New Japanese PO file for 'gcc' (version 4.6.1)
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 Japanese team of translators. The file is available at: http://translationproject.org/latest/gcc/ja.po (This file, 'gcc-4.6.1.ja.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: [libcpp,lto,fortran PATCH] Fix linemap_add use and remove unnecessary kludge
On Sat, Aug 27, 2011 at 9:04 PM, Dodji Seketeli wrote: > Hello Richard, > > Richard Guenther writes: > >> In the LTO FE the two linemap_add calls were to advance the location >> counter to cover the builtin special locations. You exchange these >> with only one - that doesn't look correct without more explanation. > > It seems to me that you don't need to worry about advancing the location > counter to cover builtin special locations because the lowest possible > location that could be handed out by the line map is set to > RESERVED_LOCATION_COUNT. You can see that by looking at linemap_init > that sets the initial highest location to RESERVED_LOCATION_COUNT - 1, > and at linemap_add that sets the starting location of the added map to > the highest location + 1. And RESERVED_LOCATION is set to 2 in > line-map.h. Hm, ok. That must have changed since the introduction of LTO then. The LTO bits are ok as well. Thanks, Richard. >> The PCH change is ok. > > Thank you. > > -- > Dodji >
[PATCH] fix -Werror=maybe-uninitialized in tree-ssa-loop-ivopts.c (PR bootstrap/50218)
The recent r178104 change to tree-ssa-loop-ivopts.c broke bootstrap on i686-linux because stage2 gcc considers the variable 'comp' used but possibly uninitialized, see PR50218. Initializing it to ERROR_MARK solves the problem, and is consistent with how other parts of r178104 represent absent/impossible 'comp' values. Ok for trunk? (I don't have svn write access.) /Mikael gcc/ 2011-08-20 Mikael Pettersson PR bootstrap/50218 * tree-ssa-loop-ivopts.c (determine_use_iv_cost_condition): Initialize comp. --- gcc-4.7-20110827/gcc/tree-ssa-loop-ivopts.c.~1~ 2011-08-26 15:36:28.0 +0200 +++ gcc-4.7-20110827/gcc/tree-ssa-loop-ivopts.c 2011-08-28 10:13:34.0 +0200 @@ -4757,7 +4757,7 @@ determine_use_iv_cost_condition (struct bool ok; int elim_inv_expr_id = -1, express_inv_expr_id = -1, inv_expr_id; tree *control_var, *bound_cst; - enum tree_code comp; + enum tree_code comp = ERROR_MARK; /* Only consider real candidates. */ if (!cand->iv)
[patch, fortran, committed] Regenerate (almost) all files in libgfortran/generated
Hello world, I have just regenerated all but one ofthe files in libgfortran/generated/ after regression-testing. This completed the fix for PR 40866, where I had left out a few files (possibly because of missing dependencies). It also clears up a minor inconsistency without any consequence (just a comment change). I'll send a separate e-mail about the one inconsistency left in a few minutes. Thomas 2011-08-28 Thomas Koenig PR fortran/40866 * generated/iall_i1.c: Regenerated. * generated/iall_i16.c: Regenerated. * generated/iall_i2.c: Regenerated. * generated/iall_i4.c: Regenerated. * generated/iall_i8.c: Regenerated. * generated/iany_i1.c: Regenerated. * generated/iany_i16.c: Regenerated. * generated/iany_i2.c: Regenerated. * generated/iany_i4.c: Regenerated. * generated/iany_i8.c: Regenerated. * generated/iparity_i1.c: Regenerated. * generated/iparity_i16.c: Regenerated. * generated/iparity_i2.c: Regenerated. * generated/iparity_i4.c: Regenerated. * generated/iparity_i8.c: Regenerated. * generated/norm2_r10.c: Regenerated. * generated/norm2_r16.c: Regenerated. * generated/norm2_r4.c: Regenerated. * generated/norm2_r8.c: Regenerated. * generated/parity_l1.c: Regenerated. * generated/parity_l16.c: Regenerated. * generated/parity_l2.c: Regenerated. * generated/parity_l4.c: Regenerated. * generated/parity_l8.c: Regenerated. * generated/misc_specifics.F90: Regenerated (comment change). * generated/_abs_c10.F90: Regenerated (comment change). * generated/_abs_c16.F90: Regenerated (comment change). * generated/_abs_c4.F90: Regenerated (comment change). * generated/_abs_c8.F90: Regenerated (comment change). * generated/_abs_i16.F90: Regenerated (comment change). * generated/_abs_i4.F90: Regenerated (comment change). * generated/_abs_i8.F90: Regenerated (comment change). * generated/_abs_r10.F90: Regenerated (comment change). * generated/_abs_r16.F90: Regenerated (comment change). * generated/_abs_r4.F90: Regenerated (comment change). * generated/_abs_r8.F90: Regenerated (comment change). * generated/_acosh_r10.F90: Regenerated (comment change). * generated/_acosh_r16.F90: Regenerated (comment change). * generated/_acosh_r4.F90: Regenerated (comment change). * generated/_acosh_r8.F90: Regenerated (comment change). * generated/_acos_r10.F90: Regenerated (comment change). * generated/_acos_r16.F90: Regenerated (comment change). * generated/_acos_r4.F90: Regenerated (comment change). * generated/_acos_r8.F90: Regenerated (comment change). * generated/_aimag_c10.F90: Regenerated (comment change). * generated/_aimag_c16.F90: Regenerated (comment change). * generated/_aimag_c4.F90: Regenerated (comment change). * generated/_aimag_c8.F90: Regenerated (comment change). * generated/_aint_r10.F90: Regenerated (comment change). * generated/_aint_r16.F90: Regenerated (comment change). * generated/_aint_r4.F90: Regenerated (comment change). * generated/_aint_r8.F90: Regenerated (comment change). * generated/_anint_r10.F90: Regenerated (comment change). * generated/_anint_r16.F90: Regenerated (comment change). * generated/_anint_r4.F90: Regenerated (comment change). * generated/_anint_r8.F90: Regenerated (comment change). * generated/_asinh_r10.F90: Regenerated (comment change). * generated/_asinh_r16.F90: Regenerated (comment change). * generated/_asinh_r4.F90: Regenerated (comment change). * generated/_asinh_r8.F90: Regenerated (comment change). * generated/_asin_r10.F90: Regenerated (comment change). * generated/_asin_r16.F90: Regenerated (comment change). * generated/_asin_r4.F90: Regenerated (comment change). * generated/_asin_r8.F90: Regenerated (comment change). * generated/_atan2_r10.F90: Regenerated (comment change). * generated/_atan2_r16.F90: Regenerated (comment change). * generated/_atan2_r4.F90: Regenerated (comment change). * generated/_atan2_r8.F90: Regenerated (comment change). * generated/_atanh_r10.F90: Regenerated (comment change). * generated/_atanh_r16.F90: Regenerated (comment change). * generated/_atanh_r4.F90: Regenerated (comment change). * generated/_atanh_r8.F90: Regenerated (comment change). * generated/_atan_r10.F90: Regenerated (comment change). * generated/_atan_r16.F90: Regenerated (comment change). * generated/_atan_r4.F90: Regenerated (comment change). * generated/_atan_r8.F90: Regenerated (comment change). * generated/_conjg_c10.F90: Regenerated (comment change
Re: [PATCH] fix -Werror=maybe-uninitialized in tree-ssa-loop-ivopts.c (PR bootstrap/50218)
Richard, Zdenek, On 08/28/2011 11:43 AM, Mikael Pettersson wrote: > The recent r178104 change to tree-ssa-loop-ivopts.c broke bootstrap on > i686-linux > because stage2 gcc considers the variable 'comp' used but possibly > uninitialized, > see PR50218. Initializing it to ERROR_MARK solves the problem, and is > consistent > with how other parts of r178104 represent absent/impossible 'comp' values. > > Ok for trunk? (I don't have svn write access.) > > /Mikael > > gcc/ > If you can ok this, I'll check it in. Thanks, - Tom > 2011-08-20 Mikael Pettersson > > PR bootstrap/50218 > * tree-ssa-loop-ivopts.c (determine_use_iv_cost_condition): Initialize > comp. > > --- gcc-4.7-20110827/gcc/tree-ssa-loop-ivopts.c.~1~ 2011-08-26 > 15:36:28.0 +0200 > +++ gcc-4.7-20110827/gcc/tree-ssa-loop-ivopts.c 2011-08-28 > 10:13:34.0 +0200 > @@ -4757,7 +4757,7 @@ determine_use_iv_cost_condition (struct >bool ok; >int elim_inv_expr_id = -1, express_inv_expr_id = -1, inv_expr_id; >tree *control_var, *bound_cst; > - enum tree_code comp; > + enum tree_code comp = ERROR_MARK; > >/* Only consider real candidates. */ >if (!cand->iv)
Re: [patch, libfortran] Fix PR 50192 - fix wide-char comparison
Regstrapped on x86_64-apple-darwin10 and powerpc-apple-darwin9 (see http://gcc.gnu.org/ml/gcc-testresults/2011-08/msg03125.html ) without regression. Thanks, Dominique
Re: [PATCH] fix -Werror=maybe-uninitialized in tree-ssa-loop-ivopts.c (PR bootstrap/50218)
On Sun, Aug 28, 2011 at 12:30:56PM +0200, Tom de Vries wrote: > > Ok for trunk? (I don't have svn write access.) > > > > /Mikael > > > > gcc/ > > > > If you can ok this, I'll check it in. This is fine. > > 2011-08-20 Mikael Pettersson > > > > PR bootstrap/50218 > > * tree-ssa-loop-ivopts.c (determine_use_iv_cost_condition): Initialize > > comp. > > > > --- gcc-4.7-20110827/gcc/tree-ssa-loop-ivopts.c.~1~ 2011-08-26 > > 15:36:28.0 +0200 > > +++ gcc-4.7-20110827/gcc/tree-ssa-loop-ivopts.c 2011-08-28 > > 10:13:34.0 +0200 > > @@ -4757,7 +4757,7 @@ determine_use_iv_cost_condition (struct > >bool ok; > >int elim_inv_expr_id = -1, express_inv_expr_id = -1, inv_expr_id; > >tree *control_var, *bound_cst; > > - enum tree_code comp; > > + enum tree_code comp = ERROR_MARK; > > > >/* Only consider real candidates. */ > >if (!cand->iv) Jakub
C++ PATCH for c++/48582 (core 342: null pointer values as template non-type arguments)
DR 342 allows null pointer values as non-type arguments in C++11; this
patch implements that. We mangle such arguments as a cast from 0 like
the EDG compiler does.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit c51119b8511eb6988e59c7decdd4c3817b9074bb
Author: Jason Merrill
Date: Fri Aug 26 22:38:12 2011 -0400
Core DR 342
PR c++/48582
* pt.c (check_valid_ptrmem_cst_expr): A null member pointer value
is valid in C++11.
(convert_nontype_argument): Likewise. Implicitly convert nullptr
and do constant folding.
* mangle.c (write_template_arg_literal): Mangle null member
pointer values as 0.
* call.c (null_member_pointer_value_p): New.
* cp-tree.h: Declare it.
diff --git a/gcc/cp/call.c b/gcc/cp/call.c
index dc35824..8421260 100644
--- a/gcc/cp/call.c
+++ b/gcc/cp/call.c
@@ -553,6 +553,23 @@ null_ptr_cst_p (tree t)
return false;
}
+/* Returns true iff T is a null member pointer value (4.11). */
+
+bool
+null_member_pointer_value_p (tree t)
+{
+ tree type = TREE_TYPE (t);
+ if (!type)
+return false;
+ else if (TYPE_PTRMEMFUNC_P (type))
+return (TREE_CODE (t) == CONSTRUCTOR
+ && integer_zerop (CONSTRUCTOR_ELT (t, 0)->value));
+ else if (TYPE_PTRMEM_P (type))
+return integer_all_onesp (t);
+ else
+return false;
+}
+
/* Returns nonzero if PARMLIST consists of only default parms,
ellipsis, and/or undeduced parameter packs. */
diff --git a/gcc/cp/cp-tree.h b/gcc/cp/cp-tree.h
index 8595943..d125642 100644
--- a/gcc/cp/cp-tree.h
+++ b/gcc/cp/cp-tree.h
@@ -4722,6 +4722,7 @@ extern tree build_addr_func (tree);
extern tree build_call_a (tree, int, tree*);
extern tree build_call_n (tree, int, ...);
extern bool null_ptr_cst_p (tree);
+extern bool null_member_pointer_value_p (tree);
extern bool sufficient_parms_p (const_tree);
extern tree type_decays_to (tree);
extern tree build_user_type_conversion (tree, tree, int);
diff --git a/gcc/cp/mangle.c b/gcc/cp/mangle.c
index 4c7cc79..1fcd999 100644
--- a/gcc/cp/mangle.c
+++ b/gcc/cp/mangle.c
@@ -2762,29 +2762,34 @@ write_template_arg_literal (const tree value)
write_char ('L');
write_type (TREE_TYPE (value));
- switch (TREE_CODE (value))
-{
-case CONST_DECL:
- write_integer_cst (DECL_INITIAL (value));
- break;
+ /* Write a null member pointer value as (type)0, regardless of its
+ real representation. */
+ if (null_member_pointer_value_p (value))
+write_integer_cst (integer_zero_node);
+ else
+switch (TREE_CODE (value))
+ {
+ case CONST_DECL:
+ write_integer_cst (DECL_INITIAL (value));
+ break;
-case INTEGER_CST:
- gcc_assert (!same_type_p (TREE_TYPE (value), boolean_type_node)
- || integer_zerop (value) || integer_onep (value));
- write_integer_cst (value);
- break;
+ case INTEGER_CST:
+ gcc_assert (!same_type_p (TREE_TYPE (value), boolean_type_node)
+ || integer_zerop (value) || integer_onep (value));
+ write_integer_cst (value);
+ break;
-case REAL_CST:
- write_real_cst (value);
- break;
+ case REAL_CST:
+ write_real_cst (value);
+ break;
-case STRING_CST:
- sorry ("string literal in function template signature");
- break;
+ case STRING_CST:
+ sorry ("string literal in function template signature");
+ break;
-default:
- gcc_unreachable ();
-}
+ default:
+ gcc_unreachable ();
+ }
write_char ('E');
}
@@ -2845,7 +2850,8 @@ write_template_arg (tree node)
/* A template appearing as a template arg is a template template arg. */
write_template_template_arg (node);
else if ((TREE_CODE_CLASS (code) == tcc_constant && code != PTRMEM_CST)
- || (abi_version_at_least (2) && code == CONST_DECL))
+ || (abi_version_at_least (2) && code == CONST_DECL)
+ || null_member_pointer_value_p (node))
write_template_arg_literal (node);
else if (DECL_P (node))
{
diff --git a/gcc/cp/pt.c b/gcc/cp/pt.c
index 3c6b2c5..1f43ff1 100644
--- a/gcc/cp/pt.c
+++ b/gcc/cp/pt.c
@@ -5240,6 +5240,8 @@ check_valid_ptrmem_cst_expr (tree type, tree expr,
STRIP_NOPS (expr);
if (expr && (null_ptr_cst_p (expr) || TREE_CODE (expr) == PTRMEM_CST))
return true;
+ if (cxx_dialect >= cxx0x && null_member_pointer_value_p (expr))
+return true;
if (complain & tf_error)
{
error ("%qE is not a valid template argument for type %qT",
@@ -5550,6 +5552,17 @@ convert_nontype_argument (tree type, tree expr, tsubst_flags_t complain)
else
expr = mark_rvalue_use (expr);
+ /* 14.3.2/5: The null pointer{,-to-member} conversion is applied
+ to a non-type argument of "nullptr". */
+ if (expr == nullptr_node
+ && (TYPE_PTR_P (type) || TYPE_PTR_TO_MEMBER_P (type)))
+expr = convert (type, expr);
+
+ /* In C++11, non-type template arguments can be arbitrary constant
+ expressions. But don't fold a PTRMEM_CST to a CONSTRUCTOR yet. */
+ if (cxx_dialect >= cxx0x && TREE_CODE
[patch libstdc++]: Add some missing errno-constants for mingw-targets
Hello, this patch adds some errno values supported by newer runtime-version. To keep backward compatibility this patch checks for their definition before trying to use them. Especially the EPERM error-number is required for having ISO-C++0x std::thread support enabled with winpthread libary for 32-bit and 64-bit Windows host. ChangeLog 2011-08-28 Kai Tietz * config/os/mingw32/error_constants.h (ECHILD): Add support, if present. (ENOSPC): Likewise. (EPERM): Likewise. Bootstrapped and regression tested for x86_64-w64-mingw32 and i686-w64-mingw32. Ok for apply? Regards, Kai Index: error_constants.h === --- error_constants.h (revision 178096) +++ error_constants.h (working copy) @@ -79,13 +79,17 @@ //network_reset = ENETRESET, //network_unreachable =ENETUNREACH, //no_buffer_space =ENOBUFS, -//no_child_process = ECHILD, +#ifdef ECHILD + no_child_process = ECHILD, +#endif //no_link =ENOLINK, no_lock_available = ENOLCK, //no_message_available = ENODATA, //no_message = ENOMSG, //no_protocol_option = ENOPROTOOPT, -//no_space_on_device = ENOSPC, +#ifdef ENOSPC + no_space_on_device = ENOSPC, +#endif //no_stream_resources =ENOSR, no_such_device_or_address = ENXIO, no_such_device = ENODEV, @@ -99,7 +103,9 @@ //not_supported = ENOTSUP, //operation_canceled = ECANCELED, //operation_in_progress = EINPROGRESS, -//operation_not_permitted =EPERM, +#ifdef EPERM + operation_not_permitted =EPERM, +#endif //operation_not_supported =EOPNOTSUPP, //operation_would_block = EWOULDBLOCK, //owner_dead = EOWNERDEAD,
Re: [4.7][google]Support for getting CPU type and feature information at run-time. (issue4893046)
On Aug 25, 2011, at 6:02 PM, H.J. Lu wrote: > On Thu, Aug 25, 2011 at 5:37 PM, Sriraman Tallam wrote: >> >> 1) Make __cpu_indicator_init a constructor in libgcc and guard to call >> it only once. > > This is unreliable and you don't need 3 symbols from libgcc. I'll add that once you start adding ctors, it is hard to impossible to ever get rid of them, and some people actually care about start up time and have static warnings for any code that has _any_ global constructors. By having it hidden behind an api, at least only that that want it, pay the price for it.
Re: [4.7][google]Support for getting CPU type and feature information at run-time. (issue4893046)
On Aug 26, 2011, at 10:58 AM, Xinliang David Li wrote: > Is there a standard way to force this init function to be called > before all ctors? Adding a ctor in one crtx.o ? Including the ctors that want to run before all other ctors? Think about that carefully. :-)
Re: [v3] Handle different versions of Solaris 8 ,
Hi again,
> FAIL: abi/header_cxxabi.c (test for excess errors)
> Excess errors:
> /var/gcc/regression/trunk/11-gcc/build/i386-pc-solaris2.11/libstdc++-v3/include/i386-pc-solaris2.11/bits/c++config.h:167:1:
> error: unknown type name 'namespace'
> /var/gcc/regression/trunk/11-gcc/build/i386-pc-solaris2.11/libstdc++-v3/include/i386-pc-solaris2.11/bits/c++config.h:168:1:
> error: expected '=', ',', ';', 'asm' or '__attribute__' before '{' token
>
> which is pretty obvious given that this test is supposed to be compiled as C
> :-)
In the meanwhile I had the time to catch up with the email and got your point
(for the accidental reader: being a C testcase *any* defined __cplusplus is
wrong)
Paolo
[PATCH] Fix ENTRY_VALUE parameter setup in var-tracking (PR debug/50215)
Hi! Depending on the computed hash of the parameter register, corresponding ENTRY_VALUE and the size of the hash table it can sometimes happen that val2 = cselib_lookup_from_insn will actually find val, which means that ENTRY_VALUEs in DEBUG_INSNs will be kept as ENTRY_VALUEs instead of REGs or similar preferrable locations if the value is still live there. Fixed by adding the ENTRY_VALUE to val->loc only after the cselib_lookup_from_insn. Bootstrapped/regtested on x86_64-linux and i686-linux, fixes PASS: gcc.dg/guality/pr45882.c -O0 line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -O1 (test for excess errors) PASS: gcc.dg/guality/pr45882.c -O1 execution test -UNSUPPORTED: gcc.dg/guality/pr45882.c -O1 line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O1 line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O1 line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -O1 line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -O1 line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -O2 (test for excess errors) PASS: gcc.dg/guality/pr45882.c -O2 execution test -UNSUPPORTED: gcc.dg/guality/pr45882.c -O2 line 16 e == 142 +PASS: gcc.dg/guality/pr45882.c -O2 line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O2 line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -O2 line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -O2 line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -O3 -fomit-frame-pointer (test for excess errors) PASS: gcc.dg/guality/pr45882.c -O3 -fomit-frame-pointer execution test +PASS: gcc.dg/guality/pr45882.c -O3 -fomit-frame-pointer line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O3 -fomit-frame-pointer line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -O3 -fomit-frame-pointer line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -O3 -fomit-frame-pointer line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -O3 -g (test for excess errors) PASS: gcc.dg/guality/pr45882.c -O3 -g execution test +PASS: gcc.dg/guality/pr45882.c -O3 -g line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O3 -g line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -O3 -g line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -O3 -g line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -Os (test for excess errors) PASS: gcc.dg/guality/pr45882.c -Os execution test +PASS: gcc.dg/guality/pr45882.c -Os line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -Os line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -Os line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -Os line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -O2 -flto -flto-partition=none (test for excess errors) PASS: gcc.dg/guality/pr45882.c -O2 -flto -flto-partition=none execution test +PASS: gcc.dg/guality/pr45882.c -O2 -flto -flto-partition=none line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O2 -flto -flto-partition=none line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -O2 -flto -flto-partition=none line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -O2 -flto -flto-partition=none line 16 e == 142 PASS: gcc.dg/guality/pr45882.c -O2 -flto (test for excess errors) PASS: gcc.dg/guality/pr45882.c -O2 -flto execution test +PASS: gcc.dg/guality/pr45882.c -O2 -flto line 16 b == 7 +PASS: gcc.dg/guality/pr45882.c -O2 -flto line 16 c == 11 +PASS: gcc.dg/guality/pr45882.c -O2 -flto line 16 d == 112 +PASS: gcc.dg/guality/pr45882.c -O2 -flto line 16 e == 142 which broke when SIMPLE_RETURN has been added to rtl.def and thus rtx codes of REG and ENTRY_VALUE changed and with that changed also the hashing of it. Ok for trunk? 2011-08-28 Jakub Jelinek PR debug/50215 * var-tracking.c (create_entry_value): Call cselib_lookup_from_insn before adding ENTRY_VALUE to val->locs. --- gcc/var-tracking.c.jj 2011-07-29 17:07:20.0 +0200 +++ gcc/var-tracking.c 2011-08-28 13:14:07.0 +0200 @@ -8488,13 +8488,13 @@ create_entry_value (rtx rtl, cselib_val cselib_val *val2; struct elt_loc_list *el; el = (struct elt_loc_list *) ggc_alloc_cleared_atomic (sizeof (*el)); - el->next = val->locs; el->loc = gen_rtx_ENTRY_VALUE (GET_MODE (rtl)); ENTRY_VALUE_EXP (el->loc) = rtl; - el->setting_insn = get_insns (); - val->locs = el; val2 = cselib_lookup_from_insn (el->loc, GET_MODE (rtl), true, VOIDmode, get_insns ()); + el->next = val->locs; + el->setting_insn = get_insns (); + val->locs = el; if (val2 && val2 != val && val2->locs Jakub
Re: [patch, libfortran] Fix PR 50192 - fix wide-char comparison
On 08/26/2011 05:40 AM, Thomas Koenig wrote: Hello world, the attached patch fixes the PR by doing comparisions for wide characters as unsigned 4-byte ints. I have put the comparison function into libgfortran.h because I will need it for MINLOC and friends for characters. OK for trunk? Which branches should this be backported to? Thomas OK, suggest if this problem produces wrong code, backport to at least 4.5, and 4.6, if applicable 4.4
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
Hi, Hello, this patch adds some errno values supported by newer runtime-version. To keep backward compatibility this patch checks for their definition before trying to use them. Any particular reason for not using [GLIBCXX_CHECK_SYSTEM_ERROR] for mingw too (by extending the list of checked syserr appropriately)? Paolo.
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
2011/8/28 Paolo Carlini : > Hi, >> >> Hello, >> >> this patch adds some errno values supported by newer runtime-version. >> To keep backward compatibility >> this patch checks for their definition before trying to use them. > > Any particular reason for not using [GLIBCXX_CHECK_SYSTEM_ERROR] for mingw > too (by extending the list of checked syserr appropriately)? > > Paolo. Well, I see here a possible clash for EOVERFLOW, which is defined for Windows crt to E2BIG. So we would have in list a double enumerator value for value_too_large and argument_list_too_long. Additionally we might want to include the socket-errors from WSA-equivalents. Well, not sure about the later, but at least I found this comment in header. Kai
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
On 08/28/2011 10:05 PM, Kai Tietz wrote: 2011/8/28 Paolo Carlini: Hi, Hello, this patch adds some errno values supported by newer runtime-version. To keep backward compatibility this patch checks for their definition before trying to use them. Any particular reason for not using [GLIBCXX_CHECK_SYSTEM_ERROR] for mingw too (by extending the list of checked syserr appropriately)? Paolo. Well, I see here a possible clash for EOVERFLOW, which is defined for Windows crt to E2BIG. So we would have in list a double enumerator value for value_too_large and argument_list_too_long. Additionally we might want to include the socket-errors from WSA-equivalents. Well, not sure about the later, but at least I found this comment in header. I don't understand: I only suggested using for mingw too the [GLIBCXX_CHECK_SYSTEM_ERROR] machinery (which is run *anyway* on mingw too), for the constants you are interested in, by adding those to the m4_foreach list. Paolo.
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
2011/8/28 Paolo Carlini : > On 08/28/2011 10:05 PM, Kai Tietz wrote: >> >> 2011/8/28 Paolo Carlini: >>> >>> Hi, Hello, this patch adds some errno values supported by newer runtime-version. To keep backward compatibility this patch checks for their definition before trying to use them. >>> >>> Any particular reason for not using [GLIBCXX_CHECK_SYSTEM_ERROR] for >>> mingw >>> too (by extending the list of checked syserr appropriately)? >>> >>> Paolo. >> >> Well, I see here a possible clash for EOVERFLOW, which is defined for >> Windows crt to E2BIG. So we would have in list a double enumerator >> value for value_too_large and argument_list_too_long. Additionally we >> might want to include the socket-errors from WSA-equivalents. Well, >> not sure about the later, but at least I found this comment in header. > > I don't understand: I only suggested using for mingw too the > [GLIBCXX_CHECK_SYSTEM_ERROR] machinery (which is run *anyway* on mingw too), > for the constants you are interested in, by adding those to the m4_foreach > list. > > Paolo. Ah, now I got you. Hmm, well, I see here no advantage in adding another macro to check for an macro. As in code of libstdc++ there is actual no use for those new ones. But if you insist on that, I can add them and replace #if guards by #if HAVE_ style. Nevertheless, for some error-numbers (eg. EAFNOSUPPORT, EBADMSG, etc) - which are socket API related - there is no constant define in errno.h (and never will be). So code to emulate this would be still a manual thing in error_constants.h header AFAICS (if we plan to support them for mingw targets). Kai
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
Hi, Ah, now I got you. Hmm, well, I see here no advantage in adding another macro to check for an macro. As in code of libstdc++ there is actual no use for those new ones. To be clear, I didn't personally invent those, but I'm still finding a bit strange doing only for mingw something completely different, considering in particular that the autoconf procedure is run *anyway* on mingw too... Nevertheless, for some error-numbers (eg. EAFNOSUPPORT, EBADMSG, etc) - which are socket API related - there is no constant define in errno.h (and never will be). So code to emulate this would be still a manual thing in error_constants.h header AFAICS (if we plan to support them for mingw targets). Once more, I don't understand: didn't you suggest changing *only* ENOSPC and EPERM? Why are we now discussing other error-numbers like ECHILD, EAFNOSUPPORT and EBADMSG? Are the latter defined in errno.h? In that case, please use [GLIBCXX_CHECK_SYSTEM_ERROR] for now. We can always simplify things at a later time. Paolo.
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
Sorry,: Once more, I don't understand: didn't you suggest changing *only* ENOSPC and EPERM? Why are we now discussing other error-numbers like ECHILD, EAFNOSUPPORT and EBADMSG? Are the latter defined in errno.h? should have written: Once more, I don't understand: didn't you suggest changing *only* ECHILD, ENOSPC and EPERM? Why are we now discussing other error-numbers like EAFNOSUPPORT and EBADMSG? Are the former defined in errno.h? Paolo.
[v3] Remove wa for c++/49045 in std::swap
Hi,
tested x86_64-linux, committed to mainline.
Paolo.
//
2011-08-28 Paolo Carlini
* include/bits/move.h (swap(_Tp(&)[_Nm], _Tp(&)[_Nm])): Remove
workaround for c++/49045.
* include/bits/algorithmfwd.h: Likewise.
Index: include/bits/move.h
===
--- include/bits/move.h (revision 178146)
+++ include/bits/move.h (working copy)
@@ -152,21 +152,13 @@
__b = _GLIBCXX_MOVE(__tmp);
}
-#ifdef __GXX_EXPERIMENTAL_CXX0X__
- // To work around c++/49045.
- template
-struct __is_nothrow_swappable
-{ static const bool value = noexcept(swap(std::declval<_Tp&>(),
- std::declval<_Tp&>())); };
-#endif
-
// _GLIBCXX_RESOLVE_LIB_DEFECTS
// DR 809. std::swap should be overloaded for array types.
template
inline void
swap(_Tp (&__a)[_Nm], _Tp (&__b)[_Nm])
#ifdef __GXX_EXPERIMENTAL_CXX0X__
-noexcept(__is_nothrow_swappable<_Tp>::value)
+noexcept(noexcept(swap(*__a, *__b)))
#endif
{
for (size_t __n = 0; __n < _Nm; ++__n)
Index: include/bits/algorithmfwd.h
===
--- include/bits/algorithmfwd.h (revision 178146)
+++ include/bits/algorithmfwd.h (working copy)
@@ -558,9 +558,9 @@
template
void
-swap(_Tp (&)[_Nm], _Tp (&)[_Nm])
+swap(_Tp (&__a)[_Nm], _Tp (&__b)[_Nm])
#ifdef __GXX_EXPERIMENTAL_CXX0X__
-noexcept(__is_nothrow_swappable<_Tp>::value)
+noexcept(noexcept(swap(*__a, *__b)))
#endif
;
sh.opt typo fix.
Hi all I found a typo in sh.opt and fix it. --Liu From 16010419db714eae298f8b7c4726eb75383bed78 Mon Sep 17 00:00:00 2001 From: Jia Liu Date: Mon, 29 Aug 2011 05:25:19 +0800 Subject: [PATCH] sh_dot_opt big endian typo fix --- gcc/ChangeLog|4 gcc/config/sh/sh.opt |2 +- 2 files changed, 5 insertions(+), 1 deletions(-) diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 56397af..0996e08 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,3 +1,7 @@ +2011-08-29 Jia Liu + + * config/sh/sh.opt (mb): Typo fix. + 2011-08-28 Mikael Pettersson PR bootstrap/50218 diff --git a/gcc/config/sh/sh.opt b/gcc/config/sh/sh.opt index e94f53a..9471596 100644 --- a/gcc/config/sh/sh.opt +++ b/gcc/config/sh/sh.opt @@ -210,7 +210,7 @@ Target Report Mask(ADJUST_UNROLL) Condition(SUPPORT_ANY_SH5) Throttle unrolling to avoid thrashing target registers unless the unroll benefit outweighs this mb -Target Report RejectNegative InverseMask(LITTLE_ENDIAN) +Target Report RejectNegative InverseMask(BIG_ENDIAN) Generate code in big endian mode mbigtable -- 1.7.4.1
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
2011/8/28 Paolo Carlini :
> Sorry,:
>>
>> Once more, I don't understand: didn't you suggest changing *only* ENOSPC
>> and EPERM? Why are we now discussing other error-numbers like ECHILD,
>> EAFNOSUPPORT and EBADMSG? Are the latter defined in errno.h?
>
> should have written:
>
> Once more, I don't understand: didn't you suggest changing *only* ECHILD,
> ENOSPC and EPERM? Why are we now discussing other error-numbers like
> EAFNOSUPPORT and EBADMSG? Are the former defined in errno.h?
>
> Paolo.
Hi,
So I altered the patch in the way you were suggesting. I verfied
values supported in more
detail and found that some other error-values should be handled for
mingw host, too.
The feared clash for E2BIG and EOVERFLOW doesn't happen, so I enabled
them, too. Additionally
in combination with winpthread, there are error-values ETIMEDOUT,
ENOTSUP, and EWOULDBLOCK present.
ChangeLog
2011-08-28 Kai Tietz
* configure: Regenerated.
* config.h.in: Regenerated.
* acinclude.m4 (GLIBCXX_CHECK_SYSTEM_ERROR) Add
checks for ECHILD, ENOSPC, EPERM, ETIMEDOUT, and
EWOULDBLOCK.
* config/os/mingw32/error_constants.h (errc): Add
enumerator values no_child_process, no_space_on_device,
not_supported, operation_not_permitted, operation_would_block,
timed_out, and value_too_large,
Regression tested for x86_64-w64-mingw32 and i686-w64-mingw32. Ok for apply?
Regards,
Kai
Index: configure
===
--- configure (revision 178145)
+++ configure (working copy)
@@ -17650,9 +17650,174 @@
$as_echo "#define HAVE_ETXTBSY 1" >>confdefs.h
fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for ECHILD" >&5
+$as_echo_n "checking for ECHILD... " >&6; }
+if test "${glibcxx_cv_system_error15+set}" = set; then :
+ $as_echo_n "(cached) " >&6
+else
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h. */
+#include
+int
+main ()
+{
+int i = ECHILD;
+ ;
+ return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+ glibcxx_cv_system_error15=yes
+else
+ glibcxx_cv_system_error15=no
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result:
$glibcxx_cv_system_error15" >&5
+$as_echo "$glibcxx_cv_system_error15" >&6; }
+if test x"$glibcxx_cv_system_error15" = x"yes"; then
+
+$as_echo "#define HAVE_ECHILD 1" >>confdefs.h
+
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for ENOSPC" >&5
+$as_echo_n "checking for ENOSPC... " >&6; }
+if test "${glibcxx_cv_system_error16+set}" = set; then :
+ $as_echo_n "(cached) " >&6
+else
+
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h. */
+#include
+int
+main ()
+{
+int i = ENOSPC;
+ ;
+ return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+ glibcxx_cv_system_error16=yes
+else
+ glibcxx_cv_system_error16=no
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+fi
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result:
$glibcxx_cv_system_error16" >&5
+$as_echo "$glibcxx_cv_system_error16" >&6; }
+if test x"$glibcxx_cv_system_error16" = x"yes"; then
+
+$as_echo "#define HAVE_ENOSPC 1" >>confdefs.h
+
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for EPERM" >&5
+$as_echo_n "checking for EPERM... " >&6; }
+if test "${glibcxx_cv_system_error17+set}" = set; then :
+ $as_echo_n "(cached) " >&6
+else
+
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h. */
+#include
+int
+main ()
+{
+int i = EPERM;
+ ;
+ return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+ glibcxx_cv_system_error17=yes
+else
+ glibcxx_cv_system_error17=no
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+fi
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result:
$glibcxx_cv_system_error17" >&5
+$as_echo "$glibcxx_cv_system_error17" >&6; }
+if test x"$glibcxx_cv_system_error17" = x"yes"; then
+
+$as_echo "#define HAVE_EPERM 1" >>confdefs.h
+
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for ETIMEDOUT" >&5
+$as_echo_n "checking for ETIMEDOUT... " >&6; }
+if test "${glibcxx_cv_system_error18+set}" = set; then :
+ $as_echo_n "(cached) " >&6
+else
+
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h. */
+#include
+int
+main ()
+{
+int i = ETIMEDOUT;
+ ;
+ return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+ glibcxx_cv_system_error18=yes
+else
+ glibcxx_cv_system_error18=no
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+fi
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result:
$glibcxx_cv_system_error18" >&5
+$as_echo "$glibcxx_cv_system_error18" >&6; }
+if test x"$glibcxx_cv_system_error18" = x"yes"; then
+
+$as_echo "#define HAVE_ETIMEDOUT 1" >>confdefs.h
+
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for EWOULDBLOCK" >&5
+$as_echo_n "checking for EWOULDBLOCK... " >&6; }
+if test "${glibcxx_cv_system_error19+set}" = set; then :
+ $as_echo_n "(cached) " >&6
+else
+
+
Re: [PATCH, PR fortran/45170] -- properly translates substring reference
On Saturday 27 August 2011 02:34:24 Steve Kargl wrote: > The attached patch allows gfortran to compile the > attached testcase. The resulting executable runs > as expected. > > Short story: > > character(len=20) :: string = 'some text here' > character(len=:), allocatable :: s > n = 5 > allocate(s, source=string(:n)) > > The length of s is determined from the expression in > the source= argument. If this expression is a lonely > substring reference as in the above, then gfortran > does set the correct length. This patch fixes this. ... does *not* set the correct length, I suppose? > OK for trunk? It is OK. Rather obvious. Thanks Mikael
Re: [patch libstdc++]: Add some missing errno-constants for mingw-targets
Hi, Once more, I don't understand: didn't you suggest changing *only* ENOSPC and EPERM? Why are we now discussing other error-numbers like ECHILD, EAFNOSUPPORT and EBADMSG? Are the latter defined in errno.h? should have written: Once more, I don't understand: didn't you suggest changing *only* ECHILD, ENOSPC and EPERM? Why are we now discussing other error-numbers like EAFNOSUPPORT and EBADMSG? Are the former defined in errno.h? Paolo. Hi, So I altered the patch in the way you were suggesting. I verfied values supported in more detail and found that some other error-values should be handled for mingw host, too. The feared clash for E2BIG and EOVERFLOW doesn't happen, so I enabled them, too. Additionally in combination with winpthread, there are error-values ETIMEDOUT, ENOTSUP, and EWOULDBLOCK present. Good. ChangeLog 2011-08-28 Kai Tietz * configure: Regenerated. * config.h.in: Regenerated. * acinclude.m4 (GLIBCXX_CHECK_SYSTEM_ERROR) Add checks for ECHILD, ENOSPC, EPERM, ETIMEDOUT, and EWOULDBLOCK. * config/os/mingw32/error_constants.h (errc): Add enumerator values no_child_process, no_space_on_device, not_supported, operation_not_permitted, operation_would_block, timed_out, and value_too_large, Regression tested for x86_64-w64-mingw32 and i686-w64-mingw32. Ok for apply? Ok with the patch tweaked to prefer uglified _GLIBCXX_HAVE_* names in error_constants.h. Let's just be consistent with the generic error_constants.h about this normal library policy too. Paolo. PS: please don't post regenerated files as part of the patch, are simply a distraction.
Re: [Fortran, Patch] Coarray: libcaf patch for _gfortran_caf_deregister
On Friday 26 August 2011 08:51:37 Tobias Burnus wrote:
> Allocatable coarrays are freed and deregistered via the libcaf function
> _gfortran_caf_deregister. Currently, the front end does not generate
> calls to the that function, however, this patch already implements the
> function.
>
> See http://gcc.gnu.org/wiki/CoarrayLib and
> http://gcc.gnu.org/ml/fortran/2010-04/msg00168.html for details.
>
> The function is called with the coarray token as argument. The token
> identifies the coarray in a way defined by the library. In case of
> single.c, it just contains the address of the allocated memory of the
> coarray. In case of mpi.c, it is an array of memory addresses on all
> images such that token[this_image()-1] is the memory location of the
> current image.
>
> The patch also adds stat= and errmsg= diagnostic.
>
> TODO: Adding calls to the function in code generated by the compiler -
> and testing the function.
>
>
> Tested by compiling with mpicc and gcc with "-Wall -Wextra -std=c99".
> OK for the trunk?
>
> Tobias
diff --git a/libgfortran/caf/mpi.c b/libgfortran/caf/mpi.c
index ea4c0f0..711c6ee 100644
--- a/libgfortran/caf/mpi.c
+++ b/libgfortran/caf/mpi.c
@@ -103,7 +103,7 @@ _gfortran_caf_finalize (void)
{
while (caf_static_list != NULL)
{
- free(caf_static_list->token[caf_this_image-1]);
+ free (caf_static_list->token[caf_this_image-1]);
caf_static_list = caf_static_list->prev;
}
Not something introduced by this patch, but I would like to point that
caf_static_list should be freed too.
@@ -187,10 +187,36 @@ error:
}
-int
-_gfortran_caf_deregister (void **token __attribute__ ((unused)))
+void
+_gfortran_caf_deregister (void **token, int *stat, char *errmsg, int
errmsg_len)
{
- return 0;
+ if (unlikely (caf_is_finalized))
+{
+ const char msg[] = "Failed to deallocate coarray - "
+ "there are stopped images";
+ if (stat)
+ {
+ *stat = STAT_STOPPED_IMAGE;
+
+ if (errmsg_len > 0)
+ {
+ int len = ((int) sizeof (msg) - 1 > errmsg_len)
+ ? errmsg_len : (int) sizeof (msg) - 1;
+ memcpy (errmsg, msg, len);
+ if (errmsg_len > len)
+ memset (&errmsg[len], ' ', errmsg_len-len);
+ }
+ return;
+ }
+ caf_runtime_error (msg);
+}
+
+ _gfortran_caf_sync_all (NULL, NULL, 0);
+
+ if (stat)
+*stat = 0;
+
+ free (token[caf_this_image-1]);
}
To be consistent with _gfortran_caf_register, I think you should be freeing
token itself (the whole array).
OK with that change.
Mikael
Re: sh.opt typo fix.
Liu wrote: > mb > -Target Report RejectNegative InverseMask(LITTLE_ENDIAN) ^^^ You missed the "Inverse" prefix, didn't you? Regards, kaz
Re: sh.opt typo fix.
On Mon, Aug 29, 2011 at 7:42 AM, Kaz Kojima wrote: > Liu wrote: >> mb >> -Target Report RejectNegative InverseMask(LITTLE_ENDIAN) > ^^^ > You missed the "Inverse" prefix, didn't you? > > Regards, > kaz > Thanks, It looks like I missed "Inverse" prefix. Regards, -Liu
Re: [4.7][google]Support for getting CPU type and feature information at run-time. (issue4893046)
Sri, please add a new api to do cpu_indicator initialization on demand
to be used in IFUNC context. Perhaps also add some debug check to make
sure no conflicting cpu model is set.
Ok for google branches for now while the discussion continues.
thanks,
David
On Thu, Aug 25, 2011 at 5:37 PM, Sriraman Tallam wrote:
> Hi,
>
> Thanks for all the comments. I am attaching a new patch
> incorporating all of the changes mentioned, mainly :
>
> 1) Make __cpu_indicator_init a constructor in libgcc and guard to call
> it only once.
> 2) Add symbol versions.
> 3) Move all builtins to the i386 port.
> 4) Add check for atom processor.
> 5) No separate passes to fold the builtins.
>
> Please let me know what you think.
> Thanks,
> -Sri.
>
> * config/i386/i386.c (build_struct_with_one_bit_fields): New function.
> (make_var_decl): New function.
> (get_field_from_struct): New function.
> (fold_builtin_target): New function.
> (ix86_fold_builtin): New function.
> (ix86_expand_builtin): Expand new builtins by folding them.
> (TARGET_FOLD_BUILTIN): New macro.
> (IX86_BUILTIN_CPU_SUPPORTS_CMOV): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_MMX): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_POPCOUNT): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_SSE): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_SSE2): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_SSE3): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_SSSE3): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_SSE4_1): New enum value.
> (IX86_BUILTIN_CPU_SUPPORTS_SSE4_2): New enum value.
> (IX86_BUILTIN_CPU_IS_AMD): New enum value.
> (IX86_BUILTIN_CPU_IS_INTEL): New enum value.
> (IX86_BUILTIN_CPU_IS_INTEL_ATOM): New enum value.
> (IX86_BUILTIN_CPU_IS_INTEL_CORE2): New enum value.
> (IX86_BUILTIN_CPU_IS_INTEL_COREI7_NEHALEM): New enum value.
> (IX86_BUILTIN_CPU_IS_INTEL_COREI7_WESTMERE): New enum value.
> (IX86_BUILTIN_CPU_IS_INTEL_COREI7_SANDYBRIDGE): New enum value.
> (IX86_BUILTIN_CPU_IS_AMDFAM10_BARCELONA): New enum value.
> (IX86_BUILTIN_CPU_IS_AMDFAM10_SHANGHAI): New enum value.
> (IX86_BUILTIN_CPU_IS_AMDFAM10_ISTANBUL): New enum value.
> * config/i386/libgcc-glibc.ver (__cpu_indicator_init): Export symbol.
> (__cpu_model): Export symbol.
> (__cpu_features): Export symbol.
> * config/i386/i386-builtin-types.def: New function type.
>
> * config/i386/i386-cpuinfo.c: New file.
> * config/i386/t-cpuinfo: New file.
> * config.host: Add t-cpuinfo to link i386-cpuinfo.o with libgcc
>
> * gcc.dg/builtin_target.c: New test.
>
>
>
>
> On Tue, Aug 23, 2011 at 4:35 AM, Michael Matz wrote:
>> Hi,
>>
>> On Mon, 22 Aug 2011, H.J. Lu wrote:
>>
>>> > void __attribute__((constructor)) bla(void)
>>> > {
>>> > __cpu_indicator_init ();
>>> > }
>>> >
>>> > I don't see any complication.?
>>> >
>>>
>>> Order of constructors. A constructor may call functions
>>> which use __cpu_indicator.
>>
>> That's why I wrote also:
>>
>>> The initializer function has to be callable from pre-.init contexts, e.g.
>>> ifunc dispatchers.
>>
>> It obviously has to be guarded against multiple calls. The ctor in libgcc
>> would be mere convenience because then non-ctor code can rely on the data
>> being initialized, and only (potential) ctor code has to check and call
>> the init function on demand.
>>
>>
>> Ciao,
>> Michael.
>
Re: [PATCH][ARM] -m{cpu,tune,arch}=native
On Sat, Aug 27, 2011 at 3:19 AM, Andrew Stubbs wrote: > Hi all, > > This patch adds support for -mcpu=native, -mtune=native, and -march=native > for ARM Linux hosts. > > So far, it only recognises Cortex-A8 and Cortex-A9, so I really need to find > out what the magic part numbers are for other cpus before this patch is > complete. I couldn't just find this information listed anywhere. I think > there are a lot of clues in the kernel code, but it's hard to mine and it > mostly only goes as far the architecture version, not the individual cpu. Could you send linaro-dev@ an email and ask them where this API is documented? Hmm. Should there be a -mfpu=native that picks between soft, VFP, and NEON based on the host? -- Michael
