NEW GCC build failure, HEAD@146132 on native

[regress at apple dot com (regress)]

With your recent patch, GCC HEAD revision 146132 had problems on:
native: build (NEW build failure)
Attached is build output for those targets.
The previous build was of revision 146101.

Log information for changes since the last build:
------------------------------------------------------------------------
r146102 | charlet | 2009-04-15 05:09:35 -0700 (Wed, 15 Apr 2009) | 53 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/a-caldel-vms.adb
   M /trunk/gcc/ada/adaint.h
   M /trunk/gcc/ada/checks.adb
   M /trunk/gcc/ada/exp_ch3.adb
   M /trunk/gcc/ada/exp_ch6.adb
   M /trunk/gcc/ada/exp_disp.adb
   M /trunk/gcc/ada/exp_disp.ads
   M /trunk/gcc/ada/g-comlin.adb
   M /trunk/gcc/ada/opt.ads
   M /trunk/gcc/ada/prj-nmsc.adb
   M /trunk/gcc/ada/rtsfind.adb
   M /trunk/gcc/ada/sem_disp.adb
   M /trunk/gcc/ada/sinfo.ads

2009-04-15  Pascal Obry  <obry@adacore.com>

	* adaint.h (__gnat_unlink): Add spec.
	(__gnat_rename): Likewise.

2009-04-15  Vincent Celier  <celier@adacore.com>

	* prj-nmsc.adb: Minor spelling error corrections in error messages

2009-04-15  Robert Dewar  <dewar@adacore.com>

	* sinfo.ads: Minor comment update

	* opt.ads: Minor comment updates

	* checks.adb (Enable_Overflow_Check): Do not set Do_Overflow_Check for
	modular type.

2009-04-15  Ed Schonberg  <schonberg@adacore.com>

	* exp_disp.ads, exp_disp.adb (Register_Primitive): Is now a function
	that generates the code needed to update a dispatch table when a
	primitive operation is declared with a subprogram body without previous
	spec. Insertion of the generated code is responsibility of the caller.
	(Make_DT): When building static tables, append the code created by
	Register_Primitive to update a secondary table after it has been
	constructed.

	* exp_ch3.adb, exp_ch6.adb: use new version of Register_Primitive.

	* sem_disp.adb (Check_Dispatching_Operation): Call Register_Primitive
	on an overriding operation that implements an interface operation only
	if not building static dispatch tables.

2009-04-15  Hristian Kirtchev  <kirtchev@adacore.com>

	* a-caldel-vms.adb (To_Duration): Declare a "safe" end of time which
	does not cause overflow when converted to Duration. Use the safe value
	as the maximum allowable time delay..

2009-04-15  Jerome Lambourg  <lambourg@adacore.com>

	* g-comlin.adb (Set_Command_Line): When adding a switch with attached
	parameter, specify that the delimiter is NUL, otherwise "-j2" will be
	translated to "-j 2".

2009-04-15  Bob Duff  <duff@adacore.com>

	* rtsfind.adb (Maybe_Add_With): Split out procedure to add implicit
	with_clauses, to avoid code duplication. Change this processing so we
	always add a with_clause on the main unit if needed.


------------------------------------------------------------------------
r146103 | charlet | 2009-04-15 05:14:57 -0700 (Wed, 15 Apr 2009) | 21 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/prj-part.adb
   M /trunk/gcc/ada/prj-tree.adb
   M /trunk/gcc/ada/prj-tree.ads
   M /trunk/gcc/ada/rtsfind.adb
   M /trunk/gcc/ada/sem_ch3.adb

2009-04-15  Robert Dewar  <dewar@adacore.com>

	* rtsfind.adb: Minor reformatting.

2009-04-15  Emmanuel Briot  <briot@adacore.com>

	* prj-part.adb, prj-tree.adb, prj-tree.ads (Restore_And_Free): renames
	Restore, and free the saved context.

2009-04-15  Gary Dismukes  <dismukes@adacore.com>

	* sem_ch3.adb (Analyze_Private_Extension_Declaration): Move error check
	for illegal private extension from a synchronized interface parent in
	front of check for illegal limited extension so that limited extension
	from a synchronized interface will be rejected.
	(Check_Ifaces): Check that a private extension that has a synchronized
	interface as a progenitor must be explicitly declared synchronized.
	Also check that a record extension cannot derive from a synchronized
	interface.


------------------------------------------------------------------------
r146104 | charlet | 2009-04-15 05:18:14 -0700 (Wed, 15 Apr 2009) | 15 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/exp_ch6.adb
   M /trunk/gcc/ada/sem_eval.adb
   M /trunk/gcc/ada/targparm.adb

2009-04-15  Robert Dewar  <dewar@adacore.com>

	* sem_eval.adb (Get_Static_Length): Go to origin node for array bounds
	in case they were rewritten by expander (Force_Evaluation).

	* targparm.adb (Get_Target_Parameters): Correct check for
	Suppress_Exception_Locations.

2009-04-15  Ed Schonberg  <schonberg@adacore.com>

	* exp_ch6.adb (Expand_Inlined_Call): If an in-parameter in a call to be
	inlined  is of an array type that is not bit-packed, use a renaming
	declaration to capture its value, rather than a constant declaration.


------------------------------------------------------------------------
r146105 | charlet | 2009-04-15 05:21:57 -0700 (Wed, 15 Apr 2009) | 13 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/exp_ch6.adb
   M /trunk/gcc/ada/gnatchop.adb
   M /trunk/gcc/ada/sem_ch13.adb

2009-04-15  Robert Dewar  <dewar@adacore.com>

	* sem_ch13.adb (Unchecked_Conversions): Store source location instead
	of node for location for warning messages.

	* gnatchop.adb: Minor reformatting

2009-04-15  Ed Schonberg  <schonberg@adacore.com>

	* exp_ch6.adb: additional guard for renaming declarations for in
	parameters of an array type.


------------------------------------------------------------------------
r146106 | ebotcazou | 2009-04-15 05:25:31 -0700 (Wed, 15 Apr 2009) | 1 line
Changed paths:
   M /trunk/gcc/ChangeLog

Fix formatting
------------------------------------------------------------------------
r146107 | charlet | 2009-04-15 05:42:16 -0700 (Wed, 15 Apr 2009) | 4 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/gcc-interface/Make-lang.in
   M /trunk/gcc/ada/gcc-interface/Makefile.in

	* gcc-interface/Make-lang.in: Update dependencies.
	* gcc-interface/Makefile.in: Reorganization of s-osinte-vxworks*
	and s-vxwext*.

------------------------------------------------------------------------
r146108 | charlet | 2009-04-15 05:43:58 -0700 (Wed, 15 Apr 2009) | 33 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/exp_ch7.adb
   M /trunk/gcc/ada/gnat_ugn.texi
   M /trunk/gcc/ada/gnatchop.adb
   M /trunk/gcc/ada/par-ch3.adb
   D /trunk/gcc/ada/s-osinte-vxworks-kernel.adb
   M /trunk/gcc/ada/s-osinte-vxworks.adb
   M /trunk/gcc/ada/s-osinte-vxworks.ads
   A /trunk/gcc/ada/s-vxwext-kernel.adb
   M /trunk/gcc/ada/s-vxwext-kernel.ads
   M /trunk/gcc/ada/s-vxwext.ads
   M /trunk/gcc/ada/system-darwin-x86_64.ads
   M /trunk/gcc/ada/system-mingw-x86_64.ads

2009-04-15  Robert Dewar  <dewar@adacore.com>

	* gnatchop.adb (BOM_Length): New global variable
	(Write_Unit): Add new parameter Write_BOM
	(Write_Chopped_Files): Check for BOM and set Write_BOM for call
	to Write_Unit

	* gnat_ugn.texi: Add note on propagation of BOM by gnatchop

2009-04-15  Geert Bosch  <bosch@adacore.com>

	* system-mingw-x86_64.ads, system-darwin-x86_64.ads
	(Backend_Overflow_Checks): Set to True.

2009-04-15  Gary Dismukes  <dismukes@adacore.com>

	* par-ch3.adb (P_Type_Declaration): Issue an error if the synchronized
	keyword is given in a record extension.

2009-04-15  Hristian Kirtchev  <kirtchev@adacore.com>

	* exp_ch7.adb (Expand_Ctrl_Function_Call): Procede with the expansion
	of a controlled function call in the context of a record aggregate.
	This does not apply to array aggregates since the call will be expanded
	into assignments.

2009-04-15  Ed Falis  <falis@adacore.com>

	* s-osinte-vxworks-kernel.adb, s-osinte-vxworks.adb,
	s-osinte-vxworks.ads s-vxwext.ads, s-vxwext-kernel.adb,
	s-vxwext-kernel.ads: Reorganize s-osinte-vxworks* and s-vxwext*.


------------------------------------------------------------------------
r146109 | charlet | 2009-04-15 05:45:27 -0700 (Wed, 15 Apr 2009) | 2 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/s-vxwext-rtp.adb
   M /trunk/gcc/ada/s-vxwext-rtp.ads

Reorganize code.

------------------------------------------------------------------------
r146110 | charlet | 2009-04-15 05:48:57 -0700 (Wed, 15 Apr 2009) | 16 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/exp_ch7.adb
   M /trunk/gcc/ada/frontend.adb
   M /trunk/gcc/ada/sem_ch8.adb

2009-04-15  Hristian Kirtchev  <kirtchev@adacore.com>

	* exp_ch7.adb (Expand_Ctrl_Function_Call): Check for the case where the
	immediate parent of the controlled function call is a component
	association.

2009-04-15  Ed Schonberg  <schonberg@adacore.com>

	* sem_ch8.adb (Use_One_Type): If the type is tagged, indicate that the
	corresponding class-wide type is also in use.

2009-04-15  Thomas Quinot  <quinot@adacore.com>

	* frontend.adb: Minor comment fix


------------------------------------------------------------------------
r146111 | charlet | 2009-04-15 05:52:48 -0700 (Wed, 15 Apr 2009) | 19 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/gnat_rm.texi
   M /trunk/gcc/ada/sem_ch5.adb
   M /trunk/gcc/ada/sem_ch6.adb
   M /trunk/gcc/ada/sem_warn.adb

2009-04-15  Bob Duff  <duff@adacore.com>

	* sem_ch5.adb (Analyze_Loop_Statement): Don't check for infinite loop
	warnings unless the loop comes from source, because checking generated
	loops is a waste of time, and makes it harder to debug
	Check_Infinite_Loop_Warning.

	* sem_warn.adb (Check_Infinite_Loop_Warning): If the local variable
	tested in the while loop is a renaming, do not warn. Otherwise, we get
	false alarms, because it's usually renaming something that we can't
	deal with (an indexed component, a global variable, ...).

	* gnat_rm.texi: Fix typo

2009-04-15  Thomas Quinot  <quinot@adacore.com>

	* sem_ch6.adb: Minor reformatting


------------------------------------------------------------------------
r146112 | charlet | 2009-04-15 05:57:34 -0700 (Wed, 15 Apr 2009) | 31 lines
Changed paths:
   M /trunk/gcc/ada/ChangeLog
   M /trunk/gcc/ada/exp_ch9.adb
   M /trunk/gcc/ada/sem_ch12.adb
   M /trunk/gcc/ada/sem_util.adb
   M /trunk/gcc/ada/sem_warn.adb
   M /trunk/gcc/ada/sinput-p.adb
   M /trunk/gcc/ada/sinput-p.ads

2009-04-15  Hristian Kirtchev  <kirtchev@adacore.com>

	* exp_ch9.adb: Comment improvements.
	(Build_Entry_Family_Name): Add parentheses around the index of a entry
	family member.

2009-04-15  Bob Duff  <duff@adacore.com>

	* sem_warn.adb (Check_Infinite_Loop_Warning): Catch cases like
	"while X /= null loop" where X is unchanged inside the loop. We were
	not warning in this case, because of the pointers -- we feared that the
	loop variable could be updated via a pointer, if there are any pointers
	around the place. But that is impossible in this case.

	* sem_util.adb (May_Be_Lvalue): This routine was overly pessimistic in
	the case of dereferences. In X.all, X cannot be an l-value. We now
	catch that case (and implicit dereferences, too).

2009-04-15  Vincent Celier  <celier@adacore.com>

	* sinput-p.ads, sinput-p.adb (Clear_Source_File_Table): New procedure

2009-04-15  Ed Schonberg  <schonberg@adacore.com>

	* sem_ch12.adb (Is_Actual_Of_Previous_Formal): Make fully recursive.
	From code reading.
	(Analyze_Package_Instantiation): If generic unit in child instance is
	the same as generic unit in parent instance, look for an outer homonym
	to locate the desired generic.


------------------------------------------------------------------------
r146113 | hubicka | 2009-04-15 06:01:20 -0700 (Wed, 15 Apr 2009) | 2 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/builtins.def

	* builtins.def (va_start, va_end, va_copy): Mark nothrow.

------------------------------------------------------------------------
r146117 | hubicka | 2009-04-15 07:07:55 -0700 (Wed, 15 Apr 2009) | 3 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/builtins.def

	* builtins.def (va_start, va_end, va_copy): Fix my previous commit.
	Wrong version of patch.

------------------------------------------------------------------------
r146120 | rguenth | 2009-04-15 07:17:35 -0700 (Wed, 15 Apr 2009) | 8 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/testsuite/ChangeLog
   A /trunk/gcc/testsuite/g++.dg/torture/pr39764.C
   M /trunk/gcc/tree-ssa-ccp.c

2009-04-15  Richard Guenther  <rguenther@suse.de>

	PR tree-optimization/39764
	* tree-ssa-ccp.c (get_value): Canonicalize value with
	canonicalize_float_value.

	* g++.dg/torture/pr39764.C: New testcase.

------------------------------------------------------------------------
r146121 | rguenth | 2009-04-15 07:18:22 -0700 (Wed, 15 Apr 2009) | 8 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/tree-mudflap.c
   M /trunk/gcc/tree-ssa-pre.c
   M /trunk/gcc/tree-ssa-sccvn.c

2009-04-15  Richard Guenther  <rguenther@suse.de>

	* tree-ssa-pre.c (eliminate): When replacing a PHI node carry
	out a necessary conversion.
	* tree-ssa-sccvn.c (run_scc_vn): Also assign value-ids to
	names we didn't value number.
	* tree-mudflap.c (mf_build_check_statement_for): Use correct types.

------------------------------------------------------------------------
r146127 | ian | 2009-04-15 08:51:26 -0700 (Wed, 15 Apr 2009) | 37 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/c-opts.c
   M /trunk/gcc/c-parser.c
   M /trunk/gcc/c-tree.h
   M /trunk/gcc/c-typeck.c
   M /trunk/gcc/c.opt
   M /trunk/gcc/doc/invoke.texi
   M /trunk/gcc/testsuite/ChangeLog
   A /trunk/gcc/testsuite/gcc.dg/Wenum-compare-1.c

gcc/:
	* c.opt (Wenum-compare): Enable for C and Objc.  Initialize to -1.
	* c-opts.c (c_common_handle_option): For C, set warn_enum_compare
	for -Wall and for -Wc++-compat.
	(c_common_post_options): For C++, set warn_enum_compare if not
	already set.
	* c-tree.h (struct c_expr): Add field original_type.
	(build_external_ref): Update declaration.
	* c-parser.c (c_parser_braced_init): Set original_type.
	(c_parser_initelt): Likewise.
	(c_parser_expr_no_commas): Likewise.
	(c_parser_conditional_expression): Likewise.
	(c_parser_cast_expression): Likewise.
	(c_parser_unary_expression): Likewise.  Pull setting of
	original_code to top of function.
	(c_parser_sizeof_expression): Set original_type.
	(c_parser_alignof_expression): Likewise.
	(c_parser_postfix_expression): Likewise.  Pull setting of
	original_code to top of function.
	(c_parser_postfix_expression_after_paren_type): Set
	original_type.
	(c_parser_postfix_expression_after_primary): Likewise.
	(c_parser_expression): Likewise.
	* c-typeck.c (build_external_ref): Add type parameter.  Change all
	callers.
	(c_expr_sizeof_expr): Set original_type field.
	(parser_build_unary_op): Likewise.
	(parser_build_binary_op): Likewise.  Optionally warn about
	comparisons of enums of different types.
	(digest_init): Set original_type field.
	(really_start_incremental_init): Likewise.
	(push_init_level, pop_init_level): Likewise.
	* doc/invoke.texi (Warning Options): -Wenum-compare now
	supported in C.
gcc/testsuite/:
	* gcc.dg/Wenum-compare-1.c: New testcase.

------------------------------------------------------------------------
r146129 | rguenth | 2009-04-15 09:12:01 -0700 (Wed, 15 Apr 2009) | 11 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/omp-low.c
   M /trunk/gcc/tree-loop-distribution.c
   M /trunk/gcc/tree-profile.c

2009-04-15  Richard Guenther  <rguenther@suse.de>

	* omp-low.c (lower_rec_input_clauses): Build correct address
	expressions.
	(expand_omp_for_generic): Fix multiplication type.
	* tree-loop-distribution.c (build_size_arg): Build a
	size_t argument.
	(generate_memset_zero): Fix types.
	* tree-profile.c (prepare_instrumented_value): Correctly
	widen a pointer.

------------------------------------------------------------------------
r146130 | eager | 2009-04-15 09:46:14 -0700 (Wed, 15 Apr 2009) | 4 lines
Changed paths:
   M /trunk/gcc/ChangeLog
   M /trunk/gcc/config/rs6000/rs6000.c
   M /trunk/gcc/config/rs6000/rs6000.md

Allows a single-precision FP register to contain a SImode value.
This is only active when TARGET_SINGLE_FPU is true (e.g., --target=powerpc-xilinx-eabi).


------------------------------------------------------------------------
r146131 | bkoz | 2009-04-15 09:54:19 -0700 (Wed, 15 Apr 2009) | 8 lines
Changed paths:
   M /trunk/libstdc++-v3/ChangeLog
   M /trunk/libstdc++-v3/doc/xml/manual/status_cxx200x.xml
   M /trunk/libstdc++-v3/doc/xml/manual/status_cxxtr1.xml

2009-04-15  Benjamin Kosnik  <bkoz@redhat.com>

	* doc/xml/manual/status_cxxtr1.xml: Update to  new table style.
	* doc/xml/manual/status_cxx200x.xml: Adjust.




------------------------------------------------------------------------
r146132 | lcwu | 2009-04-15 10:55:50 -0700 (Wed, 15 Apr 2009) | 7 lines
Changed paths:
   M /trunk/gcc/cp/ChangeLog
   M /trunk/gcc/cp/call.c
   M /trunk/gcc/cp/cvt.c
   M /trunk/gcc/testsuite/ChangeLog
   A /trunk/gcc/testsuite/g++.dg/warn/Wunused-13.C

        PR c++/39551
	* gcc/cp/call.c (build_over_call): Set TREE_NO_WARNING on the
	compiler-generated INDIRECT_REF expression.
	* gcc/cp/cvt.c (convert_to_void): Emit warning when stripping off
	INDIRECT_REF.
	* gcc/testsuite/g++.dg/warn/Wunused-13.C: New testcase.

------------------------------------------------------------------------

For more information, see <http://gcc.gnu.org/regtest/HEAD/>.

Attachment: native-log.txt
Description: Text document

-- 
Geoffrey Keating <geoffk@apple.com> 
(via an automated GCC regression-testing script.)