[Bug c++/19813] [4.0 Regression] wrong code with -finline-limit

[jakub at gcc dot gnu dot org]

--- Additional Comments From jakub at gcc dot gnu dot org  2005-02-18 08:37 
---
Fixed.

-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19813

[Bug c/20043] [4.0 Regression] transparent_union doesn't allow restrict qualifier removal

[jakub at gcc dot gnu dot org]

-- 
   What|Removed |Added

 AssignedTo|unassigned at gcc dot gnu   |jakub at gcc dot gnu dot org
   |dot org |
 Status|NEW |ASSIGNED
   Last reconfirmed|2005-02-17 23:19:58 |2005-02-18 09:29:43
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20043

[Bug libstdc++/19554] std::insert_iterator does not work together with __gnu_cxx::hash_set

[pcarlini at suse dot de]

--- Additional Comments From pcarlini at suse dot de  2005-02-18 09:41 
---
Ok, the unordered containers are actually going in! No point in keeping this
open...

-- 
   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||WONTFIX


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19554

[Bug libfortran/20037] libfortran: format termination bug in formatted write

[Thomas dot Koenig at online dot de]

--- Additional Comments From Thomas dot Koenig at online dot de  2005-02-18 
10:42 ---
I think this is identical to PR 15332.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20037

[Bug tree-optimization/19952] ICE: tree check: expected class 'declaration', have 'statement' (label_expr) in tree_verify_flow_info, at tree-cfg.c:3709

[reichelt at gcc dot gnu dot org]

--- Additional Comments From reichelt at gcc dot gnu dot org  2005-02-18 
11:04 ---
Reduced testcase (just compile with "-O -ftree-vectorize"):

==
int i;

struct A
{
~A() { ++i; }
};

struct B
{
A a;
};

void foo()
{
for (int i=0; i<2; ++i)
{
B *p;
if (p) p->~B();
}
}
==


-- 
   What|Removed |Added

 CC||reichelt at gcc dot gnu dot
   ||org
  GCC build triplet|x86_64-unknown-linux-gnu|
   GCC host triplet|x86_64-unknown-linux-gnu|
 GCC target triplet|x86_64-unknown-linux-gnu|
   Keywords||monitored


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19952

[Bug c++/20051] New: ICE when compiling SSE intrinics

[m dot broadbent at signal dot qinetiq dot com]

g++ ICEs when compiling valid SSE code using intrinsics.  The code compiles
correctly if either, the optimisation level is -O0 or the line
'_mm_store_ps(pDataIn, accum);' is removed (line 30 in main.cpp).

/opt/gcc-3.4.3/bin/g++-3.4 -O2 -march=pentium4 -save-temps -Wall -c main.cpp
main.cpp: In function `void SSEFunc(float*, float*)':
main.cpp:16: warning: 'accum' might be used uninitialized in this function
main.cpp:33: internal compiler error: in immed_double_const, at emit-rtl.c:481
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.

GCC build:

Reading specs from /opt/gcc-3.4.3/lib/gcc/i686-pc-linux-gnu/3.4.3/specs
Configured with: ../configure --prefix=/opt/gcc-3.4.3 --program-suffix=-3.4
--enable-threads --with-tune=pentium4 --enable-languages=c,c++
Thread model: posix
gcc version 3.4.3

-- 
   Summary: ICE when compiling SSE intrinics
   Product: gcc
   Version: 3.4.3
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: m dot broadbent at signal dot qinetiq dot com
CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20051

[Bug c++/20051] ICE when compiling SSE intrinics

[m dot broadbent at signal dot qinetiq dot com]

--- Additional Comments From m dot broadbent at signal dot qinetiq dot com  
2005-02-18 11:07 ---
Created an attachment (id=8225)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8225&action=view)
Preprocessed source


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20051

[Bug tree-optimization/19952] ICE: tree check: expected class 'declaration', have 'statement' (label_expr) in tree_verify_flow_info, at tree-cfg.c:3709

[reichelt at gcc dot gnu dot org]

--- Additional Comments From reichelt at gcc dot gnu dot org  2005-02-18 
11:07 ---
: Search converges between 2004-09-02-trunk (#532) and 2004-09-04-trunk (#533).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19952

[Bug fortran/20052] New: Internal error pb:get_indirect_ref_operands at tree-ssa-operands.c:1449

[jean-pierre dot benoit at irsn dot fr]

Hi, 

I try to compile a FORTRAN 77 code with gfortran and i encounter the following
problem. Unfortunatly, I cant use g77 because I meet the other problem with
#[Bug libf2c/16970] long write times via NFS (with g77-3.2.2, g77-3.3.5 and
g77-3.4.3).


[EMAIL PROTECTED] erreur_de_compil_gcc-4.0]$ make
/produits/gcc-4.0-20050213/bin/gfortran -c -O2 -v -o LIMPLI.o LIMPLI.f
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: ../gcc-4.0-20050213/configure --prefix=/ 
--enable-languages=c++,f95
Thread model: posix
gcc version 4.0.0 20050213 (experimental)
 
/export/produits01/gcc-4.0-20050213/bin/../libexec/gcc/i686-pc-linux-gnu/4.0.0/f951
 LIMPLI.f -ffixed-form -quiet -dumpbase LIMPLI.f -mtune=pentiumpro 
-auxbase-strip LIMPLI.o -O2 -version -o /tmp/ccDNH6sP.s
GNU F95 version 4.0.0 20050213 (experimental) (i686-pc-linux-gnu)
compiled by GNU C version 3.2.2 20030222 (Red Hat Linux 3.2.2-5).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
LIMPLI.f: In function 'limpli':
LIMPLI.f:1: internal compiler error: in get_indirect_ref_operands, at
tree-ssa-operands.c:1449
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.
make: *** [LIMPLI.o] Erreur 1


Here is the subroutine:
  SUBROUTINE LIMPLI (LRET)
C SEGMENT /PV/ ( VPVG  .PV , VPVD  .PV
C1 , VCAT  *I  , VTYP  *I  , VLNI*I
C2 , VCST  *I  - UPARM *I
C3 , WPBI  .PB , WPBE  .PB , WPBBL.PB
C4 , WTVD  .TV - UTVEXT.TV
C5 , WTVI  .TV - UTVAFF.TV
C6 , WPVSUB.PV - UTVSUB.TV
C7 , WPVMIN.PV , WPVMAX.PV , WPVCOM.PV
C8 , WPVD  .PV , WPVT  .PV , WPVHOM.PV
C9 , WPV   .PV , WPVE  .PV
CA , VSYM*(LVSY))
C SEGMENT PV1.PV,PV2.PV,PV3.PV,PV4.PV
C SEGMENT TV(0).PV,TV1(0).PV,TV2(0).PV,TV3(0).PV,TV4(0).PV
  COMMON/OOOCOM/OOA(1),OOT,OOV(8)
  INTEGEROOA,OOT,OOV,OOO,OO1,OO2,OO3,OO4,PV,OO5
  INTEGERPV1,PV2,PV3,PV4,TV,TV1,TV2,TV3,TV4
  INTEGEROOI(1)
  CHARACTER*4OOH
  EQUIVALENCE(OOV(1),OOI(1),OOH)
  CHARACTER*1 HSY1(20*66+8)
  CHARACTER*80HSYM
  EQUIVALENCE(HSYM,HSY1(1))
  COMMON /LSYCOM/ LHSY
  COMMON /LSYCOH/ HSY1
  EQUIVALENCE (PSPE(1),PPARO)
  INTEGER   PSPE(25)
 1  ,PPARO,PPARF,PVIRG,PEGAL,PPVIR,PAPOS,PETCO,P2PNT,PPOIN,PEXPO
 2  ,PASTE,PSLAS,PPLUS,PMOIN,PLT  ,PGT  ,PEQ  ,PGE  ,PLE  ,PNE
 3  ,PNOT ,PAND ,POR  ,PEQV ,PNEQV
 4  ,PNUM   ,PVOC ,PVOC1,PVOC2,PVOC3,PVBLAN
  COMMON /LSPCOM/ NSPE,NIVP,ITYP,ILNI,ICST
 1  ,PPARO,PPARF,PVIRG,PEGAL,PPVIR,PAPOS,PETCO,P2PNT,PPOIN,PEXPO
 2  ,PASTE,PSLAS,PPLUS,PMOIN,PLT  ,PGT  ,PEQ  ,PGE  ,PLE  ,PNE
 3  ,PNOT ,PAND ,POR  ,PEQV ,PNEQV
 4  ,PNUM(10)   ,PVOC ,PVOC1,PVOC2,PVOC3,PVBLAN
  INTEGER PYLOGI,PYINTE,PYREAL,PYCOMP,PYCHAR,PYDOUB,
 1PYIMPL,PYEXTE,PYDIME,PYCOMM,PYEQUI,PYDATA,
 2PYPARA
  COMMON /LKYCOM/ PYLOGI,PYINTE,PYREAL,PYCOMP,PYCHAR,PYDOUB,
 1PYIMPL,PYEXTE,PYDIME,PYCOMM,PYEQUI,PYDATA,
 2PYPARA
  COMMON /LTYCOM/ JTYP(256),JLNI(256)
  INTEGER  PKY(6)
  EQUIVALENCE (PKY(1),PYLOGI)
 10   CALL LNEXT (LRET)
  GO TO (950,901,902, 12)  , LRET
C12   DO 14 KTYP=1,6
12DO 14 KTYP=1,6
PV1=PKY(KTYP)
C   IF (LHSY.EQ.PV1.VSYM(/1))   THEN
  IF(LHSY.EQ.OOA(OOA(OOT+PV1)+23))THEN
C IF (HSYM(1:LHSY).EQ.PV1.VSYM(1:LHSY)) GO TO 16
  IF(HSYM(1:LHSY).EQ.OOH(OOA(OOT+PV1)*4+84+1:OOA(OOT+PV1)*4+84+LHSY)
 *)GOTO16
ENDIF
 14 CONTINUE
   GO TO 904
 16   IF (KTYP.EQ.6)   THEN
CALL LXTYP3 (LRET,KTYP,KLNI)
IF (LRET.EQ.1) GO TO 950
C ELSEIF (NSPE.EQ.JASTE)   THEN
  ELSEIF(NSPE.EQ.11)THEN
CALL LNEXT (LRET)
GO TO (950,901, 18,903), LRET
C18 IF (ITYP.NE.TYPINT)GO TO 902
18IF(ITYP.NE.2)GOTO902
CALL LXTYP2 (LRET,KTYP,KLNI,ICST)
IF (LRET.EQ.1) GO TO 950
  ELSE
KLNI=1
  ENDIF
C IF (NSPE.NE.JPARO)   GO TO 901
  IF(NSPE.NE.1)GOTO901
 30   CALL LNEXT (LRET)
  GO TO (950,901,902, 32)  , LRET
 32   IF (LHSY.NE.1)   GO TO 905
  I1=ICHAR(HSY1(1))
  I2=I1
C IF (NSPE.EQ.JMOIN)   THEN
  IF(NSPE.EQ.14)THEN
CALL LNEXT (LRET)
GO TO (950,901,902, 34), LRET
 34 IF (LHSY.NE.1) GO TO 905
I2=ICHAR(HSY1(1))
  ENDIF
C DO 36 I=I1,I2
  DO 36 I=I1,I2
IF (JTYP(I+1).GT.0)

[Bug tree-optimization/19952] ICE: tree check: expected class 'declaration', have 'statement' (label_expr) in tree_verify_flow_info, at tree-cfg.c:3709

[reichelt at gcc dot gnu dot org]

--- Additional Comments From reichelt at gcc dot gnu dot org  2005-02-18 
11:43 ---
Devang, the regression was introduced with the introduction of
the if-conversion for the vectorizer by your patch
http://gcc.gnu.org/ml/gcc-cvs/2004-09/msg00207.html

Could you please have a look?


-- 
   What|Removed |Added

 CC||dpatel at gcc dot gnu dot
   ||org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19952

[Bug c++/20053] New: casting pointers to arrays in c++ produces wrong output when optimizing

[linuxadmin at yandex dot ru]

To: [EMAIL PROTECTED]
Subject: casting pointer to array in c++ produces wrong output when optimizing
From: [EMAIL PROTECTED] (Ivan G. Shevchenko)
Reply-To:
Cc:

>Submitter-Id:
>Originator:Ivan G. Shevchenko
>Organization:  -
>Confidential:  no
>Synopsis:  casting pointers to arrays in c++ produces wrong output when
optimizing
>Severity:  critical
>Priority:  medium
>Category:  c++
>Class: wrong-code
>Release:   3.4.3  AND  4.0.0 20050213 (experimental)
>Environment:
System: Linux acid 2.6.10-my #1 Fri Feb 11 12:13:56 UTC 2005 i686 athlon i386
GNU/Linux
Architecture: i686
host: i686-pc-linux-gnu
build: i686-pc-linux-gnu
target: i686-pc-linux-gnu
configured with: ../gcc-sources/configure --program-suffix=snapshot
--enable-libstdcxx-debug --enable-libstdcxx-debug-flags=-w -O0 -ggdb3
>Description:
accessing an uint32_t array (that is casted from an uint64_t pointer)
from an overloaded operator+ leads to wrong results when optimizing
with -O3, but produces correct results when optimizing with -O1
>How-To-Repeat:

#include
#include
using namespace std;

class TEST{
uint64_t data_;
public:
  uint32_t& _0() {return ((uint32_t*)&data_)[0];}
const uint32_t& _0()const{return ((uint32_t*)&data_)[0];}
  uint32_t& _1() {return ((uint32_t*)&data_)[1];}
const uint32_t& _1()const{return ((uint32_t*)&data_)[1];}
/*
// same effect
  uint32_t& _0() {return *((uint32_t*)&data_+0);}
const uint32_t& _0()const{return *((uint32_t*)&data_+0);}
  uint32_t& _1() {return *((uint32_t*)&data_+1);}
const uint32_t& _1()const{return *((uint32_t*)&data_+1);}
*/
};

TEST operator+(const TEST& a,const TEST& b){
TEST temp;
temp._0()=a._0()+b._0(); temp._1()=a._1()+b._1();
return temp;
}

int main(){
TEST a,b;
a._0()=a._1()=1;
b._0()=b._1()=2;

cout

[Bug tree-optimization/19951] ICE in tree_split_edge, at tree-cfg.c:3199 with -ftree-vectorize

[reichelt at gcc dot gnu dot org]

--- Additional Comments From reichelt at gcc dot gnu dot org  2005-02-18 
12:06 ---
Reduced testcase:

===
struct A
{
~A();
};

void foo();

void bar(int i)
{
A a;

if (i)
foo();
else
for (;;)
foo();
}
===


-- 
   What|Removed |Added

 CC||reichelt at gcc dot gnu dot
   ||org
   Keywords||monitored


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19951

[Bug tree-optimization/19951] ICE in tree_split_edge, at tree-cfg.c:3199 with -ftree-vectorize

[reichelt at gcc dot gnu dot org]

--- Additional Comments From reichelt at gcc dot gnu dot org  2005-02-18 
12:19 ---
: Search converges between 2004-11-19-014001-trunk (#644) and
2004-11-19-161001-trunk (#645).

It looks like this was introduced by one of your patches on 2004-11-19, Dorit.

Btw, here's an even shorter testcase:


struct A
{
~A();
};

void foo();

void bar()
{
A a;

foo();
for (;;)
foo();
}



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19951

[Bug tree-optimization/19903] ACATS cxa4006 cxa4017 fail at runtime

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
12:19 ---
Proposed fix: http://gcc.gnu.org/ml/gcc-patches/2005-02/msg01024.html


-- 
   What|Removed |Added

   Keywords||patch


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19903

[Bug c++/20008] [4.0 Regression] internal compiler error: in expand_case, at stmt.c:2397

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
12:21 ---
Subject: Bug 20008

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 12:20:50

Modified files:
gcc: ChangeLog stmt.c 
gcc/testsuite  : ChangeLog 
Added files:
gcc/testsuite/g++.dg/opt: switch4.C 

Log message:
gcc/ChangeLog:
PR c++/20008
* stmt.c (expand_case): Don't assume cleanup_tree_cfg will remove
cases that are out-of-range for the index type.
gcc/testsuite/ChangeLog:
PR c++/20008
* g++.dg/opt/switch4.C: New.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.7522&r2=2.7523
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/stmt.c.diff?cvsroot=gcc&r1=1.415&r2=1.416
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5046&r2=1.5047
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/opt/switch4.C.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20008

[Bug c++/20008] [4.0 Regression] internal compiler error: in expand_case, at stmt.c:2397

[jakub at gcc dot gnu dot org]

--- Additional Comments From jakub at gcc dot gnu dot org  2005-02-18 12:26 
---
Fixed.

-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20008

[Bug ada/18819] ACATS cdd2a02 fails at runtime

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
12:27 ---
It's SRA again, but the problem really originates in the FE's type frobbing with
VIEW_CONVERT_EXPR and the like.  I think we should consider disabling SRA for
the Ada compiler until this mess is sorted out.


-- 
   What|Removed |Added

 CC||ebotcazou at gcc dot gnu dot
   ||org
 AssignedTo|ebotcazou at gcc dot gnu dot|unassigned at gcc dot gnu
   |org |dot org
 Status|ASSIGNED|NEW


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18819

[Bug ada/18659] [4.0 Regression] ACATS ICE in bug in tree-sra.c:1507 on x86 and ppc-darwin (6 FAIL)

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
12:31 ---
SRA is an Ada killer because of the the abuse of VIEW_CONVERT_EXPRs by the Ada
front-end.  I think it should be disabled in the Ada compiler for 4.0.x.


-- 
   What|Removed |Added

 CC|laurent at guerby dot net   |ebotcazou at gcc dot gnu dot
   ||org
   Last reconfirmed|2005-01-18 00:45:39 |2005-02-18 12:31:03
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18659

[Bug ada/18302] ACATS test c953002 (and others) hangs

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
12:35 ---
Presumably fixed.  Reopen if not.


-- 
   What|Removed |Added

 Status|WAITING |RESOLVED
 Resolution||FIXED
   Target Milestone|--- |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18302

[Bug c++/20053] casting pointers to arrays in c++ produces wrong output when optimizing

[falk at debian dot org]

--- Additional Comments From falk at debian dot org  2005-02-18 12:38 
---
This code is invalid; you're accessing an object of type uint64_t via a pointer
to uint32_t, which is not allowed. Use -fno-strict-aliasing, or use a union.


-- 
   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution||INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20053

[Bug ada/19140] ACATS c37402a segfaults at runtime

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
12:38 ---
I get the segfault on x86 as of today.


-- 
   What|Removed |Added

 CC||ebotcazou at gcc dot gnu dot
   ||org
 Status|UNCONFIRMED |NEW
 Ever Confirmed||1
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 12:38:43
   date||
Summary|ACATS c37402a segfault at   |ACATS c37402a segfaults at
   |runtime on x86_64   |runtime


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19140

[Bug ada/19140] ACATS c37402a segfaults at runtime

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
12:43 ---
Investigating.


-- 
   What|Removed |Added

 AssignedTo|unassigned at gcc dot gnu   |ebotcazou at gcc dot gnu dot
   |dot org |org
 Status|NEW |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19140

[Bug ada/18659] [4.0 Regression] ACATS ICE in bug in tree-sra.c:1507 on x86 and ppc-darwin (6 FAIL)

[charlet at adacore dot com]

--- Additional Comments From charlet at adacore dot com  2005-02-18 12:58 
---
Subject: Re:  [4.0 Regression] ACATS ICE in bug in tree-sra.c:1507 on x86 and 
ppc-darwin (6 FAIL)

> SRA is an Ada killer because of the the abuse of VIEW_CONVERT_EXPRs by the Ada
> front-end.  I think it should be disabled in the Ada compiler for 4.0.x.

Sounds reasonable to me.

Arno


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18659

[Bug c++/20032] ICE on non-constant array size in template

[reichelt at gcc dot gnu dot org]

--- Additional Comments From reichelt at gcc dot gnu dot org  2005-02-18 
13:00 ---
The problem can be reduced to the following testacse:

==
int Sum(int);

template struct A
{
int X[Sum(2)];
void foo() { sizeof(X); }
};

void bar()
{
A<0> a;
a.foo();
}
==

The code is invalid, since the array size "Sum(2)" is not a constant.
Even the definition 
  inline int Sum(unsigned int n) { return n*(n+1)/2; }
from the original test case doesn't change this.

The ICE was fixed by Mark's patch for PR 18429
http://gcc.gnu.org/ml/gcc-cvs/2004-11/msg00582.html
We now (i.e. in gcc 4.0.0) emit a proper error message.

Closing, since this is not a regression.


-- 
   What|Removed |Added

 CC||reichelt at gcc dot gnu dot
   ||org
 Status|UNCONFIRMED |RESOLVED
   Keywords||ice-on-invalid-code
 Resolution||FIXED
Summary|internal error  |ICE on non-constant array
   ||size in template
   Target Milestone|--- |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20032

[Bug target/20054] New: ICE in change_address_1

[jakub at gcc dot gnu dot org]

/* { dg-do compile } */
/* { dg-options "-O2 -fno-strict-aliasing" } */

unsigned int *foo (void);

char *
bar (double *d)
{
  return (char *) (d + 1) - sizeof (unsigned int);
}

char
baz (double x)
{
  unsigned int h = *foo ();
  unsigned int l = *(unsigned int *) bar (&x);

  return (h & ~0x8000L) == 0x7FF0 && l == 0;
}

when compiled with -m64 -O2 -fno-strict-aliasing dies with ICE in
change_address_1.

This is a regression from 3.4.x.

-- 
   Summary: ICE in change_address_1
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: target
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: jakub at gcc dot gnu dot org
CC: gcc-bugs at gcc dot gnu dot org
GCC target triplet: s390x-redhat-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug target/20054] ICE in change_address_1

[jakub at gcc dot gnu dot org]

--- Additional Comments From jakub at gcc dot gnu dot org  2005-02-18 13:32 
---
The problem is alter_subreg is called during rest_of_handle_flow2 splitting
on (subreg:DI (mem:SI (reg:DI 12 %r12 [orig:46 D.1072 ] [46]) [0 S4 A32]) 0).

As flow2 is already past reload, I think this shouldn't be valid RTL, as per
documentation:
"The reload pass ensures that paradoxical references are only made to hard
registers."

The insn in question is
(insn 77 13 16 0 (set (reg:DI 1 %r1 [51])
(and:DI (subreg:DI (mem:SI (reg:DI 12 %r12 [orig:46 D.1072 ] [46]) [0 S4
A32]) 0)
(const_int 2147483647 [0x7fff]))) 98 {*llgt_didi} (nil)


-- 
   What|Removed |Added

 CC||uweigand at gcc dot gnu dot
   ||org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug other/20055] New: Many source files are unnecessarily executable

[gcc-bugzilla at gcc dot gnu dot org]

I just noticed that many source files are unnecessarily marked executable.
This needs to be fixed by someone with direct access to the CVS repository.

The list below is pruned of anything which obviously needs to be executable,,
like configure scripts etc.

boehm-gc/AmigaOS.c
boehm-gc/cord/de_win.ICO
boehm-gc/digimars.mak
boehm-gc/doc/README.arm.cross
gcc/ada/a-chzla1.ads
gcc/ada/a-chzla9.ads
gcc/ada/a-elchha.ads
gcc/ada/a-lfztio.ads
gcc/ada/a-liztio.ads
gcc/ada/a-llfzti.ads
gcc/ada/a-llizti.ads
gcc/ada/a-sfztio.ads
gcc/ada/a-siztio.ads
gcc/ada/a-ssizti.ads
gcc/ada/a-stzbou.adb
gcc/ada/a-stzbou.ads
gcc/ada/a-stzfix.adb
gcc/ada/a-stzfix.ads
gcc/ada/a-stzhas.adb
gcc/ada/a-stzhas.ads
gcc/ada/a-stzmap.adb
gcc/ada/a-stzmap.ads
gcc/ada/a-stzsea.adb
gcc/ada/a-stzsea.ads
gcc/ada/a-stzsup.adb
gcc/ada/a-stzsup.ads
gcc/ada/a-stzunb.adb
gcc/ada/a-stzunb.ads
gcc/ada/a-swunau.adb
gcc/ada/a-swunau.ads
gcc/ada/a-szmzco.ads
gcc/ada/a-szunau.adb
gcc/ada/a-szunau.ads
gcc/ada/a-szunha.adb
gcc/ada/a-szunha.ads
gcc/ada/a-szuzti.adb
gcc/ada/a-szuzti.ads
gcc/ada/a-tiunio.ads
gcc/ada/a-wwunio.ads
gcc/ada/a-ztcoau.adb
gcc/ada/a-ztcoau.ads
gcc/ada/a-ztcoio.adb
gcc/ada/a-ztcoio.ads
gcc/ada/a-ztcstr.adb
gcc/ada/a-ztcstr.ads
gcc/ada/a-ztdeau.adb
gcc/ada/a-ztdeau.ads
gcc/ada/a-ztdeio.adb
gcc/ada/a-ztdeio.ads
gcc/ada/a-ztedit.adb
gcc/ada/a-ztedit.ads
gcc/ada/a-ztenau.adb
gcc/ada/a-ztenau.ads
gcc/ada/a-ztenio.adb
gcc/ada/a-ztenio.ads
gcc/ada/a-ztexio.adb
gcc/ada/a-ztexio.ads
gcc/ada/a-ztfiio.adb
gcc/ada/a-ztfiio.ads
gcc/ada/a-ztflau.adb
gcc/ada/a-ztflau.ads
gcc/ada/a-ztflio.adb
gcc/ada/a-ztflio.ads
gcc/ada/a-ztgeau.adb
gcc/ada/a-ztgeau.ads
gcc/ada/a-ztinau.adb
gcc/ada/a-ztinau.ads
gcc/ada/a-ztinio.adb
gcc/ada/a-ztinio.ads
gcc/ada/a-ztmoau.adb
gcc/ada/a-ztmoau.ads
gcc/ada/a-ztmoio.adb
gcc/ada/a-ztmoio.ads
gcc/ada/a-zttest.adb
gcc/ada/a-zttest.ads
gcc/ada/a-zzunio.ads
gcc/ada/g-memdum.adb
gcc/ada/g-memdum.ads
gcc/ada/g-sestin.ads
gcc/ada/g-utf_32.adb
gcc/ada/g-utf_32.ads
gcc/ada/g-zstspl.ads
gcc/ada/i-vxwork-x86.ads
gcc/ada/s-addope.adb
gcc/ada/s-addope.ads
gcc/ada/s-exnint.adb
gcc/ada/s-exnllf.adb
gcc/ada/s-exnlli.adb
gcc/ada/s-expint.adb
gcc/ada/s-explli.adb
gcc/ada/s-restri.adb
gcc/ada/s-restri.ads
gcc/ada/seh_init.c
gcc/config/i386/host-mingw32.c
gcc/testsuite/g++.dg/ext/dll-MI1.h
gcc/testsuite/g++.dg/ext/dllexport-MI1.C
gcc/testsuite/g++.dg/ext/dllexport1.C
gcc/testsuite/g++.dg/ext/dllimport-MI1.C
gcc/testsuite/g++.dg/ext/dllimport10.C
gcc/testsuite/g++.dg/ext/dllimport3.C
gcc/testsuite/g++.dg/ext/dllimport4.C
gcc/testsuite/g++.dg/ext/dllimport5.C
gcc/testsuite/g++.dg/ext/dllimport6.C
gcc/testsuite/g++.dg/ext/dllimport7.C
gcc/testsuite/g++.dg/ext/dllimport9.C
gcc/testsuite/g++.dg/template/array11.C
gcc/testsuite/g++.dg/template/array3.C
gcc/testsuite/g++.dg/template/friend33.C
gcc/testsuite/g++.dg/template/partial3.C
gcc/testsuite/g++.dg/template/spec15.C
gcc/testsuite/g++.dg/template/spec16.C
gcc/testsuite/g++.dg/template/ttp7.C
gcc/testsuite/gcc.c-torture/execute/20031012-1.c
gcc/testsuite/gcc.c-torture/execute/970923-1.c
gcc/testsuite/gcc.dg/dll-5.c
gcc/testsuite/gcc.dg/vect/pr16105.c
gcc/testsuite/gcc.dg/vect/pr18400.c
gcc/testsuite/gcc.dg/vect/tree-vect.h
gcc/testsuite/gcc.dg/vect/vect-1.c
gcc/testsuite/gcc.dg/vect/vect-10.c
gcc/testsuite/gcc.dg/vect/vect-11.c
gcc/testsuite/gcc.dg/vect/vect-12.c
gcc/testsuite/gcc.dg/vect/vect-13.c
gcc/testsuite/gcc.dg/vect/vect-14.c
gcc/testsuite/gcc.dg/vect/vect-15.c
gcc/testsuite/gcc.dg/vect/vect-16.c
gcc/testsuite/gcc.dg/vect/vect-17.c
gcc/testsuite/gcc.dg/vect/vect-18.c
gcc/testsuite/gcc.dg/vect/vect-19.c
gcc/testsuite/gcc.dg/vect/vect-2.c
gcc/testsuite/gcc.dg/vect/vect-20.c
gcc/testsuite/gcc.dg/vect/vect-21.c
gcc/testsuite/gcc.dg/vect/vect-22.c
gcc/testsuite/gcc.dg/vect/vect-23.c
gcc/testsuite/gcc.dg/vect/vect-24.c
gcc/testsuite/gcc.dg/vect/vect-25.c
gcc/testsuite/gcc.dg/vect/vect-26.c
gcc/testsuite/gcc.dg/vect/vect-27.c
gcc/testsuite/gcc.dg/vect/vect-28.c
gcc/testsuite/gcc.dg/vect/vect-29.c
gcc/testsuite/gcc.dg/vect/vect-3.c
gcc/testsuite/gcc.dg/vect/vect-30.c
gcc/testsuite/gcc.dg/vect/vect-31.c
gcc/testsuite/gcc.dg/vect/vect-32.c
gcc/testsuite/gcc.dg/vect/vect-33.c
gcc/testsuite/gcc.dg/vect/vect-34.c
gcc/testsuite/gcc.dg/vect/vect-36.c
gcc/testsuite/gcc.dg/vect/vect-4.c
gcc/testsuite/gcc.dg/vect/vect-40.c
gcc/testsuite/gcc.dg/vect/vect-41.c
gcc/testsuite/gcc.dg/vect/vect-42.c
gcc/testsuite/gcc.dg/vect/vect-43.c
gcc/testsuite/gcc.dg/vect/vect-44.c
gcc/testsuite/gcc.dg/vect/vect-45.c
gcc/testsuite/gcc.dg/vect/vect-46.c
gcc/testsuite/gcc.dg/vect/vect-47.c
gcc/testsuite/gcc.dg/vect/vect-48.c
gcc/testsuite/gcc.dg/vect/vect-49.c
gcc/testsuite/gcc.dg/vect/vect-5.c
gcc/testsuite/gcc.dg/vect/vect-50.c
gcc/testsuite/gcc.dg/vect/vect-51.c
gcc/testsuite/gcc.dg/vect/vect-52.c
gcc/testsuite/gcc.dg/vect/vect-53.c
gcc/testsuite/gcc.dg/vect/vect-54.c
gcc/testsuite/gcc.dg/vect/vect-55.c
gcc/testsuite/gcc.dg/vect/vect-56.c
gcc/testsuite/gcc.dg/vect/vect-57.c
gcc/testsuite/gcc.dg/vect/vect-58.c
gcc/testsuite

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[pinskia at gcc dot gnu dot org]

-- 
   What|Removed |Added

   Keywords||ice-on-valid-code
Summary|ICE in change_address_1 |[4.0 Regression] ICE in
   ||change_address_1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug middle-end/20030] [4.0 Regression] Broken arguments access

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
14:32 ---
Subject: Bug 20030

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 14:31:54

Modified files:
gcc: ChangeLog fold-const.c 

Log message:
2005-02-18  Andrew Pinski  <[EMAIL PROTECTED]>

PR middle-end/20030
* fold-const.c (fold_indirect_ref_1): Use the correct index for zero 
access,
the lower bound of the array type if it exists.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.7523&r2=2.7524
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/fold-const.c.diff?cvsroot=gcc&r1=1.514&r2=1.515



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20030

[Bug middle-end/20030] [4.0 Regression] Broken arguments access

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
14:37 ---
Fixed.
--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
14:37 ---
Subject: Bug 20030

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 14:36:55

Modified files:
gcc/testsuite  : ChangeLog 
Added files:
gcc/testsuite/gfortran.fortran-torture/execute: 
character_passing.f90 

Log message:
2004-02-18  Andrew Pinski  <[EMAIL PROTECTED]>

PR middle-end/20030
* gfortran.fortran-torture/execute/character_passing.f90: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5047&r2=1.5048
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.fortran-torture/execute/character_passing.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED
   Target Milestone|--- |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20030

[Bug middle-end/20030] [4.0 Regression] Broken arguments access

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
14:37 ---
Fixed.
--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
14:37 ---
Subject: Bug 20030

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 14:36:55

Modified files:
gcc/testsuite  : ChangeLog 
Added files:
gcc/testsuite/gfortran.fortran-torture/execute: 
character_passing.f90 

Log message:
2004-02-18  Andrew Pinski  <[EMAIL PROTECTED]>

PR middle-end/20030
* gfortran.fortran-torture/execute/character_passing.f90: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5047&r2=1.5048
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gfortran.fortran-torture/execute/character_passing.f90.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED
   Target Milestone|--- |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20030

[Bug target/20051] ICE when compiling SSE intrinics

[pinskia at gcc dot gnu dot org]

-- 
   What|Removed |Added

  Component|c++ |target
   Keywords||ice-on-valid-code


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20051

[Bug tree-optimization/20052] Internal error pb:get_indirect_ref_operands at tree-ssa-operands.c:1449

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
14:50 ---
Hmm, this is fixed now via the following two patches:
2005-02-18  Andrew Pinski  <[EMAIL PROTECTED]>

PR middle-end/20030
* fold-const.c (fold_indirect_ref_1): Use the correct index for zero 
access,
the lower bound of the array type if it exists.

2005-02-13  Jason Merrill  <[EMAIL PROTECTED]>

PR c++/16405
* fold-const.c (fold_indirect_ref_1): Split out from...
(build_fold_indirect_ref): Here.
(fold_indirect_ref): New fn.
* tree.h: Declare it.
* gimplify.c (gimplify_compound_lval): Call fold_indirect_ref.
(gimplify_modify_expr_rhs): Likewise.
(gimplify_expr): Likewise.



-- 
   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
  Component|fortran |tree-optimization
   Keywords||ice-on-valid-code
 Resolution||FIXED
   Target Milestone|--- |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20052

[Bug java/20056] New: [4.0 regression] 'verification failed: incompatible type on stack' with --indirect-dispatch

[bonniot at users dot sf dot net]

The testcase is a tiny bytecode class. gcj 3.3 and 3.4 can compile it, and the
bcel verifier says it is valid. However, gcj 4.0.0 fails on it when
--indirect-dispatch is enabled:

$ /usr/lib/gcc-snapshot/bin/gcj --indirect-dispatch test/ACatch.class
test/main.nice: In class 'test.ACatch':
test/main.nice: In constructor '(test.Monotype)':
test/main.nice:1: error: verification failed: incompatible type on stack

-- 
   Summary: [4.0 regression] 'verification failed: incompatible type
on stack' with --indirect-dispatch
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: java
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: bonniot at users dot sf dot net
CC: gcc-bugs at gcc dot gnu dot org,java-prs at gcc dot gnu
dot org
  GCC host triplet: i386-debian-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20056

[Bug java/20056] [4.0 regression] 'verification failed: incompatible type on stack' with --indirect-dispatch

[bonniot at users dot sf dot net]

--- Additional Comments From bonniot at users dot sf dot net  2005-02-18 
14:53 ---
Created an attachment (id=8226)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8226&action=view)
testcase


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20056

[Bug target/20016] Compiling libgcc2.c with -Os for avr-gcc?

[pinskia at gcc dot gnu dot org]

-- 
   What|Removed |Added

 Status|UNCONFIRMED |NEW
 Ever Confirmed||1
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 14:55:32
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20016

[Bug target/20045] gcc.dg/ia64-fptr-1.c fails on ia64-hpux

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
14:57 ---
I think this testcase should only be testing on ia64-*-linux*

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20045

[Bug fortran/20057] New: internal compiler error: Bus error

[dir at lanl dot gov]

I get a bus error with macintosh gfortran from the 20050213 package. Here is
a simple routine that has the problem -

[dranta:~/tests/gfortran] dir% gfortran -c mesh3d.f
mesh3d.f: In function 'mesh3d':
mesh3d.f:9: internal compiler error: Bus error
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.
[dranta:~/tests/gfortran] dir% cat mesh3d.f
  SUBROUTINE MESH3D (XN,YN,ZN,IDEF,NI,NJ,NK,CIJ,NSEG1,IPRINT,NSKIP)
C
  COMMON /IODATA/ DATA(128)
  COMMON /IOSTRG/ STRG(128)
  CHARACTER*12 STRG
C
  DO 290 N=1,32,4
  NM=NEAR(DATA(N))
  IF (NM.EQ.0) GO TO 400
  IF (NM.LT.5.OR.NM.GT.NPT) THEN
WRITE(STRG,1000) NPT
CALL ERROR2(114,NM,IMIN,FLOAT(JMIN),FLOAT(KMIN),COMAND,STRG)
  END IF
  290 CONTINUE
  GO TO 400
  400 CALL STEVE(INT,XN,YN,ZN,NI,NJ,NF,CIJ,NSEG1)
 1000 FORMAT(I2)
  END

-- 
   Summary: internal compiler error: Bus error
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: dir at lanl dot gov
CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: powerpc-apple-darwin7.8.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20057

[Bug other/20055] Many source files are unnecessarily executable

[pinskia at gcc dot gnu dot org]

-- 
   What|Removed |Added

 CC||overseers at gcc dot gnu dot
   ||org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20055

[Bug fortran/20057] internal compiler error: Bus error

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
15:23 ---


*** This bug has been marked as a duplicate of 15966 ***

-- 
   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution||DUPLICATE


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20057

[Bug fortran/15966] ICE and segmentation fault on internal write

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
15:23 ---
*** Bug 20057 has been marked as a duplicate of this bug. ***

-- 
   What|Removed |Added

 CC||dir at lanl dot gov


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15966

[Bug tree-optimization/19828] [4.0 Regression] LIM is pulling out a pure function even though there is something which can modify global memory

[jakub at gcc dot gnu dot org]

--- Additional Comments From jakub at gcc dot gnu dot org  2005-02-18 15:55 
---
Just got hit by this too on subversion.
Distilled testcase is:
typedef __SIZE_TYPE__ size_t;
extern size_t strlen (const char *s);
extern int strncmp (const char *s1, const char *s2, size_t n);
extern void abort (void);

const char *a[16] = { "a", "bc", "de", "fgh" };

int
foo (char *x, const char *y, size_t n)
{
  size_t i, j = 0;
  for (i = 0; i < n; i++)
{
  if (strncmp (x + j, a[i], strlen (a[i])) != 0)
return 2;
  j += strlen (a[i]);
  if (y)
j += strlen (y);
}
  return 0;
}

int
main (void)
{
  if (foo ("abcde", (const char *) 0, 3) != 0)
abort ();
  return 0;
}

With Zdenek's definition of pure the only pure functions in builtins.def
would be is*/to*, none of the string/memory functions could be.  But the
documentation explicitely mentions strlen and memcmp as pure functions.

-- 
   What|Removed |Added

 CC||jakub at redhat dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19828

[Bug fortran/20058] New: Error on kind 16 hex data statement

[dir at lanl dot gov]

I get an error with the Macintosh gfortran from the 20050213 package. This
compilies with out problem on all the f90 and f95 compiliers that I have tried.
Puting "Z'40028000'_INT8" as the data statement will let gfortran
compile it, but then Absoft 9.0, Layhey 6.2 and IBM will reject it.

[dranta:~/tests/gfortran] dir% gfortran -c sagechanges.f90
 In file sagechanges.f90:5

data ieee64_two / Z'40028000'/
1
Error: Integer too big for integer kind 4 at (1)
[dranta:~/tests/gfortran] dir% cat sagechanges.f90
  subroutine get_real_type()
implicit none
integer, parameter :: INT8  = SELECTED_INT_KIND(16)
integer(INT8) :: ieee64_two
data ieee64_two / Z'40028000'/
  end subroutine get_real_type

-- 
   Summary: Error on kind 16 hex data statement
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: dir at lanl dot gov
CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: powerpc-apple-darwin7.8.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20058

[Bug tree-optimization/19828] [4.0 Regression] LIM is pulling out a pure function even though there is something which can modify global memory

[rakdver at gcc dot gnu dot org]

--- Additional Comments From rakdver at gcc dot gnu dot org  2005-02-18 
16:16 ---
OK, I agree that definition of ``pure'' needs to be changed in order to be 
useful (and to match the expectations); obviously, any function that is not 
total does not match the current definition.

What I find somewhat troublesome is that the "upgraded" definition of pure puts 
some of obligation on user of the function, rather than function itself.  The 
definition matching the expected semantics would need to be something like
"Pure function is guaranteed to be always called in such a way that it has no
side effects."

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19828

[Bug fortran/20059] New: internal compiler error: Segmentation Fault - For common blocks

[dir at lanl dot gov]

I get an Segmentation Fault with the Macintosh gfortran from the 20050213
package for this program.


[dranta:~/tests/gfortran] dir% gfortran -c plotz.f
[address=6f6c0004 pc=000140d0]
plotz.f: In function 'plotz':
plotz.f:1: internal compiler error: Segmentation Fault
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.
[dranta:~/tests/gfortran] dir% cat plotz.f
  SUBROUTINE PLOTZ
  IMPLICIT DOUBLE PRECISION (A-H,O-Z)
  COMMON /HEAD/ HED,P(3)
  CHARACTER HED*80,P*1
  COMMON /ULAB/ LAB(5)
  CHARACTER*80 LAB
  COMMON /IODATA/ DATA(64)
  COMMON /TAPES/ NOUT,NTAPE2,NTAPE3,NTAPE4,NTAPE5,NTAPE6,NTAPE7,
 $ NTAPE8,NTAPE9
  COMMON /CCPOOL/ RMIN,RMAX,ZMIN,ZMAX,IMIN,JMIN,IMAX,JMAX,NFLOP,
 $ HTP,SCALEP,ANGP,CCXMAX,WIDTH
C
RETURN
  END

-- 
   Summary: internal compiler error: Segmentation Fault - For common
blocks
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: dir at lanl dot gov
CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: powerpc-apple-darwin7.8.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20059

[Bug java/18362] internal compiler error: in size_binop, at fold-const.c:1598

[bonniot at users dot sf dot net]

--- Additional Comments From bonniot at users dot sf dot net  2005-02-18 
16:48 ---
Created an attachment (id=8228)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8228&action=view)
testcase

This is a 6K jar that exhibits the same behaviour:

$ /usr/lib/gcc-snapshot/bin/gcj --indirect-dispatch N.jar
bossa/syntax/ast.nice:0: internal compiler error: in size_binop, at
fold-const.c:1637

That's with gcj 4.0.0 20050212.

Note that if you remove any of the three classes from the jar, the bug does not
occur. Neither does it if you unpack the jar and call gcj on the three classes
directly.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18362

[Bug java/18362] internal compiler error: in size_binop, at fold-const.c:1598

[bonniot at users dot sf dot net]

-- 
   What|Removed |Added

 CC||bonniot at users dot sf dot
   ||net


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18362

configuration problem using --enable-intermodule on x86_64-unknown-linux-gnu

[Kenneth Zadeck]

The build gets partially thru the stage2 and then crashes when trying to 
compile most of the back end in one step with the following message:

gcc: cannot specify -o with -c or -S and multiple compilations
This same step works fine on my pentium (suse 9.2) and darwin machines.  
I am compiling against the gcc/mainline. My x86_64 machine is running 
suse 9.2 with all the current patches installed.

my configuration command was:
../lorien/gccBaseline/configure --enable-intermodule
Kenny

[Bug c++/20060] New: Ambiguous MSG in array initialization when expression is not integral constant.

[msadoghi at ca dot ibm dot com]

Sourc code for t.cpp:

int mymain()
{
int n = 10;
int (*p)[] = new int [2][n]; // error - 2nd dimension not constant
return 0;
}





Expected Behaviour:

"t.cpp", line 4.34: 1540-0016 (S) The expression must be an integral
non-volatile constant expression.




Actual Behaviour:

t.cpp: In function `int mymain()':
t.cpp:4: cannot convert `int (*)[((n - 1) + 1)]' to `int (*)[]' in 
initialization




t.ii generated with -save-temps option

# 1 "t.cpp"
# 1 ""
# 1 ""
# 1 "t.cpp"
int mymain()
{
int n = 10;
int (*p)[] = new int [2][n]; 
return 0;
}





Release:

GCC Version: 3.2.0





Environment:
System Type:

Reading specs from 
/usr/local/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/3.2/specs
Configured with: /scratch/gcc-3.2/configure --prefix=/usr/local/gcc.3.2.0
--enable-threads=aix --disable-nls
Thread model: aix
gcc version 3.2
 /usr/local/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/3.2/cpp0 -lang-c++
-D__GNUG__=3 -D__DEPRECATED -D__EXCEPTIONS -v -iprefix
/usr/local/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/3.2/ -D__GNUC__=3
-D__GNUC_MINOR__=2 -D__GNUC_PATCHLEVEL__=0 -D__GXX_ABI_VERSION=102 -D_IBMR2
-D_POWER -D_LONG_LONG -D_AIX -D_AIX32 -D_AIX41 -D_AIX43 -D_AIX51 -D_IBMR2
-D_POWER -D_LONG_LONG -D_AIX -D_AIX32 -D_AIX41 -D_AIX43 -D_AIX51 -Asystem=unix
-Asystem=aix -D__NO_INLINE__ -D__STDC_HOSTED__=1 -D_XOPEN_SOURCE=500
-D_XOPEN_SOURCE_EXTENDED=1 -D_LARGE_FILE_API -D_ALL_SOURCE
-D__WCHAR_TYPE__=short unsigned int -D_ARCH_COM
/home/msadoghi/scenario/common/t.cpp t.ii
GNU CPP version 3.2 (cpplib)
ignoring nonexistent directory
"/usr/local/lib/gcc-lib/../../powerpc-ibm-aix5.1.0.0/include"
ignoring nonexistent directory 
"/usr/local/gcc.3.2.0/powerpc-ibm-aix5.1.0.0/include"
ignoring duplicate directory "/usr/local/gcc.3.2.0/include/c++/3.2"
ignoring duplicate directory
"/usr/local/gcc.3.2.0/include/c++/3.2/powerpc-ibm-aix5.1.0.0"
ignoring duplicate directory "/usr/local/gcc.3.2.0/include/c++/3.2/backward"
ignoring duplicate directory
"/usr/local/gcc.3.2.0/lib/gcc-lib/powerpc-ibm-aix5.1.0.0/3.2/include"
#include "..." search starts here:
#include <...> search starts here:
 /usr/local/lib/gcc-lib/../../include/c++/3.2
 /usr/local/lib/gcc-lib/../../include/c++/3.2/powerpc-ibm-aix5.1.0.0
 /usr/local/lib/gcc-lib/../../include/c++/3.2/backward
 /usr/local/lib/gcc-lib/powerpc-ibm-aix5.1.0.0/3.2/include
 /usr/local/include
 /usr/local/gcc.3.2.0/include
 /usr/include
End of search list.
 /usr/local/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/3.2/cc1plus -fpreprocessed
t.ii -quiet -dumpbase t.cpp -version -o t.s
GNU CPP version 3.2 (cpplib)
GNU C++ version 3.2 (powerpc-ibm-aix5.1.0.0)
compiled by GNU C version 3.2.
/home/msadoghi/scenario/common/t.cpp: In function `int mymain()':
/home/msadoghi/scenario/common/t.cpp:4: cannot convert `int (*)[((n - 1) + 1)]'
   to `int (*)[]' in initialization





How-To-Repeat:

g++ -v -save-temps -c t.cpp



Note:

This is a diagnostic test case. Compilation failure is expected. However the 
message
produces by gcc compiler it ambiguous and can be improved. The expression [((n -
1) + 1)]
should be replaced simply with [n].


-- 
   Summary: Ambiguous MSG in array initialization when expression is
not integral constant.
   Product: gcc
   Version: 3.2
Status: UNCONFIRMED
  Severity: minor
  Priority: P2
 Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: msadoghi at ca dot ibm dot com
CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20060

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[jakub at gcc dot gnu dot org]

--- Additional Comments From jakub at gcc dot gnu dot org  2005-02-18 17:33 
---
I think the reason why this paradoxical subreg of memory survived post reload
is that the subreg is hardcoded in the llgt_sidi/llgt_sidi_split patterns.
All other patterns have explicit subreg only around register_operands, these
are the only two that use it for memory_operand, though I see arm and h8300
also use this in their machine descriptions.

Wouldn't it be better to describe llgt_sidi as
(zero_extend:DI (and:SI (match_operand:SI "memory_operand") (0x7fff)))
instead?

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug c/20061] New: Certain flags generate warnings when they should not.

[wmahoney at unomaha dot edu]

I am interested in looking at the GCC loop optimization. There are some tree 
dump flags that are registered (via register_dump_files() in tree-optimize.c). 
In particular, -fdump-tree-loopinit-raw and -fdump-tree-loopdone-raw that 
generate warnings when they should not. Specifically:

gcc -O3 -S -fdump-tree-loopinit-raw short.c

The problem is that there is also -fdump-tree-loop-raw as well, and "loop" is 
a prefix of "loopinit" and "loopdone". The function dump_switch_p_1() in tree-
dump.c calls skip_leading_substring() which works correctly for "loopinit" but 
then skips only the first four bytes when processing "loop". This 
leaves "init" dangling, and generates:

cc1: warning: ignoring unknown option 'init' in '-fdump-tree-loop'

Same for "done" of course.

No need to do -save-temps or look at any of the compiler output - this happens 
before any processing of the input file (a 4-line C program will suffice). 

Also, no urgency because the --first-- encounter where it does match the 
entire option works, so you do get the requested dumps, just the incorrect 
warning.

-- 
   Summary: Certain flags generate warnings when they should not.
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: enhancement
  Priority: P3
 Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: wmahoney at unomaha dot edu
CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20061

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[uweigand at gcc dot gnu dot org]

--- Additional Comments From uweigand at gcc dot gnu dot org  2005-02-18 
17:42 ---
I think the middle-end *should* allow back-ends to use SUBREGs
in that way.  cleanup_subreg_operands should really act only on
SUBREGs that are part of an operand, and not on SUBREGs that 
are part of the invariant insn pattern.

In any case, I'll try to reimplement the pattern in a different way.
The problem is that it needs to have a form that combine will match ...

-- 
   What|Removed |Added

 AssignedTo|unassigned at gcc dot gnu   |uweigand at gcc dot gnu dot
   |dot org |org
 Status|UNCONFIRMED |ASSIGNED
 Ever Confirmed||1
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 17:42:25
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug tree-optimization/19786] [4.0 Regression] Aliasing optimisation bug

[aoliva at gcc dot gnu dot org]

--- Additional Comments From aoliva at gcc dot gnu dot org  2005-02-18 
17:45 ---
I can't duplicate this with a tree updated some time earlier today.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19786

[Bug libstdc++/10606] uncaught_exception() returns false too early

[david dot moore at intel dot com]

--- Additional Comments From david dot moore at intel dot com  2005-02-18 
17:56 ---
As the person who has been working on similar problems in the Intel Compiler, 
Martin and I had some lengthy discussions about this during which he convinced 
me he was correct.

The result was issue 475 in the active C++ core issues list. Its now clear that 
uncaught_exception should be on during the copy construction of temporary that 
is thrown. 

Thinking about the user model, it seems that this is the only model that makes 
sense.  I believe the standard clarifications in this issue will be made but 
that in any case, the issue is clear enough at this point to proceed.

M2CW. David Moore


-- 
   What|Removed |Added

 CC||david dot moore at intel dot
   ||com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10606

[Bug c++/20062] New: hash_map leak memory

[gcc-bugzilla at gcc dot gnu dot org]

hash_map implementation of g++ extensions of STL leaks memory

Environment:
System: Linux rose 2.6.8-2004 #2 SMP Wed Nov 24 11:15:13 CET 2004 i686 
GNU/Linux
Architecture: i686


host: i486-pc-linux-gnu
build: i486-pc-linux-gnu
target: i486-pc-linux-gnu
configured with: ../src/configure -v 
--enable-languages=c,c++,java,f77,pascal,objc,ada,treelang --prefix=/usr 
--mandir=/usr/share/man --infodir=/usr/share/info 
--with-gxx-include-dir=/usr/include/c++/3.3 --enable-shared --with-system-zlib 
--enable-nls --without-included-gettext --enable-__cxa_atexit 
--enable-clocale=gnu --enable-debug --enable-java-gc=boehm 
--enable-java-awt=xlib --enable-objc-gc i486-linux

How-To-Repeat:
Compile the code, rerun command, I included in the description
--- Additional Comments From arturas at inf dot unibz dot it  2005-02-18 
18:10 ---
Fix:
I have no idea.

-- 
   Summary: hash_map leak memory
   Product: gcc
   Version: 3.3.3
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: arturas at inf dot unibz dot it
CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i486-pc-linux-gnu
  GCC host triplet: i486-pc-linux-gnu
GCC target triplet: i486-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug libstdc++/20062] hash_map leak memory

[pinskia at gcc dot gnu dot org]

-- 
   What|Removed |Added

  Component|c++ |libstdc++


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug libstdc++/20062] hash_map leak memory

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
18:18 ---
(In reply to comment #0)
> Compile the code, rerun command, I included in the description

I don't see any.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug libstdc++/20062] hash_map leak memory

[arturas at inf dot unibz dot it]

--- Additional Comments From arturas at inf dot unibz dot it  2005-02-18 
18:25 ---
Subject: Re:  hash_map leak memory

Hi Pinskia,

I think I made a mistake including the code.

Here's the code:

- a.cc -
#include 
#include 

#undef __DEPRECATED
#include 

class Eqstr {
  public:
  bool operator()(const char* s1, const char* s2) const {

  if (strcmp(s1, s2) == 0)
return true;
  else
return false;
  }
};

using std::vector;
using std::cout;
using std::endl;

main () {
  hash_map, Eqstr> signatures;
  hash_map, Eqstr>::iterator it;

  char *str;
  int *pi, j;
  char qgram[4];
  qgram[0] = qgram[1] = qgram[2] = 'a';
  qgram[3] = 0;
  vector< char * > p;

  while (strcmp( qgram, "zzz" ) != 0) {
str = new char [4];
strcpy( str, qgram );
pi = new int [1];

signatures[str] = pi;
//scanning thourgh all strings from aaa to zzz
if (qgram[2] != 'z')
  qgram[2]++;
else {
  qgram[2] = 'a';
  if (qgram[1] != 'z')
qgram[1]++;
  else {
qgram[1] = 'a';
qgram[0]++;
  }
}
  }

  for (it = signatures.begin(); it != signatures.end(); it++) {
p.push_back( (char *)it->first );
delete[] it->second;
  }

  for (j = 0; j < p.size(); j++)
delete[] (p[j]);
}
-

memory leak was obtain in the following way:

$ g++ -g a.cc
$ LeakTracer/LeakCheck ./a.out
$ LeakCheck ./a.out # (From http://www.andreasen.org/LeakTracer/)
$ [EMAIL PROTECTED]:/tmp$ ~/edit/LeakTracer/leak-analyze ./a.out
Gathered 56 (1 unique) points of data.
(gdb) 
#-- Leak: counted 56x / total Size: 296360
  No source file for address 0x400a9a5b.

I've played a little with __builtin_return_address to retrieve the place
of the leak. It seems that it is line 127 of stl_vector.h file.

Should I also upload the file?

Thanks,
arturas


On Fri, Feb 18, 2005 at 06:18:01PM -, pinskia at gcc dot gnu dot org wrote:
> 
> --- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
> 18:18 ---
> (In reply to comment #0)
> > Compile the code, rerun command, I included in the description
> 
> I don't see any.
> 
> -- 
> 
> 
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062
> 
> --- You are receiving this mail because: ---
> You reported the bug, or are watching the reporter.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug c++/20060] Ambiguous MSG in array initialization when expression is not integral constant.

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
18:28 ---
Fixed in 3.4.0 and above:
t.cc: In function `int mymain()':
t.cc:4: error: `n' cannot appear in a constant-expression


-- 
   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
   Keywords||diagnostic
 Resolution||FIXED
   Target Milestone|--- |3.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20060

[Bug libstdc++/20062] hash_map leak memory

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
18:31 ---
Can you read  
and make sure 
that this is not the documented "leak" (I put leak in quotes because it is not 
a leak really).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

Re: configuration problem using --enable-intermodule on x86_64-unknown-linux-gnu

[Andrew Pinski]

On Feb 18, 2005, at 11:50 AM, Kenneth Zadeck wrote:
The build gets partially thru the stage2 and then crashes when trying 
to compile most of the back end in one step with the following 
message:

gcc: cannot specify -o with -c or -S and multiple compilations
This same step works fine on my pentium (suse 9.2) and darwin 
machines.  I am compiling against the gcc/mainline. My x86_64 machine 
is running suse 9.2 with all the current patches installed.

my configuration command was:
../lorien/gccBaseline/configure --enable-intermodule
Yes this was reported a while back by Jan.  Basically multilib is 
getting the way
some how.

-- Pinski

[Bug middle-end/20061] Certain flags generate warnings when they should not.

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
18:38 ---
Confirmed. Note this flag only turns on the dumping for the tree based loop 
optimizators and not for 
the RTL based ones.

-- 
   What|Removed |Added

   Severity|enhancement |minor
 Status|UNCONFIRMED |NEW
  Component|c   |middle-end
 Ever Confirmed||1
   Keywords||diagnostic
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 18:38:09
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20061

[Bug fortran/20059] internal compiler error: Segmentation Fault - For common blocks

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
18:44 ---
Confirmed, but note I think this is the same as PR 19387 even though there is 
no ICE in that bug.

-- 
   What|Removed |Added

  BugsThisDependsOn||19387
 Status|UNCONFIRMED |NEW
 Ever Confirmed||1
   Keywords||ice-on-valid-code
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 18:44:20
   date||
Summary|internal compiler error:|internal compiler error:
   |Segmentation Fault - For|Segmentation Fault - For
   |common blocks   |common blocks


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20059

[Bug libstdc++/20062] hash_map leak memory

[pcarlini at suse dot de]

--- Additional Comments From pcarlini at suse dot de  2005-02-18 18:51 
---
With valgrind, I cannot reproduce the problem on 3.3.3:

==32676== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 17 from 1)
==32676== malloc/free: in use at exit: 296360 bytes in 56 blocks.
==32676== malloc/free: 35224 allocs, 35168 frees, 894872 bytes allocated.
==32676== For counts of detected errors, rerun with: -v
==32676== searching for pointers to 56 not-freed blocks.
==32676== checked 2563516 bytes.
==32676==
==32676== LEAK SUMMARY:
==32676==definitely lost: 0 bytes in 0 blocks.
==32676==possibly lost:   0 bytes in 0 blocks.
==32676==still reachable: 296360 bytes in 56 blocks.
==32676== suppressed: 0 bytes in 0 blocks.
==32676== Reachable blocks (those to which a pointer was found) are not shown.

(N.B: usual spurious reachable due to the pool allocator)

neither on 3.4.3:

==32669== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 17 from 1)
==32669== malloc/free: in use at exit: 0 bytes in 0 blocks.
==32669== malloc/free: 52749 allocs, 52749 frees, 809664 bytes allocated.
==32669== For counts of detected errors, rerun with: -v
==32669== No malloc'd blocks -- no leaks are possible.

In any case, even if hash_map leaks (I don't think so)... see libstdc++/19554.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[uweigand at gcc dot gnu dot org]

--- Additional Comments From uweigand at gcc dot gnu dot org  2005-02-18 
18:45 ---
OK, now I see the problem: cleanup_subreg_operands *does* indeed
touch only operands.  However, the result of the splitter matches
the *llgt_didi insn instead of *llgt_disi, because after reload
nonimmediate_operands matches a SUBREG (MEM).  Thus the SUBREG is
indeed part of the operand, which is broken.

The proper fix is to move the *llgt_disi pattern to *before* the
*llgt_didi pattern to ensure it matches first.

I'll commit this fix once testing completes.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug ada/19140] ACATS c37402a segfaults at runtime

[uweigand at gcc dot gnu dot org]

--- Additional Comments From uweigand at gcc dot gnu dot org  2005-02-18 
19:01 ---
This test case fails on s390(x) as well; in fact it is quite annoying
because it causes the whole test suite to hang ...

The reason for this appears to be that the Ada runtime has installed
a segfault handler which attempts to take some lock already held by
the code that segfaulted.  As the Ada test suite doesn't have timeout
handlers, this blocks the whole test suite :-(


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19140

[Bug tree-optimization/19952] ICE: tree check: expected class 'declaration', have 'statement' (label_expr) in tree_verify_flow_info, at tree-cfg.c:3709

[dpatel at apple dot com]

--- Additional Comments From dpatel at apple dot com  2005-02-18 19:02 
---
Subject: Re:  ICE: tree check: expected class 'declaration', have 'statement' 
(label_expr) in tree_verify_flow_info, at tree-cfg.c:3709


ok
-
Devang



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19952

[Bug libstdc++/20062] hash_map leak memory

[arturas at inf dot unibz dot it]

--- Additional Comments From arturas at inf dot unibz dot it  2005-02-18 
19:04 ---
Subject: Re:  hash_map leak memory

Hi Andrew,

Thanks for the reply. I think you are correct. I overlooked that
information completely. It seems there's no memory leak.

I apologies for taking your time.

arturas

On Fri, Feb 18, 2005 at 06:51:04PM -, pcarlini at suse dot de wrote:
> 
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug libstdc++/20062] hash_map leak memory

[pcarlini at suse dot de]

--- Additional Comments From pcarlini at suse dot de  2005-02-18 19:06 
---
Agreed...

-- 
   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution||INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20062

[Bug c++/19508] [4.0 regression] dwarf2, ICE on __attribute__(aligned) in class template

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
19:12 ---
Subject: Bug 19508

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 19:11:58

Modified files:
gcc/cp : ChangeLog decl2.c 
gcc/testsuite  : ChangeLog 
Added files:
gcc/testsuite/g++.dg/ext: attrib20.C 

Log message:
2005-01-20  Giovanni Bajo  <[EMAIL PROTECTED]>

PR c++/19508
* g++.dg/ext/attrib20.C: New test.

2005-01-20  Giovanni Bajo  <[EMAIL PROTECTED]>

PR c++/19508
* decl2.c (grokfield): Do not apply attributes to template parameters
as they are ignored by tsubst anyway.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4634&r2=1.4635
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/decl2.c.diff?cvsroot=gcc&r1=1.766&r2=1.767
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5048&r2=1.5049
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/ext/attrib20.C.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19508

[Bug c++/19508] [4.0 regression] dwarf2, ICE on __attribute__(aligned) in class template

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
19:14 ---
The ICE itself is fixed and the rest of the bug is referenced in PR 17743 so 
closing as fixed.

-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19508

[Bug ada/18659] [4.0 Regression] ACATS ICE in bug in tree-sra.c:1507 on x86 and ppc-darwin (6 FAIL)

[ebotcazou at gcc dot gnu dot org]

--- Additional Comments From ebotcazou at gcc dot gnu dot org  2005-02-18 
19:20 ---
> > SRA is an Ada killer because of the the abuse of VIEW_CONVERT_EXPRs by the 
> > Ada
> > front-end.  I think it should be disabled in the Ada compiler for 4.0.x.
> 
> Sounds reasonable to me.

OK, I'll submit a patch once 4.0.x is branched.  As of today, excluding the
problematic cxa3004 and cxb3006, as well as cxa4006 and cxa4017 for which I've
submitted a fix, we have 6 failures at -O2 -fno-tree-sra.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18659

[Bug ada/19959] [4.0 Regression] Can't compile gnattools for the cross targets

[berndtrog at yahoo dot com]

--- Additional Comments From berndtrog at yahoo dot com  2005-02-18 20:22 
---
Fails for --target=i386-elf too.

-- 
   What|Removed |Added

 GCC target triplet|avr |
Summary|[4.0 Regression] Can't  |[4.0 Regression] Can't
   |compile gnattools for the   |compile gnattools for the
   |AVR target  |cross targets


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19959

[Bug target/19830] cris-elf testsuite failure: gcc.c-torture/execute/920501-8.c execute tests.

[hp at gcc dot gnu dot org]

--- Additional Comments From hp at gcc dot gnu dot org  2005-02-18 20:25 
---
Upon closer examination, the miscompiled vfprintf.c is missing a compare insn.
This seems to have been the same bug as PR 19898.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19830

[Bug target/19636] Can't compile ethernut OS (avr-gcc)

[berndtrog at yahoo dot com]

-- 
   What|Removed |Added

 CC||berndtrog at yahoo dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19636

Re: configuration problem using --enable-intermodule on x86_64-unknown-linux-gnu

[Kenneth Zadeck]

Thanks, for looking into it.
kenny
Andrew Pinski wrote:
On Feb 18, 2005, at 11:50 AM, Kenneth Zadeck wrote:
The build gets partially thru the stage2 and then crashes when trying 
to compile most of the back end in one step with the following message:

gcc: cannot specify -o with -c or -S and multiple compilations
This same step works fine on my pentium (suse 9.2) and darwin 
machines.  I am compiling against the gcc/mainline. My x86_64 machine 
is running suse 9.2 with all the current patches installed.

my configuration command was:
../lorien/gccBaseline/configure --enable-intermodule

Yes this was reported a while back by Jan.  Basically multilib is 
getting the way
some how.

-- Pinski

[Bug target/19830] cris-elf testsuite failure: gcc.c-torture/execute/920501-8.c execute tests.

[hp at gcc dot gnu dot org]

--- Additional Comments From hp at gcc dot gnu dot org  2005-02-18 20:28 
---
.dbr:

(insn:QI 852 851 853
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1310 (set
(cc0)
(compare (reg/v/f:SI 2 r2 [orig:130 bp ] [130])
(reg/f:SI 9 r9 [orig:738 bp ] [738]))) 9 {cmpsi} (nil)
(nil))

(jump_insn 853 852 5371
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1310 (set
(pc)
(if_then_else (leu (cc0)
(const_int 0 [0x0]))
(label_ref 854)
(pc))) 159 {bleu} (nil)
(expr_list:REG_BR_PRED (const_int 18 [0x12])
(expr_list:REG_BR_PROB (const_int 1100 [0x44c])
(nil

(note 5371 853 5364 [bb 434] NOTE_INSN_BASIC_BLOCK)

(note 5364 5371 5366
("/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c") 1311)

(insn 5366 5364 5367
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1311 (set
(mem:QI (post_inc:SI (reg/f:SI \
9 r9 [orig:738 bp ] [738])) [0 S1 A8])
(reg:QI 11 r11)) 41 {movqi} (nil)
(expr_list:REG_INC (reg/f:SI 9 r9 [orig:738 bp ] [738])
(nil)))

(insn 5367 5366 5368
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1311 (set
(mem/i:SI (plus:SI (reg/f:SI 14\
 sp)
(const_int 1687 [0x697])) [19 rve+0 S4 A8])
(reg/f:SI 9 r9 [orig:738 bp ] [738])) 32 {*movsi_internal} (nil)
(nil))

(note 5368 5367 5369
("/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c") 1310)

(insn 5369 5368 5370
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1310 (set
(cc0)
(compare (reg/v/f:SI 2 r2 [orig:130 bp ] [130])
(reg/f:SI 9 r9 [orig:738 bp ] [738]))) 9 {cmpsi} (nil)
(nil))

(jump_insn 5370 5369 5737
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1310 (set
(pc)
(if_then_else (gtu (cc0)
(const_int 0 [0x0]))
(label_ref 8646)
(pc))) 153 {bgtu} (nil)
(expr_list:REG_BR_PRED (const_int 4 [0x4])
(expr_list:REG_BR_PROB (const_int 8900 [0x22c4])
(nil

Yet:
cmp.d $r9,$r2   ;# 852  cmpsi/2
bls .L128
nop ;# 853  bleu
;# basic block 434
;#
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1311
.loc 1 1311 0
move.b $r11,[$r9+]  ;# 5366 movqi/2
move.d $r9,[$sp+1687]   ;# 5367 *movsi_internal/11
;#
/home/hp/cvs_areas/combined/cvs_write/newlib/libc/stdio/vfprintf.c:1310
.loc 1 1310 0
bhi .L612
nop ;# 5370 bgtu

Which concludes the analysis.  Case closed.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19830

[Bug target/19830] cris-elf testsuite failure: gcc.c-torture/execute/920501-8.c execute tests.

[hp at gcc dot gnu dot org]

--- Additional Comments From hp at gcc dot gnu dot org  2005-02-18 20:32 
---
One final note: this code is at the point in the simulator trace where the
number of digits is calculated and first seen to be different between the faulty
code and working code (that is, working due to unrelated rearrangement).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19830

[Bug tree-optimization/19786] [4.0 Regression] Aliasing optimisation bug

[sylvain dot pion at sophia dot inria dot fr]

--- Additional Comments From sylvain dot pion at sophia dot inria dot fr  
2005-02-18 20:48 ---
(In reply to comment #13)
> I can't duplicate this with a tree updated some time earlier today.

It has also disappeared for me on Fedora Core 3.
But I can still reproduce it on Fedora Core 1,
with a g++ which I have just built (20050218).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19786

[Bug fortran/20063] New: gfortran assembler error - : Bad Absolute Expression, absolute 0 assumed.

[dir at lanl dot gov]

I get this error with the Macintosh gfortran from the 20050213 package with some
of my programs.

[dranta:~/tests/gfortran] dir% gfortran -c block.f
/var/tmp//ccu5xNom.s:28:Bad Absolute Expression, absolute 0 assumed.
/var/tmp//ccu5xNom.s:28:Repeat < 0, .space ignored
/var/tmp//ccu5xNom.s:88:Bad Absolute Expression, absolute 0 assumed.
/var/tmp//ccu5xNom.s:88:Repeat < 0, .space ignored
[dranta:~/tests/gfortran] dir% cat  block.f
  block data blkdat
  character*8 names vax
  common/vfile/names(60)vax
  data names / 'rstxyz', 'disk05', 'disk06', 'thrxyz', '  ',noto
 1 '  ', 'disk01', 'disk02', 'n3plot', 'n3dump',noto
 2 'disk11', 'disk12', 'strxyz', 'n3runr', 'itrcnt',noto
 3 '  ', '  ', 'xyz123', 28*'  ',   noto
 4 'n3dhsp','n3data', 10*'  ', 'tty', '  '/ noto
  end

-- 
   Summary: gfortran assembler error - : Bad Absolute Expression,
absolute 0 assumed.
   Product: gcc
   Version: 4.0.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: dir at lanl dot gov
CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: powerpc-apple-darwin7.8.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20063

[Bug tree-optimization/18977] [4.0 regression] LAPACK test xeigtsts segfaults with optimization

[wilson at gcc dot gnu dot org]

--- Additional Comments From wilson at gcc dot gnu dot org  2005-02-18 
20:53 ---
This is a tree-optimization problem not a target problem.

-- 
   What|Removed |Added

 AssignedTo|unassigned at gcc dot gnu   |wilson at gcc dot gnu dot
   |dot org |org
 Status|NEW |ASSIGNED
  Component|target  |tree-optimization
   Last reconfirmed|2005-02-12 04:00:42 |2005-02-18 20:53:32
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18977

[Bug fortran/20063] gfortran assembler error - : Bad Absolute Expression, absolute 0 assumed.

[sgk at troutmask dot apl dot washington dot edu]

--- Additional Comments From sgk at troutmask dot apl dot washington dot 
edu  2005-02-18 21:01 ---
Your code compiles fine on amd64-*-freebsd where I'm using a gfortran
form 20050218 sources.  I do get a warning though

troutmask:kargl[205] /usr/tmp/sgk/bin/gfc -c e.f
/var/tmp/ccvV0Vua.s: Assembler messages:
/var/tmp/ccvV0Vua.s:39: Warning: .space repeat count is zero, ignored
/var/tmp/ccvV0Vua.s:99: Warning: .space repeat count is zero, ignored

I suspect that this a target specific bug or it could be related to
a serious bug introduced on 2/13 and fix last night.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20063

[Bug fortran/20063] gfortran assembler error, n*'xxxx' is broken

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
21:06 ---
Confirmed, reduced to:
  block data blkdat
  character*8 names
  common/vfile/names(3)
  data names / 'a',2*''/
  end

-- 
   What|Removed |Added

 Status|UNCONFIRMED |NEW
 Ever Confirmed||1
   Keywords||wrong-code
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 21:06:33
   date||
Summary|gfortran assembler error - :|gfortran assembler error,
   |Bad Absolute Expression,|n*'' is broken
   |absolute 0 assumed. |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20063

[Bug fortran/20063] gfortran assembler error, n*'xxxx' is broken

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
21:07 ---
(In reply to comment #1)
> Your code compiles fine on amd64-*-freebsd where I'm using a gfortran
> form 20050218 sources.  I do get a warning though
Darwin's as treats those warnings as errors as they really should as this is a 
bug
in the front-end.
If you take a look at my reduced testcase and look at the assembly ouput you 
will see that
the assembly outputed is wrong.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20063

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
21:11 ---
Subject: Bug 20054

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 21:11:07

Modified files:
gcc: ChangeLog 
gcc/config/s390: s390.md 
gcc/testsuite  : ChangeLog 
Added files:
gcc/testsuite/gcc.dg: pr20054.c 

Log message:
ChangeLog:

PR target/20054
* config/s390/s390.md ("*llgt_sidi", "*llgt_sidi_split"): Move to
before the "*llgt_didi" pattern.

testsuite/ChangeLog:

PR target/20054
* gcc.dg/pr20054.c: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.7526&r2=2.7527
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/s390/s390.md.diff?cvsroot=gcc&r1=1.153&r2=1.154
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5049&r2=1.5050
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/pr20054.c.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[uweigand at gcc dot gnu dot org]

--- Additional Comments From uweigand at gcc dot gnu dot org  2005-02-18 
21:15 ---
Fixed.

-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug target/20054] [4.0 Regression] ICE in change_address_1

[uweigand at gcc dot gnu dot org]

-- 
   What|Removed |Added

   Target Milestone|--- |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20054

[Bug target/19886] [4.0 Regression] g++.dg/eh/registers1.C execution test fails

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
21:23 ---
Subject: Bug 19886

CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED]   2005-02-18 21:23:16

Modified files:
gcc: ChangeLog 
gcc/config/ia64: ia64.c ia64.h 

Log message:
PR target/19886
* config/ia64/ia64.h (struct machine_function): Add state_num.
* config/ia64/ia64.c (process_epilogue,
process_for_unwind_directive): Use new unwind state numbers each
time rather than state 1.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.7527&r2=2.7528
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/ia64/ia64.c.diff?cvsroot=gcc&r1=1.345&r2=1.346
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/ia64/ia64.h.diff?cvsroot=gcc&r1=1.191&r2=1.192



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19886

[Bug c++/20064] New: Types shall not be defined in casts in the strict mode.

[msadoghi at ca dot ibm dot com]

Sourc code for t.cpp:

int mymain()
{
void *p = (void *)(struct {int i;} *)0; /* error - can't define a type 
in a cast */
return 0;
}






Expected Behaviour:

t.cpp:3: error: ISO C++ forbids defining types within cast




Actual Behaviour:

t.cpp:3: warning: ISO C++ forbids defining types within cast




t.ii generated with -save-temps option

# 1 "t.cpp"
# 1 "t.cpp"
# 1 ""
# 1 ""
# 1 "t.cpp"
int mymain()
{
void *p = (void *)(struct {int i;} *)0;
return 0;
}






Release:

GCC Version: 3.2.0





Environment:
System Type:

Reading specs from /usr/libexec/gcc/darwin/ppc/3.3/specs
Thread model: posix
gcc version 3.3 20030304 (Apple Computer, Inc. build 1495)
 /usr/libexec/gcc/darwin/ppc/3.3/cc1plus -E -D__GNUG__=3 -quiet -v -D__GNUC__=3
-D__GNUC_MINOR__=3 -D__GNUC_PATCHLEVEL__=0 -D__APPLE_CC__=1495 -D__DYNAMIC__
/home/msadoghi/scenario/common/t.cpp -fPIC -std=iso9899:199409 -pedantic
-D__private_extern__=extern t.ii
cc1plus: warning: "-std=iso9899:199409" is valid for C/ObjC but not for C++
ignoring nonexistent directory "/usr/ppc-darwin/include"
ignoring nonexistent directory "/Local/Library/Frameworks"
#include "..." search starts here:
#include <...> search starts here:
 /usr/include/gcc/darwin/3.3/c++
 /usr/include/gcc/darwin/3.3/c++/ppc-darwin
 /usr/include/gcc/darwin/3.3/c++/backward
 /usr/local/include
 /usr/include/gcc/darwin/3.3
 /usr/include
End of search list.
Framework search starts here:
 /System/Library/Frameworks
 /Library/Frameworks
End of framework search list.
 /usr/libexec/gcc/darwin/ppc/3.3/cc1plus -fpreprocessed t.ii -fPIC -quiet
-dumpbase t.cpp -auxbase t -pedantic -std=iso9899:199409 -version
-D__private_extern__=extern -o t.s
cc1plus: warning: "-std=iso9899:199409" is valid for C/ObjC but not for C++
GNU C++ version 3.3 20030304 (Apple Computer, Inc. build 1495) (ppc-darwin)
compiled by GNU C version 3.3 20030304 (Apple Computer, Inc. build 
1495).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=131072
/home/msadoghi/scenario/common/t.cpp: In function `int mymain()':
/home/msadoghi/scenario/common/t.cpp:3: warning: ISO C++ forbids defining types
   within cast
 /usr/libexec/gcc/darwin/ppc/as -arch ppc -o t.o t.s





How-To-Repeat:

g++ -v -save-temps -c -std=iso9899:199409 -pedantic t.cpp



Note:

This is a diagnostic test case. Compilation failure is expected on the strict 
mode,
where type is defined in a cast.

C++ ANSI Standard:
"Types shall not be defined in casts in the strict mode."


AIX, MACOS and LINUX all have a similar behaviors.


-- 
   Summary: Types shall not be defined in casts in the strict mode.
   Product: gcc
   Version: 3.2
Status: UNCONFIRMED
  Severity: normal
  Priority: P2
 Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: msadoghi at ca dot ibm dot com
CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20064

[Bug tree-optimization/18178] Missed opportunity for removing bounds checking

[cvs-commit at gcc dot gnu dot org]

--- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-18 
21:29 ---
Subject: Bug 18178

CVSROOT:/cvs/gcc
Module name:gcc
Branch: tree-cleanup-branch
Changes by: [EMAIL PROTECTED]   2005-02-18 21:28:58

Modified files:
gcc: ChangeLog.tcb Makefile.in tree-into-ssa.c 
 tree-vrp.c 
gcc/testsuite  : ChangeLog.tcb 
gcc/testsuite/gcc.dg/tree-ssa: ssa-pre-2.c ssa-pre-3.c 
   ssa-pre-4.c ssa-pre-5.c 
   ssa-pre-6.c 
Added files:
gcc/testsuite/g++.dg/tree-ssa: pr18178.C 

Log message:
PR tree-optimization/18178
* Makefile.in (tree-vrp.o): Depend on CFGLOOP_H,
tree-scalar-evolution.h and tree-chrec.h.
* tree-into-ssa.c (prepare_block_for_update): Also rewrite
operands of statements that define new names.
* tree-vrp.c: Include cfgloop.h, tree-scalar-evolution.h and
tree-chrec.h.
(cfg_loops): New local variable.
(compare_values): Forward declare.
(copy_value_range): Remove.
(set_value_range): Add range integrity checks.
Decay to VR_VARYING ranges that take all possible values in
the type domain.
(compare_values): Do some symbolic comparisons.
(value_inside_range): Move earlier in the file.
(value_ranges_intersect_p): Likewise.
(extract_range_from_assert): If the ASSERT_EXPR conditional
and the variable have intersecting ranges, use the
intersection to derive a narrower range.
(extract_range_from_ssa_name): New function.
(extract_range_from_binary_expr): Re-arrange to always call
set_value_range to set the newly computed range.
(extract_range_from_unary_expr): Likewise.
Do not special case ABS_EXPR.
If a type cast operation changes to a type of different size,
set the resulting range to VR_VARYING.
If the new range has the limits swapped around, set the result
to VR_VARYING.
(extract_range_from_expr): Call extract_range_from_ssa_name.
(compare_ranges): Allow symbolic ranges.
Fix calls to compare_values to always check for specific
return values.
(compare_range_with_value): Likewise.
(adjust_range_with_scev): New function.
(vrp_visit_assignment): Call it if the statement is inside a
loop.
(vrp_meet): Always call set_value_range to set the newly
computed range.
(vrp_visit_phi_node): Remove FIXME regarding derivation.
(execute_vrp): Call loop_optimizer_init, scev_initialize,
scev_finalize and loop_optimizer_finalize.

testsuite/ChangeLog.tcb:

* testsuite/gcc.dg/tree-ssa/ssa-pre-2.c: Adjust expected pattern.
* testsuite/gcc.dg/tree-ssa/ssa-pre-3.c: Likewise.
* testsuite/gcc.dg/tree-ssa/ssa-pre-4.c: Likewise.
* testsuite/gcc.dg/tree-ssa/ssa-pre-5.c: Likewise.
* testsuite/gcc.dg/tree-ssa/ssa-pre-6.c: Likewise.

PR tree-optimization/18178
* testsuite/g++.dg/tree-ssa/pr18178.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.tcb.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.2.23&r2=1.1.2.24
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/Makefile.in.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1396.2.18&r2=1.1396.2.19
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-into-ssa.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=2.21.2.12&r2=2.21.2.13
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-vrp.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.2.4&r2=1.1.2.5
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.tcb.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.2.6&r2=1.1.2.7
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/tree-ssa/pr18178.C.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=NONE&r2=1.1.2.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/ssa-pre-2.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.3&r2=1.3.10.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/ssa-pre-3.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.10.1&r2=1.1.10.2
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/ssa-pre-4.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.2.1&r2=1.1.2.2
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/ssa-pre-5.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.2.1&r2=1.1.2.2
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/ssa-pre-6.c.diff?cvsroot=gcc&only_with_tag=tree-cleanup-branch&r1=1.1.2.1&r2=1.1.2.2



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178

[Bug c++/20064] Types shall not be defined in casts in the strict mode.

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
21:29 ---
There are two things here, -pedantic does not enable strict mode, just the 
warnings for strict mode, 
use -pedantic-errors instead.

Anyways this is rejected by default with 3.4.0 and above.

-- 
   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution||FIXED
   Target Milestone|--- |3.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20064

[Bug fortran/20063] n*'xxxx' disables padding

[tobi at gcc dot gnu dot org]

--- Additional Comments From tobi at gcc dot gnu dot org  2005-02-18 21:35 
---
>From the generated assembly
.ascii  "a   "
.zero   4294967292
.ascii  ""
.zero   4
.ascii  ""
.zero   4
.zero   4
i.e. the elements that are filled in with a repeat count are not padded to the
correct length.

-- 
   What|Removed |Added

Summary|gfortran assembler error,   |n*'' disables padding
   |n*'' is broken  |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20063

[Bug fortran/20063] n*'xxxx' disables padding

[tobi at gcc dot gnu dot org]

--- Additional Comments From tobi at gcc dot gnu dot org  2005-02-18 21:37 
---
Well, more precisely: they're zero-padded instead of being padded with blanks.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20063

[Bug tree-optimization/18178] Missed opportunity for removing bounds checking

[dnovillo at gcc dot gnu dot org]

--- Additional Comments From dnovillo at gcc dot gnu dot org  2005-02-18 
21:44 ---

Fix.  http://gcc.gnu.org/ml/gcc-patches/2005-02/msg01074.html.

The patch allows VRP to eliminate 42% more conditionals on cc1-i-files and 173%
more conditionals on DLV.  But produces little difference in MICO and TRAMP3D.  

-- 
   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178

[Bug tree-optimization/19659] GCC does not remove an "if" statement that never triggers.

[dnovillo at gcc dot gnu dot org]

-- 
Bug 19659 depends on bug 18178, which changed state.

Bug 18178 Summary: Missed opportunity for removing bounds checking
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178

   What|Old Value   |New Value

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19659

[Bug tree-optimization/18178] Missed opportunity for removing bounds checking

[schlie at comcast dot net]

--- Additional Comments From schlie at comcast dot net  2005-02-18 21:57 
---
(In reply to comment #6)
> Fix.  http://gcc.gnu.org/ml/gcc-patches/2005-02/msg01074.html.

For 4.0 ?



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178

[Bug java/20056] [4.0 regression] 'verification failed: incompatible type on stack' with --indirect-dispatch

[tromey at gcc dot gnu dot org]

--- Additional Comments From tromey at gcc dot gnu dot org  2005-02-18 
22:00 ---
I'm looking at this


-- 
   What|Removed |Added

 AssignedTo|unassigned at gcc dot gnu   |tromey at gcc dot gnu dot
   |dot org |org
 Status|UNCONFIRMED |ASSIGNED
 Ever Confirmed||1
   Last reconfirmed|-00-00 00:00:00 |2005-02-18 22:00:32
   date||


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20056

[Bug tree-optimization/18178] Missed opportunity for removing bounds checking

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
22:00 ---
(In reply to comment #7)
> (In reply to comment #6)
> > Fix.  http://gcc.gnu.org/ml/gcc-patches/2005-02/msg01074.html.
> 
> For 4.0 ?

No, but for 4.1, reopening as there is no 4.1 right now.

-- 
   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178

[Bug tree-optimization/19659] GCC does not remove an "if" statement that never triggers.

[pinskia at gcc dot gnu dot org]

-- 
Bug 19659 depends on bug 18178, which changed state.

Bug 18178 Summary: Missed opportunity for removing bounds checking
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178

   What|Old Value   |New Value

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19659

[Bug tree-optimization/18178] Missed opportunity for removing bounds checking

[pinskia at gcc dot gnu dot org]

--- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-18 
22:01 ---
Suspending as fixed and moving the target milestone to 4.1.0 so we know to 
retest it when tcb gets 
merged in.

-- 
   What|Removed |Added

 Status|REOPENED|SUSPENDED
   Target Milestone|--- |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18178