHREAD_USE_COND_INIT_FUNC
> +# undef __GTHREAD_COND_INIT
> +# define __GTHREAD_COND_INIT_FUNCTION __gthread_cond_init_function
> +#endif
> +
> #if SUPPORTS_WEAK && GTHREAD_USE_WEAK
> # ifndef __gthrw_pragma
> # define __gthrw_pragma(pragma)
> @@ -730,6 +744,15 @@ __gthread_setspecific (__gthread_key_t __key, const void
> *__ptr)
>return __gthrw_(pthread_setspecific) (__key, __ptr);
> }
>
> +#ifdef _GTHREAD_USE_MUTEX_INIT_FUNC
> +static inline void
> +__gthread_mutex_init_function (__gthread_mutex_t *__mutex)
> +{
> + if (__gthread_active_p ())
> +__gthrw_(pthread_mutex_init) (__mutex, NULL);
> +}
> +#endif
> +
> static inline int
> __gthread_mutex_destroy (__gthread_mutex_t *__mutex)
> {
> @@ -778,7 +801,8 @@ __gthread_mutex_unlock (__gthread_mutex_t *__mutex)
> return 0;
> }
>
> -#ifndef PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP
> +#if !defined( PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP) \
> + || defined(_GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC)
> static inline int
> __gthread_recursive_mutex_init_function (__gthread_recursive_mutex_t
> *__mutex)
> {
> @@ -828,6 +852,15 @@ __gthread_recursive_mutex_unlock
> (__gthread_recursive_mutex_t *__mutex)
>return __gthread_mutex_unlock (__mutex);
> }
>
> +#ifdef _GTHREAD_USE_COND_INIT_FUNC
> +static inline void
> +__gthread_cond_init_function (__gthread_cond_t *__cond)
> +{
> + if (__gthread_active_p ())
> +__gthrw_(pthread_cond_init) (__cond, NULL);
> +}
> +#endif
> +
> static inline int
> __gthread_cond_broadcast (__gthread_cond_t *__cond)
> {
> diff --git a/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> b/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> index ccefeaf..421478d 100644
> --- a/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> +++ b/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> @@ -39,4 +39,7 @@
> // -flat_namespace to work around the way that it doesn't.
> #define _GLIBCXX_WEAK_DEFINITION __attribute__ ((weak))
>
> +// Static initializer macro is buggy in darwin, see libstdc++/51906
> +#define _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC
> +
> #endif
Re: Gthreads patch to disable static initializer macros
On 5 February 2012 23:53, Jack Howarth wrote: > > There are no unexpected libstdc++ testsuite regressions on > x86_64-apple-darwin11 with this > patch applied to current gcc trunk. Thanks, Jack. If I can't get approval for the change I'll workaround it in libstdc++, but IMHO it would be cleaner to change gthr-posix.h to not define macros that can't be used safely.
libstdc++/52104 - fix linker error for non-TLS targets
The out-of-line destructor in future.cc uses std::call_once which uses
a lambda on TLS targets. Because we compile with
-fno-implicit-templates an explicit instantiation is needed, but can't
be done because the closure type created by the lambda can't be named.
This puts the destructor inline in the header for TLS targets. A
better fix would be to avoid using a lambda in std::call_once, but I
don't want to touch that code instage4.
PR libstdc++/52104
* include/std/future (__future_base::_Async_state_common): Define
destructor inline for targets without TLS.
* src/c++11/future.cc (__future_base::_Async_state_common): Only
define destructor for TLS targets.
Tested x86_64-linux by me and on sparc-solaris by Eric Botcazou.
Committed to trunk.
diff --git a/libstdc++-v3/include/std/future b/libstdc++-v3/include/std/future
index 1093e3f..962400b 100644
--- a/libstdc++-v3/include/std/future
+++ b/libstdc++-v3/include/std/future
@@ -1425,7 +1425,11 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
class __future_base::_Async_state_common : public __future_base::_State_base
{
protected:
+#ifdef _GLIBCXX_HAVE_TLS
~_Async_state_common();
+#else
+~_Async_state_common() { _M_join(); }
+#endif
// Allow non-timed waiting functions to block until the thread completes,
// as if joined.
diff --git a/libstdc++-v3/src/c++11/future.cc b/libstdc++-v3/src/c++11/future.cc
index dab0774..61a9729 100644
--- a/libstdc++-v3/src/c++11/future.cc
+++ b/libstdc++-v3/src/c++11/future.cc
@@ -85,11 +85,13 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
__future_base::_State_base::~_State_base() = default;
+#ifdef _GLIBCXX_HAVE_TLS
__future_base::_Async_state_common::~_Async_state_common() { _M_join(); }
// Explicit instantiation due to -fno-implicit-instantiation.
template void call_once(once_flag&, void (thread::*&&)(),
reference_wrapper&&);
#endif
+#endif
_GLIBCXX_END_NAMESPACE_VERSION
} // namespace std
[patch] document which edition of Effective C++ is used for -Weffc++
PR c++/48680
* doc/invoke.texi (C++ Dialect Options): Use @option markup for
-Weffc++ and specify guidelines come from second edition.
Tested with 'make doc html'
OK for trunk?
diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
index 6105710..6c61e53 100644
--- a/gcc/doc/invoke.texi
+++ b/gcc/doc/invoke.texi
@@ -2437,7 +2437,7 @@ the compiler to never throw an exception.
Warn when a class has virtual functions and accessible non-virtual
destructor, in which case it would be possible but unsafe to delete
an instance of a derived class through a pointer to the base class.
-This warning is also enabled if -Weffc++ is specified.
+This warning is also enabled if @option{-Weffc++} is specified.
@item -Wreorder @r{(C++ and Objective-C++ only)}
@opindex Wreorder
@@ -2467,7 +2467,7 @@ The following @option{-W@dots{}} options are not affected
by @option{-Wall}.
@opindex Weffc++
@opindex Wno-effc++
Warn about violations of the following style guidelines from Scott Meyers'
-@cite{Effective C++} book:
+@cite{Effective C++, Second Edition} book:
@itemize @bullet
@item
Re: [patch] document which edition of Effective C++ is used for -Weffc++
On Mon, 6 Feb 2012, Jonathan Wakely wrote: > PR c++/48680 > * doc/invoke.texi (C++ Dialect Options): Use @option markup for > -Weffc++ and specify guidelines come from second edition. > > Tested with 'make doc html' > OK for trunk? Yes. The first is a bug fix, the second obvious (for those in the know, before your patch :-). Gerald
Re: [patch] document which edition of Effective C++ is used for -Weffc++
On 6 February 2012 01:04, Gerald Pfeifer wrote: > On Mon, 6 Feb 2012, Jonathan Wakely wrote: >> PR c++/48680 >> * doc/invoke.texi (C++ Dialect Options): Use @option markup for >> -Weffc++ and specify guidelines come from second edition. >> >> Tested with 'make doc html' >> OK for trunk? > > Yes. The first is a bug fix, the second obvious (for those in > the know, before your patch :-). Thanks, I've checked it in.
Re: [Patch] PR 52118 (Description of Wunused-local-typedefs)
OK. Jason
Re: Gthreads patch to disable static initializer macros
On Sun, Feb 05, 2012 at 08:26:22PM +, Jonathan Wakely wrote:
> This has been testing on darwin, if Rainer tests it successfully on
> Tru64 will the gthr-posix.h change be acceptable?
Ok with a suitable ChangeLog entry.
> diff --git a/libgcc/gthr-posix.h b/libgcc/gthr-posix.h
> index 46054f6..45b15a8 100644
> --- a/libgcc/gthr-posix.h
> +++ b/libgcc/gthr-posix.h
> @@ -74,6 +74,20 @@ typedef struct timespec __gthread_time_t;
> #define __GTHREAD_COND_INIT PTHREAD_COND_INITIALIZER
> #define __GTHREAD_TIME_INIT {0,0}
>
> +#ifdef _GTHREAD_USE_MUTEX_INIT_FUNC
> +# undef __GTHREAD_MUTEX_INIT
> +# define __GTHREAD_MUTEX_INIT_FUNCTION __gthread_mutex_init_function
> +#endif
> +#ifdef _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC
> +# undef __GTHREAD_RECURSIVE_MUTEX_INIT
> +# undef __GTHREAD_RECURSIVE_MUTEX_INIT_FUNCTION
> +# define __GTHREAD_RECURSIVE_MUTEX_INIT_FUNCTION
> __gthread_recursive_mutex_init_function
> +#endif
> +#ifdef _GTHREAD_USE_COND_INIT_FUNC
> +# undef __GTHREAD_COND_INIT
> +# define __GTHREAD_COND_INIT_FUNCTION __gthread_cond_init_function
> +#endif
> +
> #if SUPPORTS_WEAK && GTHREAD_USE_WEAK
> # ifndef __gthrw_pragma
> # define __gthrw_pragma(pragma)
> @@ -730,6 +744,15 @@ __gthread_setspecific (__gthread_key_t __key, const void
> *__ptr)
>return __gthrw_(pthread_setspecific) (__key, __ptr);
> }
>
> +#ifdef _GTHREAD_USE_MUTEX_INIT_FUNC
> +static inline void
> +__gthread_mutex_init_function (__gthread_mutex_t *__mutex)
> +{
> + if (__gthread_active_p ())
> +__gthrw_(pthread_mutex_init) (__mutex, NULL);
> +}
> +#endif
> +
> static inline int
> __gthread_mutex_destroy (__gthread_mutex_t *__mutex)
> {
> @@ -778,7 +801,8 @@ __gthread_mutex_unlock (__gthread_mutex_t *__mutex)
> return 0;
> }
>
> -#ifndef PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP
> +#if !defined( PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP) \
> + || defined(_GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC)
> static inline int
> __gthread_recursive_mutex_init_function (__gthread_recursive_mutex_t
> *__mutex)
> {
> @@ -828,6 +852,15 @@ __gthread_recursive_mutex_unlock
> (__gthread_recursive_mutex_t *__mutex)
>return __gthread_mutex_unlock (__mutex);
> }
>
> +#ifdef _GTHREAD_USE_COND_INIT_FUNC
> +static inline void
> +__gthread_cond_init_function (__gthread_cond_t *__cond)
> +{
> + if (__gthread_active_p ())
> +__gthrw_(pthread_cond_init) (__cond, NULL);
> +}
> +#endif
> +
> static inline int
> __gthread_cond_broadcast (__gthread_cond_t *__cond)
> {
> diff --git a/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> b/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> index ccefeaf..421478d 100644
> --- a/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> +++ b/libstdc++-v3/config/os/bsd/darwin/os_defines.h
> @@ -39,4 +39,7 @@
> // -flat_namespace to work around the way that it doesn't.
> #define _GLIBCXX_WEAK_DEFINITION __attribute__ ((weak))
>
> +// Static initializer macro is buggy in darwin, see libstdc++/51906
> +#define _GTHREAD_USE_RECURSIVE_MUTEX_INIT_FUNC
> +
> #endif
Jakub
