[Bug c++/18384] [3.3/3.4/4.0 Regression] ICE on zero-length array with empty initializer...

--- Additional Comments From mark at codesourcery dot com 2004-11-18 07:59 --- Subject: Re: [3.3/3.4/4.0 Regression] ICE on zero-length array with empty initializer... giovannibajo at libero dot it wrote: >--- Additional Comments From giovannibajo at libero dot it 2004-11-18 >0

[Bug fortran/18537] no warning about tabs with std=f95 option

-- What|Removed |Added Keywords||diagnostic http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18537

[Bug tree-optimization/18308] ICE in do_jump, at dojump.c:274

--- Additional Comments From dpatel at apple dot com 2004-11-18 02:33 --- Subject: Re: ICE in do_jump, at dojump.c:274 Andrew, You can try following to fix tree level if-conversion. I have not tested it completely yet. - Devang Index: tree-if-conv.c

[Bug fortran/18537] New: no warning about tabs with std=f95 option

Tabs in source code are not standard Fortran but are accepted by many compilers. Gfortran should at least warn about the use of tabs when invoked with the -std=f95 option. The Compaq Visual Fortran and Intel compilers do so. -- Summary: no warning about tabs with std=f95 option

[Bug tree-optimization/18308] ICE in do_jump, at dojump.c:274

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-18 02:31 --- And that would mean it was caused by: * dojump.c (do_jump) : Abort on gimplified cases. -- What|Removed |Added --

[Bug tree-optimization/18308] ICE in do_jump, at dojump.c:274

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-18 02:29 --- So someone removed too much from do_jump in the first place. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18308

[Bug tree-optimization/18308] ICE in do_jump, at dojump.c:274

--- Additional Comments From dpatel at apple dot com 2004-11-18 02:26 --- After I update tree-level if-conversion to force gimple operands appropriately, rewrite_out_of_ssa() is converting following ... bar()

[Bug tree-optimization/18536] ICE: in emit_move_insn, at expr.c:2590 with -ftree-vectorize -msse2

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-18 01:52 --- Confirmed, the problem I think is the same as the PPC64 problem in PR 18403 but I did not check the patch which will fix that one for sure. Breakpoint 1, fancy_abort (file=0x81fcb4 "/Users/pinskia/src/

[Bug c++/18384] [3.3/3.4/4.0 Regression] ICE on zero-length array with empty initializer...

--- Additional Comments From giovannibajo at libero dot it 2004-11-18 01:44 --- Mark, I guess this is the same problem of PR 18327. Will your fix take care of both, or should I look into PR 18327? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18384

[Bug rtl-optimization/17950] Over Aggressive Use of Data Cache Touch Instructions During FDO

--- Additional Comments From jgrimm2 at us dot ibm dot com 2004-11-18 01:32 --- Hmmm... looks like a couple things (maybe more) come in to play here. First, value-profile code doesn't seem to merge prefetches (among other things). That is, it should be pretty easy (crude?) to track the

[Bug tree-optimization/18536] New: ICE: in emit_move_insn, at expr.c:2590 with -ftree-vectorize -msse2

/tmp/gcc4/libexec/gcc/i686-pc-linux-gnu/4.0.0/cc1 -quiet test.c -quiet -dumpbase test.c -msse2 -auxbase test -O -ftree-vectorize -o test.s -version GNU C version 4.0.0 20041117 (experimental) (i686-pc-linux-gnu

[Bug middle-end/18535] New: fix_irreducible_loops could be improved

If get_loop_exit_edges would return a VEC (edges) instead of a malloced array, we could improve the else stament of when we get the edges to be just an assigment of the SUCCS so we don't have an extra setting loop in this case. The code in quesiton: if (bb->loop_father->header == bb)

[Bug tree-optimization/18308] ICE in do_jump, at dojump.c:274

--- Additional Comments From dpatel at apple dot com 2004-11-18 01:28 --- I meant, Bigger issue is : if-convert COND_EXPR is not vectorized and do_jump() does not handle it. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18308

[Bug tree-optimization/18308] ICE in do_jump, at dojump.c:274

--- Additional Comments From dpatel at apple dot com 2004-11-18 01:27 --- Bigger issue is : if-convert COND_EXPR is not vectorized and do_jump() does not aborts(). -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18308

[Bug rtl-optimization/16088] [4.0 Regression] Generates wrong code

--- Additional Comments From giovannibajo at libero dot it 2004-11-18 01:26 --- OK closing this as invalid then. It would be great if you could reduce it to something attachable. A big preprocessed source code which segfaults is already a starting point, for instance. -- W

[Bug rtl-optimization/16088] [4.0 Regression] Generates wrong code

--- Additional Comments From lucier at math dot purdue dot edu 2004-11-18 01:13 --- Maybe this bug should be closed for lack of specificity. The problem is the following: gcc-4.0 miscompiles C code generated by the Scheme->C compiler Gambit-C (http://www.iro.umontreal.ca/~gambit). I ha

[Bug bootstrap/18533] EXTRA_MULTILIB_PARTS are used to build libgcc_s.so for multilib

-- What|Removed |Added OtherBugsDependingO||18532 nThis|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18533

[Bug bootstrap/18532] libgcc.mk isn't parallel build safe for multilib

-- What|Removed |Added BugsThisDependsOn||18533 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18532

[Bug target/17735] [4.0 Regression] make stops with "initializer for integer value is too complicated" while building an avr-cross compiler

--- Additional Comments From pbrook at gcc dot gnu dot org 2004-11-18 00:21 --- Patch Here. http://gcc.gnu.org/ml/gcc-patches/2004-11/msg01437.html -- What|Removed |Added

[Bug c++/18534] cout operator<< not in right order

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 22:26 --- No this is undefined code. the order of calling a in a() + a() also undefined by the same rule. -- What|Removed |Added ---

[Bug other/17464] The newly built gcc shared libraries aren't used for bootstap and check

-- Bug 17464 depends on bug 18526, which changed state. Bug 18526 Summary: fortran testsuite broken http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18526 What|Old Value |New Value

[Bug other/17466] Testsuites in gcc override LD_LIBRARY_PATH

--- Additional Comments From hjl at lucon dot org 2004-11-17 22:22 --- *** Bug 18526 has been marked as a duplicate of this bug. *** -- What|Removed |Added CC

[Bug fortran/18526] fortran testsuite broken

--- Additional Comments From hjl at lucon dot org 2004-11-17 22:22 --- *** This bug has been marked as a duplicate of 17466 *** -- What|Removed |Added Status|AS

[Bug tree-optimization/17635] [4.0 regression] ICE in verify_ssa: type mismatch

--- Additional Comments From dpatel at apple dot com 2004-11-17 22:13 --- Subject: Re: [4.0 regression] ICE in verify_ssa: type mismatch On Nov 13, 2004, at 5:55 AM, reichelt at gcc dot gnu dot org wrote: > > --- Additional Comments From reichelt at gcc dot gnu dot org > 2004-1

[Bug c++/18534] New: cout operator<< not in right order

Running on a SuSe 9.1 pro x86_64 machine. Reading specs from /usr/lib64/gcc/x86_64-suse-linux/3.4.3/specs Configured with: ./configure --enable-threads=posix --prefix=/usr --with-local-prefix=/usr/local --infodir=/usr/share/info --mandir=/usr/share/man --enable-languages=c,c++,--disable-checking -

[Bug tree-optimization/18307] merge_pointed_to_info called incorrectly

--- Additional Comments From dnovillo at gcc dot gnu dot org 2004-11-17 21:11 --- Fix: http://gcc.gnu.org/ml/gcc-patches/2004-11/msg01414.html. -- What|Removed |Added

[Bug tree-optimization/18307] merge_pointed_to_info called incorrectly

--- Additional Comments From cvs-commit at gcc dot gnu dot org 2004-11-17 21:07 --- Subject: Bug 18307 CVSROOT:/cvs/gcc Module name:gcc Changes by: [EMAIL PROTECTED] 2004-11-17 21:07:08 Modified files: gcc: ChangeLog tree-ssa-alias.c Log mess

[Bug bootstrap/18533] New: EXTRA_MULTILIB_PARTS are used to build libgcc_s.so for multilib

For multilib, EXTRA_MULTILIB_PARTS are used to build libgcc_s.so. I don't think it should be there at all. -- Summary: EXTRA_MULTILIB_PARTS are used to build libgcc_s.so for multilib Product: gcc Version: 3.4.4 Status: UNCONFIRMED

[Bug bootstrap/18532] New: libgcc.mk isn't parallel build safe for multilib

With multilib, libgcc.mk has crtendS.o: stmp-dirs $(MAKE) GCC_FOR_TARGET="$(GCC_FOR_TARGET)" \ AR_FOR_TARGET="$(AR_FOR_TARGET)" \ AR_CREATE_FOR_TARGET="$(AR_CREATE_FOR_TARGET)" \ AR_EXTRACT_FOR_TARGET="$(AR_EXTRACT_FOR_TARGET)" \ AR_FLAGS_FOR_TARGET=

[Bug middle-end/16266] [4.0 regression] gcc.dg/c99-intconst-1.c compilation is very slow

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 18:16 --- In fact I just tested that patch and fixes the problem. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16266

[Bug middle-end/16266] [4.0 regression] gcc.dg/c99-intconst-1.c compilation is very slow

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 18:13 --- I assume that this is a latenet bug exposed by my patch, see for a patch from Dale which might solve this. -- What|Removed

[Bug middle-end/16266] [4.0 regression] gcc.dg/c99-intconst-1.c compilation is very slow

--- Additional Comments From fjahanian at apple dot com 2004-11-17 18:02 --- Following patch has broken many dejagnu tests on apple-ppc-darwing with -mcpu=G5. http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/function.c.diff?cvsroot=gcc&r1=1.581&r2=1.582 FAIL: gcc.c-torture/execute/20041

[Bug tree-optimization/18374] [meta-bug] Argument and return value marshalling at tree level

> Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 > Here is a bug which says the oposite should be done (as arguments are > already > done): PR 15484. Unless I misunderstand; C's promotion/evaluation semantics need not be literally applied, but must only yield equivalent result

[Bug c++/18530] Bogus warning about shadowed variable in locale_facets.tcc

-- What|Removed |Added Component|libstdc++ |c++ http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18530

[Bug target/18531] Rationalize Altivec logical ops

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 17:27 --- Note vandc was already done (and approved): http://gcc.gnu.org/ml/gcc-patches/2004-11/msg01328.html -- What|Removed |Added ---

[Bug fortran/18526] fortran testsuite broken

-- What|Removed |Added CC|hjl at lucon dot org| AssignedTo|unassigned at gcc dot gnu |hjl at lucon dot org |dot org

[Bug libstdc++/18530] Bogus warning about shadowed variable in locale_facets.tcc

--- Additional Comments From martin at mpa-garching dot mpg dot de 2004-11-17 17:00 --- Hi Paolo, I didn't even look that closely, sorry. If the warning is bogus, there is indeed a compiler bug; I have upgraded the severity back to "normal" and updated the title. --

[Bug c++/16625] Discarded Linkonce sections in .rodata

--- Additional Comments From gustavo dot boiko at kdemail dot net 2004-11-17 16:59 --- any news on this problem? I've seen some problems like this one in OpenOffice.org cvs HEAD compilation with gcc-3.4.3 because of some inlined class methods at the cppu and the xmlscript modules. --

[Bug libstdc++/18530] Warning about shadowed variable in locale_facets.tcc

--- Additional Comments From pcarlini at suse dot de 2004-11-17 16:44 --- Hi. Frankly, I believe the warning is bogus, and is actually a compiler bug: there is no "member of 'this'" named __ct, AFAICS. Also the fact that appears only at -O3 (not at -O2, f.i.) seems very suspect... --

[Bug target/18531] Rationalize Altivec logical ops

-- What|Removed |Added Status|UNCONFIRMED |ASSIGNED Ever Confirmed||1 Last reconfirmed|-00-00 00:00:00 |200

[Bug target/18531] New: Rationalize Altivec logical ops

Altivec patterns are defined at all integer vector modes for GCC named patterns, but not for combiner patterns (e.g., vandc). The patterns probably should use comparison_operator, %q output template, and MD file macro for multiple integer vector modes. -- Summary: Rationalize Altivec

[Bug fortran/18525] [gfortran] ICE on valid code in gfc_get_symbol_decl()

--- Additional Comments From martin at mpa-garching dot mpg dot de 2004-11-17 16:18 --- Thanks, I have updated the title and keywords now. -- What|Removed |Added

[Bug libstdc++/18530] Warning about shadowed variable in locale_facets.tcc

--- Additional Comments From martin at mpa-garching dot mpg dot de 2004-11-17 16:14 --- The warning is only displayed at -O3. Lower optimization levels don't trigger it. -- What|Removed |Added

[Bug libstdc++/18530] New: Warning about shadowed variable in locale_facets.tcc

ratch/gcc/configure --quiet --prefix=/afs/mpa/data/martin/ugcc --enable-languages=c++,f95 --with-gmp=/afs/mpa/data/martin/mygmp Thread model: posix gcc version 4.0.0 20041117 (experimental) /afs/mpa/data/martin/ugcc/libexec/gcc/i686-pc-linux-gnu/4.0.0/cc1plus -quiet -v -D_GNU_SOURCE bug.cc -quiet

[Bug tree-optimization/18507] block_defs_stack varrray should not be GC'ed

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 16:00 --- Patch here: . -- What|Removed |Added

[Bug tree-optimization/18374] [meta-bug] Argument and return value marshalling at tree level

-- What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirmed||1 Last reconfirmed|-00-00 00:00:00 |2004-11-

[Bug target/18528] miscompilation of inline assembler with -fweb

-- What|Removed |Added Component|c |target Keywords||wrong-code http://gcc.gnu.org/bugzilla/show_bug.cgi?id

[Bug c/18314] Abnormal behavior in optimization

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 15:46 --- So this is invalid so closing. -- What|Removed |Added Status|UNCONFIRMED

[Bug tree-optimization/18529] When the lower bound of a loop is non-constant we cannot find the number of iterations

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 15:44 --- Note this is like PR 18527 but a different problem (I don't know how offten this shows up in SPEC but I found this when I was working on my loops to memset pass). -- What|Removed

[Bug tree-optimization/18529] New: When the lower bound of a loop is non-constant we cannot find the number of iterations

long *a; void f(int start, int end) { int i; for (i = start; i http://gcc.gnu.org/ml/gcc/2004-08/msg01315.html>. I filed this to keep track of it. -- Summary: When the lower bound of a loop is non-constant we cannot find the number of iterations Pro

[Bug fortran/18525] [gfortran] ICE on invalid(?) code in gfc_get_symbol_decl()

--- Additional Comments From Thomas dot Koenig at online dot de 2004-11-17 15:39 --- The code is perfectly valid. Here, n in the subroutine hides the host parameter n (which is then inaccessible in the subroutine xx). -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18525

[Bug tree-optimization/18527] cannot determine loop bounds for loops with <=

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 15:36 --- Confirmed, note PR 16803 is related (in fact when this is fixed that one should be fixed also). -- What|Removed |Added --

[Bug tree-optimization/18067] [4.0 Regression] ICE at loc_descriptor_from_tree_1 in dwarf2out.c (VLA)

--- Additional Comments From phython at gcc dot gnu dot org 2004-11-17 15:13 --- As Andrew mentioned, this bug affects more than powerpc. An enable-checking build is needed to reproduce this bug. With a disable-checking build the debugging code ignores SSA_NAMEs. -- What

[Bug fortran/17675] [4.0 Regression] Alignment constraints not honored in EQUIVALENCE

--- Additional Comments From phython at gcc dot gnu dot org 2004-11-17 15:09 --- I'm getting the same problems on sparc32 builds. -- What|Removed |Added Status|

[Bug other/18508] [3.4/4.0 Regression] "basename: too few arguments" when building without bootstrap

-- What|Removed |Added Keywords||build Summary|"basename: too few |[3.4/4.0 Regression] |arguments" when

[Bug rtl-optimization/17950] Over Aggressive Use of Data Cache Touch Instructions During FDO

--- Additional Comments From rakdver at atrey dot karlin dot mff dot cuni dot cz 2004-11-17 14:41 --- Subject: Re: Over Aggressive Use of Data Cache Touch Instructions During FDO Hello, > I cannot reproduce this with the 20041118 CVS head on > powerpc64-unknown-linux-gnu. I think th

[Bug rtl-optimization/17950] Over Aggressive Use of Data Cache Touch Instructions During FDO

--- Additional Comments From nathan at gcc dot gnu dot org 2004-11-17 14:34 --- I cannot reproduce this with the 20041118 CVS head on powerpc64-unknown-linux-gnu. Can you confirm the command line options and provide the bug.gcno, bug.gcda files. -- http://gcc.gnu.org/bugzilla/show_

[Bug tree-optimization/18519] [4.0 Regression] ICE: Segmentation fault with optimization

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 12:46 --- Fixed. -- What|Removed |Added Status|ASSIGNED|RESOLVED

[Bug tree-optimization/18374] [meta-bug] Argument and return value marshalling at tree level

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 13:13 --- Here is a bug which says the oposite should be done (as arguments are already done): PR 15484. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18374

[Bug fortran/18526] fortran testsuite broken

--- Additional Comments From Thomas dot Koenig at online dot de 2004-11-17 12:45 --- Created an attachment (id=7561) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=7561&action=view) log file from testsuite There appears to be some problem with loading the shared library libunwind.s

[Bug fortran/18525] [gfortran] ICE on invalid(?) code in gfc_get_symbol_decl()

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 12:58 --- Confirmed, this is related to PR 13082. -- What|Removed |Added BugsThisDependsOn|

[Bug c/18314] Abnormal behavior in optimization

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 13:08 --- I think you want: *(volatile int*)&test volatile only matters on pointers and types of variables and not casts IIRC. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18314

[Bug tree-optimization/18527] New: Missed optimization

When loop bound condition is <=, number_of_iterations_in_loop returns 'unknown' although the loop is countable. Turns out this case is pretty common in SPEC and blocks vectorization opportunities. E.g., int foo () { int a[N]; int i; int n; for (i = 0; i <= n; i++) { ca[i] = 2;

[Bug tree-optimization/15484] [tree-ssa] bool and short function arguments promoted to int

--- Additional Comments From nathan at gcc dot gnu dot org 2004-11-17 13:33 --- The underlying ABI might require short & bool arguments to be passed as int. Some do, some don't. Anyway, that's an argument marshalling issue, which is not currently addressed at the tree level. The cast

[Bug libfortran/17999] [4.0 Regression] libfortran: uses some C99 functions (snprintf)

--- Additional Comments From ebotcazou at gcc dot gnu dot org 2004-11-17 13:57 --- Fixing. -- What|Removed |Added AssignedTo|unassigned at gcc dot gnu |ebotcazou

[Bug libfortran/16135] [4.0 Regression] libfortran doesn't build, use of C99 types

--- Additional Comments From ebotcazou at gcc dot gnu dot org 2004-11-17 13:56 --- Fixing. -- What|Removed |Added AssignedTo|unassigned at gcc dot gnu |ebotcazou

[Bug c/18528] miscompilation of inline assembler with -fweb

--- Additional Comments From janne at hybrid dot fi 2004-11-17 13:52 --- Created an attachment (id=7562) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=7562&action=view) bug case for -fweb Sample case for -fweb bug. Just for convenience. This is the original source and not .i fil

[Bug fortran/18526] fortran testsuite broken

--- Additional Comments From pinskia at gcc dot gnu dot org 2004-11-17 12:51 --- A fall out (most likely forgotten part) when add libunwind. -- What|Removed |Added

[Bug c/18528] New: miscompilation of inline assembler with -fweb

On arm-elf-gcc, I encountered a miscompilation of inline assembly. The miscompilation only occurs when compiled with -fweb. -fno-web makes the sample compile correctly even with -O3 (and -fnew-ra). Here is the configuration: Reading specs from /home/build/gnu/lib/gcc/arm-elf/3.4.3/specs Configur

[Bug c/18314] Abnormal behavior in optimization

--- Additional Comments From schwab at suse dot de 2004-11-17 13:45 --- 6.7.3#3: The properties associated with qualified types are meaningful only for expressions that are lvalues. This is not a constraint, so not an error. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18314

[Bug fortran/18526] New: fortran testsuite broken

>From a freshly extracted gcc-4.0-20041114. "make bootstrap" worked like a charm. cd-ing into the gcc subdirectory of my build directoy: $ ./xgcc -v Reading specs from /home/zfkts/lib/gcc/ia64-unknown-linux-gnu/4.0.0/specs Configured with: ../gcc-4.0-20041114/configure --prefix=/home/zfkts --ena

[Bug target/18263] [3.4 only] Build broken for ARC.

--- Additional Comments From giovannibajo at libero dot it 2004-11-17 12:30 --- Fixed. -- What|Removed |Added Status|NEW |RESOLVED

[Bug target/18263] [3.4 only] Build broken for ARC.

--- Additional Comments From cvs-commit at gcc dot gnu dot org 2004-11-17 12:30 --- Subject: Bug 18263 CVSROOT:/cvs/gcc Module name:gcc Branch: gcc-3_4-branch Changes by: [EMAIL PROTECTED] 2004-11-17 12:29:17 Modified files: gcc: Change

[Bug fortran/18525] [gfortran] ICE on invalid(?) code in gfc_get_symbol_decl()

-- What|Removed |Added Known to fail||4.0.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18525

[Bug fortran/18525] New: [gfortran] ICE on invalid(?) code in gfc_get_symbol_decl()

rtin/ugcc --enable-languages=c++,f95 --with-gmp=/afs/mpa/data/martin/mygmp Thread model: posix gcc version 4.0.0 20041117 (experimental) /afs/mpa/data/martin/ugcc/libexec/gcc/i686-pc-linux-gnu/4.0.0/f951 bug.f90 -quiet -dumpbase bug.f90 -mtune=pentiumpro -auxbase bug -version -o /tmp/ccVLWgWD.

[Bug target/14010] large stack frame long double ice

--- Additional Comments From amodra at bigpond dot net dot au 2004-11-17 11:40 --- Seems to be fixed on the 3.4.x branch too. -- What|Removed |Added Status|NEW

[Bug target/18506] Altivec definitions of vec_init

--- Additional Comments From dorit at il dot ibm dot com 2004-11-17 11:09 --- (In reply to comment #1) > PR 10469 looks like a testcase where this could improve the code generation.\ gcc.dg/vect/vect-25.c is a testcase too. This is what is currently generated (the invariant is in r3):

[Bug target/18503] _mm_move_ss SSE intrinsic causes erroneous

--- Additional Comments From uros at gcc dot gnu dot org 2004-11-17 10:10 --- Patch here: http://gcc.gnu.org/ml/gcc-patches/2004-11/msg01377.html -- What|Removed |Added

[Bug target/18503] _mm_move_ss SSE intrinsic causes erroneous

--- Additional Comments From uros at gcc dot gnu dot org 2004-11-17 09:19 --- . -- What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirme

[Bug target/15397] [g77] c float function called from fortran gives wrong result

--- Additional Comments From Thomas dot Koenig at online dot de 2004-11-17 09:18 --- (In reply to comment #3) I checked this with gcc-4.0-20041107 on a ia64, and the problem didn't occur: $ gfortran -v Reading specs from /home/zfkts/lib/gcc/ia64-unknown-linux-gnu/4.0.0/specs Configured

[Bug target/18503] _mm_move_ss SSE intrinsic causes erroneous

--- Additional Comments From uros at kss-loka dot si 2004-11-17 09:17 --- The problem was that "sse_movss" pattern has wrong vec_merge constant, so it was wrongly combined to "vmsubv4sf3" pattern. Instead of 0x1, "sse_movss" pattern should have 0x14: (define_insn "sse_movss" [(set (ma

[Bug fortran/18518] d1mach(4) (BLAS) broken, g77 regression

--- Additional Comments From c dot lemmen at fz-juelich dot de 2004-11-17 08:46 --- On AMD x32 the correct result is obtained with g77. g95 and pgfortran work as expected, only gfortran stops: > gfortran d1mach.f && ./a.out 2.225073858507201E-308 At line 215 of file d1mach.f STOP 778