* Re: egcs-971105 powerpc-ibm-aix4.1.4.0 new regression compile/920501-4.c
@ 1997-11-18 7:52 Kaveh R. Ghazi
0 siblings, 0 replies; 2+ messages in thread
From: Kaveh R. Ghazi @ 1997-11-18 7:52 UTC (permalink / raw)
To: egcs
The following error still occurs in egcs-971114 and is one
case where egcs does worse than the most recent gcc snapshot.
--Kaveh
> I got a c-torture regression in egcs-971105 on powerpc-ibm-aix4.1.4.0.
>
> I successfully bootstrapped a clean snapshot using BOOT_CFLAGS='-g -O3
> -mminimal-toc' and also a clean snapshot plus Jim's unroll patch at
> BOOT_CFLAGS='-g -O3 -funroll-all-loops -mminimal-toc -pipe'.
>
> Both bootstrapps have these new c-torture problems:
>
> ERROR: 920501-4.c: compiler got signal 6: -O2 -fomit-frame-pointer
> ERROR: 920501-4.c: compiler got signal 6: -O3 -fomit-frame-pointer
> ERROR: 920501-4.c: compiler got signal 6: -O3 -funroll-all-loops
>
> Below is some more info about the failure.
>
> --Kaveh
>
> ---------------------------------------------------------------------------
>
> This is what gets output when compiled by hand with -O2
> -fomit-frame-pointer:
>
> compile/920501-4.c: In function `foo':
> compile/920501-4.c:49: Could not find a spill register
> (insn:QI 657 668 771 (parallel[
> (set (mem:SI (plus:SI (reg:SI 1 r1)
> (const_int 32808)))
> (lshiftrt:SI (neg:SI (abs:SI (reg/v:SI 66 ctr)))
> (const_int 31)))
> (clobber (scratch:SI))
> ] ) 540 {sleu+25} (nil)
> (expr_list:REG_UNUSED (scratch:SI)
> (nil)))
>
> IOT/Abort trap in raise at 0xd0078900
>
>
> ---------------------------------------------------------------------------
>
> Here is the backtrace from cc1 using dbx:
>
>
> raise(??) at 0xd0078900
> abort() at 0xd0019a14
> fatal_insn(message = (nil), insn = internal error: assertion failed at line 947
> 0xd0078900), line 1135 in "toplev.c"
> allocate_reload_reg(??, ??, ??, ??), line 5196 in "reload1.c"
> unnamed block $b6424, line 5746 in "reload1.c"
> choose_reload_regs(??, ??), line 5746 in "reload1.c"
> unnamed block $b6384, line 4127 in "reload1.c"
> unnamed block $b6382, line 4127 in "reload1.c"
> unnamed block $b6381, line 4127 in "reload1.c"
> reload_as_needed(??, ??), line 4127 in "reload1.c"
> reload1.reload(??, ??, ??), line 2081 in "reload1.c"
> global_alloc(??), line 584 in "global.c"
> unnamed block $b1372, line 3521 in "toplev.c"
> unnamed block $b1371, line 3521 in "toplev.c"
> rest_of_compilation(decl = internal error: assertion failed at line 947 in file
> 0x20058c64), line 3521 in "toplev.c"
> finish_function(??), line 7050 in "c-decl.c"
> yyparse(), line 144 in "c-parse.y"
> compile_file(name = internal error: assertion failed at line 947 in file frame.c
> "foo.i"), line 2513 in "toplev.c"
--
Kaveh R. Ghazi Project Manager / Custom Development
ghazi@caip.rutgers.edu ICon CMT Corp.
^ permalink raw reply [flat|nested] 2+ messages in thread
* egcs-971105 powerpc-ibm-aix4.1.4.0 new regression compile/920501-4.c
@ 1997-11-09 8:06 Kaveh R. Ghazi
0 siblings, 0 replies; 2+ messages in thread
From: Kaveh R. Ghazi @ 1997-11-09 8:06 UTC (permalink / raw)
To: egcs; +Cc: ghazi
I got a c-torture regression in egcs-971105 on powerpc-ibm-aix4.1.4.0.
I successfully bootstrapped a clean snapshot using BOOT_CFLAGS='-g -O3
-mminimal-toc' and also a clean snapshot plus Jim's unroll patch at
BOOT_CFLAGS='-g -O3 -funroll-all-loops -mminimal-toc -pipe'.
Both bootstrapps have these new c-torture problems:
ERROR: 920501-4.c: compiler got signal 6: -O2 -fomit-frame-pointer
ERROR: 920501-4.c: compiler got signal 6: -O3 -fomit-frame-pointer
ERROR: 920501-4.c: compiler got signal 6: -O3 -funroll-all-loops
Below is some more info about the failure.
--Kaveh
---------------------------------------------------------------------------
This is what gets output when compiled by hand with -O2
-fomit-frame-pointer:
compile/920501-4.c: In function `foo':
compile/920501-4.c:49: Could not find a spill register
(insn:QI 657 668 771 (parallel[
(set (mem:SI (plus:SI (reg:SI 1 r1)
(const_int 32808)))
(lshiftrt:SI (neg:SI (abs:SI (reg/v:SI 66 ctr)))
(const_int 31)))
(clobber (scratch:SI))
] ) 540 {sleu+25} (nil)
(expr_list:REG_UNUSED (scratch:SI)
(nil)))
IOT/Abort trap in raise at 0xd0078900
---------------------------------------------------------------------------
Here is the backtrace from cc1 using dbx:
raise(??) at 0xd0078900
abort() at 0xd0019a14
fatal_insn(message = (nil), insn = internal error: assertion failed at line 947
0xd0078900), line 1135 in "toplev.c"
allocate_reload_reg(??, ??, ??, ??), line 5196 in "reload1.c"
unnamed block $b6424, line 5746 in "reload1.c"
choose_reload_regs(??, ??), line 5746 in "reload1.c"
unnamed block $b6384, line 4127 in "reload1.c"
unnamed block $b6382, line 4127 in "reload1.c"
unnamed block $b6381, line 4127 in "reload1.c"
reload_as_needed(??, ??), line 4127 in "reload1.c"
reload1.reload(??, ??, ??), line 2081 in "reload1.c"
global_alloc(??), line 584 in "global.c"
unnamed block $b1372, line 3521 in "toplev.c"
unnamed block $b1371, line 3521 in "toplev.c"
rest_of_compilation(decl = internal error: assertion failed at line 947 in file
0x20058c64), line 3521 in "toplev.c"
finish_function(??), line 7050 in "c-decl.c"
yyparse(), line 144 in "c-parse.y"
compile_file(name = internal error: assertion failed at line 947 in file frame.c
"foo.i"), line 2513 in "toplev.c"
main(argc = 5, argv = 0x2ff22aac, envp = 0xffffffff), line 4394 in "toplev.c"
--
Kaveh R. Ghazi Project Manager
ghazi@caip.rutgers.edu ICon CMT Corp.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~1997-11-18 7:52 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1997-11-18 7:52 egcs-971105 powerpc-ibm-aix4.1.4.0 new regression compile/920501-4.c Kaveh R. Ghazi
-- strict thread matches above, loose matches on Subject: below --
1997-11-09 8:06 Kaveh R. Ghazi
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).