* Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
@ 1998-06-23 22:55 Alexandre Oliva
1998-06-24 10:08 ` Craig Burley
0 siblings, 1 reply; 11+ messages in thread
From: Alexandre Oliva @ 1998-06-23 22:55 UTC (permalink / raw)
To: egcs-results, egcs
Native configuration is sparc-sun-sunos4.1.3
=== g++ tests ===
Running target unix
FAIL: g++.jason/rfg15.C returning a value from a void function (test for errors, line 5)
FAIL: g++.jason/rfg15.C (test for excess errors)
FAIL: g++.law/profile1.C Execution test
FAIL: g++.mike/p9706.C Execution test
FAIL: g++.pt/static3.C (test for excess errors)
FAIL: g++.robertl/980310-2.C Execution test
FAIL: g++.robertl/980310-3.C Execution test
FAIL: g++.robertl/eb100.C (test for excess errors)
FAIL: g++.robertl/eb101.C (test for excess errors)
FAIL: g++.robertl/eb117.C Execution test
FAIL: g++.robertl/eb129.C (test for excess errors)
FAIL: g++.robertl/eb130.C (test for excess errors)
FAIL: g++.robertl/eb131.C (test for excess errors)
FAIL: g++.robertl/eb25.C (test for excess errors)
FAIL: g++.robertl/eb40.C (test for excess errors)
FAIL: g++.robertl/eb47.C (test for excess errors)
FAIL: g++.robertl/eb53.C (test for excess errors)
FAIL: g++.robertl/eb56.C (test for excess errors)
FAIL: g++.robertl/eb68.C (test for excess errors)
FAIL: g++.robertl/eb72.C (test for excess errors)
FAIL: g++.robertl/eb89.C (test for excess errors)
FAIL: g++.robertl/eb91.C (test for excess errors)
FAIL: g++.robertl/float1.C Execution test
FAIL: g++.robertl/rtti5.C Execution test
=== g++ Summary ===
# of expected passes 4170
# of unexpected failures 24
# of expected failures 94
# of untested testcases 6
/l/dsk01/temp/egcs/gcc/testsuite/../xgcc version egcs-2.91.42 19980621 (gcc2 ss-980502 experimental)
=== g77 tests ===
Running target unix
FAIL: g77.f-torture/execute/cabs.f execution, -O0
FAIL: g77.f-torture/execute/cabs.f execution, -O1
FAIL: g77.f-torture/execute/cabs.f execution, -O2
FAIL: g77.f-torture/execute/cabs.f execution, -O2 -fomit-frame-pointer -finline-functions
FAIL: g77.f-torture/execute/cabs.f execution, -Os
=== g77 Summary ===
# of expected passes 232
# of unexpected failures 5
/l/dsk01/temp/egcs/gcc/g77 version egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) (from FSF-g77 version 0.5.23)
gcc version egcs-2.91.42 19980621 (gcc2 ss-980502 experimental)
=== gcc tests ===
Running target unix
FAIL: gcc.c-torture/execute/980505-1.c execution, -O2
FAIL: gcc.c-torture/execute/980505-1.c execution, -O2 -fomit-frame-pointer -finline-functions
FAIL: gcc.c-torture/execute/980505-1.c execution, -O2 -g
FAIL: gcc.c-torture/execute/980505-1.c execution, -Os
FAIL: gcc.c-torture/execute/980506-2.c execution, -O2
FAIL: gcc.c-torture/execute/980506-2.c execution, -O2 -g
FAIL: gcc.c-torture/execute/980506-2.c execution, -Os
FAIL: gcc.c-torture/execute/980526-1.c execution, -O2 -fomit-frame-pointer -finline-functions
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O0
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O1
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O2
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O2 -fomit-frame-pointer -finline-functions
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O2 -fomit-frame-pointer -finline-functions -funroll-loops
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O2 -fomit-frame-pointer -finline-functions -funroll-all-loops
FAIL: gcc.c-torture/execute/loop-2f.c execution, -O2 -g
FAIL: gcc.c-torture/execute/loop-2f.c execution, -Os
=== gcc Summary ===
# of expected passes 7397
# of unexpected failures 16
# of expected failures 7
# of unsupported tests 17
/l/dsk01/temp/egcs/gcc/xgcc version egcs-2.91.42 19980621 (gcc2 ss-980502 experimental)
=== libio tests ===
Running target unix
=== libio Summary ===
# of expected passes 40
=== libstdc++ tests ===
Running target unix
=== libstdc++ Summary ===
# of expected passes 30
configure flags: --with-gnu-as --enable-shared
BOOT_CFLAGS: -O4 -g
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-23 22:55 Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3 Alexandre Oliva
@ 1998-06-24 10:08 ` Craig Burley
1998-06-24 14:50 ` Alexandre Oliva
0 siblings, 1 reply; 11+ messages in thread
From: Craig Burley @ 1998-06-24 10:08 UTC (permalink / raw)
To: oliva; +Cc: egcs
>FAIL: g77.f-torture/execute/cabs.f execution, -O0
>FAIL: g77.f-torture/execute/cabs.f execution, -O1
>FAIL: g77.f-torture/execute/cabs.f execution, -O2
>FAIL: g77.f-torture/execute/cabs.f execution, -O2 -fomit-frame-pointer -finline-functions
>FAIL: g77.f-torture/execute/cabs.f execution, -Os
Could somebody with a SunOS 4.1.3 (or 4.1.x, I guess) system look
into this?
In particular, I wonder if this is a case of the wrong c_abs, z_abs,
or f__cabs being linked in (e.g. from a non-libf2c/libg2c library)?
We had that problem come up years ago during g77 alpha-testing, IIRC
on this same machine. I think Kate Hedstrom ran into that, and it
was basically due to not linking the libraries in the right order.
(And, now that I've changed some naming and placement of the libf2c
library...maybe the problem has resurfaced?)
tq vm, (burley)
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-24 10:08 ` Craig Burley
@ 1998-06-24 14:50 ` Alexandre Oliva
1998-06-24 20:08 ` Jeffrey A Law
` (2 more replies)
0 siblings, 3 replies; 11+ messages in thread
From: Alexandre Oliva @ 1998-06-24 14:50 UTC (permalink / raw)
To: Craig Burley; +Cc: egcs
Craig Burley <burley@gnu.org> writes:
>> FAIL: g77.f-torture/execute/cabs.f execution, -O0
> Could somebody with a SunOS 4.1.3 (or 4.1.x, I guess) system look
> into this?
oliva@atibaia% egcsc cabs.f -lg2c
/n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42/libg2c.a(cabs.o): In function `f__cabs':
/n/temp1/gcctest/bin/../src/egcs/libf2c/libF77/cabs.c:25: undefined reference to `sqrt'
collect2: ld returned 1 exit status
oliva@atibaia% egcsc cabs.f -lg2c -lm
oliva@atibaia% ./a.out
Does anyone know how we can add `-lm' to the link command of this test
program, iff libm exists?
--
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-24 14:50 ` Alexandre Oliva
@ 1998-06-24 20:08 ` Jeffrey A Law
1998-06-25 6:50 ` Fred Fish
1998-06-24 21:23 ` Craig Burley
1998-06-25 6:50 ` Dave Love
2 siblings, 1 reply; 11+ messages in thread
From: Jeffrey A Law @ 1998-06-24 20:08 UTC (permalink / raw)
To: Alexandre Oliva; +Cc: Craig Burley, egcs
In message < orzpf2d4w6.fsf@zecarneiro.lsd.dcc.unicamp.br >you write:
> collect2: ld returned 1 exit status
> oliva@atibaia% egcsc cabs.f -lg2c -lm
> oliva@atibaia% ./a.out
>
> Does anyone know how we can add `-lm' to the link command of this test
> program, iff libm exists?
Not that I'm aware of.
We should probably just add -lm all the time. I'd be amazed if systems
without libm would even get far enough to try running the testsuite :-)
jeff
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-24 14:50 ` Alexandre Oliva
1998-06-24 20:08 ` Jeffrey A Law
@ 1998-06-24 21:23 ` Craig Burley
1998-06-25 6:50 ` Dave Love
2 siblings, 0 replies; 11+ messages in thread
From: Craig Burley @ 1998-06-24 21:23 UTC (permalink / raw)
To: oliva; +Cc: egcs
>oliva@atibaia% egcsc cabs.f -lg2c
>/n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42/libg2c.a(cabs.o): In function `f__cabs':
>/n/temp1/gcctest/bin/../src/egcs/libf2c/libF77/cabs.c:25: undefined reference to `sqrt'
>collect2: ld returned 1 exit status
>oliva@atibaia% egcsc cabs.f -lg2c -lm
>oliva@atibaia% ./a.out
>
>Does anyone know how we can add `-lm' to the link command of this test
>program, iff libm exists?
The g77 driver does this automatically. By "this" I mean adds -lg2c -lm.
In fact, I think if you specify -lg2c, it adds -lm after it; and if
you specify -lm, it adds -lg2c before it. Haven't tested that for
awhile. (But most of this is only since snapshot 19980615 or so;
before that, adding -lg2c -lm was done rather "mechanically".)
tq vm, (burley)
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-24 20:08 ` Jeffrey A Law
@ 1998-06-25 6:50 ` Fred Fish
0 siblings, 0 replies; 11+ messages in thread
From: Fred Fish @ 1998-06-25 6:50 UTC (permalink / raw)
To: law; +Cc: oliva, burley, egcs
> > Does anyone know how we can add `-lm' to the link command of this test
> > program, iff libm exists?
> Not that I'm aware of.
>
> We should probably just add -lm all the time. I'd be amazed if systems
> without libm would even get far enough to try running the testsuite :-)
Please don't do this. BeOS should be able to run the testsuite just
fine, once I install an updated dejagnu.
-Fred
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-24 14:50 ` Alexandre Oliva
1998-06-24 20:08 ` Jeffrey A Law
1998-06-24 21:23 ` Craig Burley
@ 1998-06-25 6:50 ` Dave Love
1998-06-26 7:16 ` Alexandre Oliva
2 siblings, 1 reply; 11+ messages in thread
From: Dave Love @ 1998-06-25 6:50 UTC (permalink / raw)
To: Alexandre Oliva; +Cc: Craig Burley, egcs
>>>>> "AO" == Alexandre Oliva <oliva@dcc.unicamp.br> writes:
AO> oliva@atibaia% egcsc cabs.f -lg2c
AO> /n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42/libg2c.a(cabs.o): In function `f__cabs':
AO> /n/temp1/gcctest/bin/../src/egcs/libf2c/libF77/cabs.c:25: undefined reference to `sqrt'
AO> collect2: ld returned 1 exit status
AO> oliva@atibaia% egcsc cabs.f -lg2c -lm
AO> oliva@atibaia% ./a.out
I suspect that's not testing the right thing, though. You should use
the `g77' driver, which is analagous to the `g++' one; the `-v' switch
can be helpful. What's the output from the test in the testsuite
g77.log?
AO> Does anyone know how we can add `-lm' to the link command of this test
AO> program, iff libm exists?
The relevant line from `g77 -v cabs.f' on a non-egcs installation I
have available on sunos4 is this (which works):
/usr/local/lib/gcc-lib/sparc-sun-sunos4.1.3/2.7.2.3.f.1/ld -e start -dc -dp /lib/crt0.o -L/usr/local/lib/gcc-lib/sparc-sun-sunos4.1.3/2.7.2.3.f.1 -L/usr/local/lib /tmp/cca196901.o -lf2c -lm -lgcc -lc -lgcc
(You should see -lg2c now, not -lf2c.) If it's really a problem with
-lm, the question is whether the library list needs to be permuted or
added to somehow. Thanks for any light you can shed.
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-25 6:50 ` Dave Love
@ 1998-06-26 7:16 ` Alexandre Oliva
1998-06-26 13:03 ` Craig Burley
0 siblings, 1 reply; 11+ messages in thread
From: Alexandre Oliva @ 1998-06-26 7:16 UTC (permalink / raw)
To: Dave Love; +Cc: Craig Burley, egcs
Dave Love <d.love@dl.ac.uk> writes:
> You should use the `g77' driver, which is analagous to the `g++' one
Of course! :-)
> What's the output from the test in the testsuite g77.log?
Executing on host: /l/dsk01/temp/egcs/gcc/g77 -B/l/dsk01/temp/egcs/gcc/ /tmp_mnt/n/temp1/tmp/gcctest/src/egcs-main/gcc/testsuite/g77.f-torture/execute/cabs.f -w -O0 -L/l/dsk01/temp/egcs/libraries/libf2c -lm -o /l/dsk01/temp/egcs/gcc/testsuite/cabs.x
PASS: g77.f-torture/execute/cabs.f compilation, -O0
spawn [open ...]
Fortran abort routine called
The problem seems to have something to do with linking libm before
libg2c. Since the testsuite links with -lm but not with -lg2c, ld is
called like this:
/n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42/collect2 -e start -dc -dp /lib/crt0.o -L/n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42 -L/n/temp1/gcctest/egcs/SunOS-4.1.3/sparc-sun-sunos4.1.3/lib -L/n/temp1/gcctest/egcs/SunOS-4.1.3/lib /tmp/cca073921.o -lm -lg2c -lm -lgcc -lc -lgcc
Example:
oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f && ./a.out
oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f -lg2c -lm && ./a.out
oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f -lm && ./a.out
Fortran abort routine called
Abort (core dumped)
oliva@atibaia% gdb ./a.out core
GNU gdb 4.17
[snip]
Program terminated with signal 6, Abort.
Reading symbols from /usr/lib/libc.so.1.8...done.
Reading symbols from /usr/lib/libdl.so.1.0...done.
#0 0xf775c120 in kill ()
(gdb) where
#0 0xf775c120 in kill ()
#1 0xf777449c in abort ()
#2 0x565c in sig_die (s=0x3f18 "Fortran abort routine called", kill=1)
at /n/temp1/gcctest/bin/../src/egcs/libf2c/libF77/sig_die.c:34
#3 0x3f50 in G77_abort_0 ()
at /n/temp1/gcctest/bin/../src/egcs/libf2c/libF77/abort_.c:14
#4 0x22fc in MAIN__ () at cabs.f:9
#5 0x3eec in main (argc=1, argv=0xf7fff64c)
at /n/temp1/gcctest/bin/../src/egcs/libf2c/libF77/main.c:61
--
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-26 7:16 ` Alexandre Oliva
@ 1998-06-26 13:03 ` Craig Burley
1998-06-27 19:07 ` Alexandre Oliva
1998-06-27 22:14 ` Jeffrey A Law
0 siblings, 2 replies; 11+ messages in thread
From: Craig Burley @ 1998-06-26 13:03 UTC (permalink / raw)
To: oliva; +Cc: d.love, egcs, egcs-patches, burley
>The problem seems to have something to do with linking libm before
>libg2c. Since the testsuite links with -lm but not with -lg2c, ld is
>called like this:
>
>/n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42/collect2 -e start -dc -dp /lib/crt0.o -L/n/temp1/gcctest/egcs/SunOS-4.1.3/lib/gcc-lib/sparc-sun-sunos4.1.3/egcs-2.91.42 -L/n/temp1/gcctest/egcs/SunOS-4.1.3/sparc-sun-sunos4.1.3/lib -L/n/temp1/gcctest/egcs/SunOS-4.1.3/lib /tmp/cca073921.o -lm -lg2c -lm -lgcc -lc -lgcc
>
>Example:
>
>oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f && ./a.out
>oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f -lg2c -lm && ./a.out
>oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f -lm && ./a.out
>Fortran abort routine called
>Abort (core dumped)
Here is a patch that should fix this problem. I'm putting it into
g77 0.5.24 as well. Dave Love, once this patch is confirmed as
working by Alexandre or someone else on SunOS4, could you please
install it?
The news.texi that I've written for 0.5.24 (and will later
submit patches to clean up egcs-g77's news.texi shortly before
final testing) is:
@item
The @code{g77} driver now ensures that @samp{-lg2c}
is specified in the link phase prior to any
occurrence of @samp{-lm}.
This prevents accidentally linking to a routine
in the SunOS4 @samp{-lm} library
when the generated code wants to link to the one
in @code{libf2c} (@code{libg2c}).
Fri Jun 26 11:54:19 1998 Craig Burley <burley@gnu.org>
* g77spec.c (lang_specific_driver): Put `-lg2c' in
front of any `-lm' that is seen.
*** g77-e/gcc/f/g77spec.c.~1~ Fri Jun 19 16:02:11 1998
--- g77-e/gcc/f/g77spec.c Fri Jun 26 11:55:48 1998
*************** Report bugs to <egcs-bugs@cygnus.org>.\n
*** 512,515 ****
--- 512,517 ----
if (saw_library == 1)
saw_library = 2; /* -l<library> -lm. */
+ else
+ append_arg (FORTRAN_LIBRARY);
}
else if (strcmp (argv[i], FORTRAN_LIBRARY) == 0)
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-26 13:03 ` Craig Burley
@ 1998-06-27 19:07 ` Alexandre Oliva
1998-06-27 22:14 ` Jeffrey A Law
1 sibling, 0 replies; 11+ messages in thread
From: Alexandre Oliva @ 1998-06-27 19:07 UTC (permalink / raw)
To: Craig Burley; +Cc: d.love, egcs, egcs-patches
Craig Burley <burley@gnu.org> writes:
>> oliva@atibaia% ~/egcs/SunOS-4.1.3/bin/g77 -g cabs.f -lm && ./a.out
>> Fortran abort routine called
>> Abort (core dumped)
> Here is a patch that should fix this problem. I'm putting it into
> g77 0.5.24 as well. Dave Love, once this patch is confirmed as
> working by Alexandre or someone else on SunOS4, could you please
> install it?
Confirmed, it works, thanks.
--
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3
1998-06-26 13:03 ` Craig Burley
1998-06-27 19:07 ` Alexandre Oliva
@ 1998-06-27 22:14 ` Jeffrey A Law
1 sibling, 0 replies; 11+ messages in thread
From: Jeffrey A Law @ 1998-06-27 22:14 UTC (permalink / raw)
To: Craig Burley; +Cc: oliva, d.love, egcs, egcs-patches
In message < 199806261602.MAA12907@melange.gnu.org >you write:
> Here is a patch that should fix this problem. I'm putting it into
> g77 0.5.24 as well. Dave Love, once this patch is confirmed as
> working by Alexandre or someone else on SunOS4, could you please
> install it?
I took the liberty of installing this patch now instead of waiting.
I'd like it to be in the next snapshot (which will spin tonight) for
testing purposes.
If this causes problems, blame at me, not David :-)
jeff
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~1998-06-27 22:14 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1998-06-23 22:55 Results for egcs-2.91.42 19980621 (gcc2 ss-980502 experimental) testsuite on sparc-sun-sunos4.1.3 Alexandre Oliva
1998-06-24 10:08 ` Craig Burley
1998-06-24 14:50 ` Alexandre Oliva
1998-06-24 20:08 ` Jeffrey A Law
1998-06-25 6:50 ` Fred Fish
1998-06-24 21:23 ` Craig Burley
1998-06-25 6:50 ` Dave Love
1998-06-26 7:16 ` Alexandre Oliva
1998-06-26 13:03 ` Craig Burley
1998-06-27 19:07 ` Alexandre Oliva
1998-06-27 22:14 ` Jeffrey A Law
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).