[Gc] Failing test of gc 7.1 on Solaris 10 amd64

Mon Jan 16 00:16:49 PST 2012

Hi Ivan,

Am 16.01.2012 um 07:37 schrieb Ivan Maidanski:
> Please try the recent snapshot from Git release (or master) branch.

One minor issue on the docs: The description
  http://www.hpl.hp.com/personal/Hans_Boehm/gc/
should have
  ln -s  ../libatomic_ops bdwgc/libatomic_ops
instead of
  ln -s  D/libatomic_ops D/bdwgc/libatomic_ops
and libatomic_ops needs also to be bootstrapped with autoreconf/automake.

The testsuite now runs fine from GIT on Solaris 10 amd64 64 bit:

> gmake  check-TESTS
> gmake[2]: Entering directory `/home/dam/tmp/libgc/bdwgc'
> SUCCEEDED
> PASS: cordtest
> Switched to incremental mode
> Emulating dirty bits with mprotect/signals
> Completed 1 tests
> Allocated 747393 collectable objects
> Allocated 202 uncollectable objects
> Allocated 1250000 atomic objects
> Allocated 21760 stubborn objects
> Finalized 2206/2206 objects - finalization is probably ok
> Total number of bytes allocated is 104880884
> Final heap size is 17858560 bytes
> Collector appears to work
> PASS: gctest
> Leaked composite object at 444fe0 (tests/leak_test.c:12, sz=4, NORMAL)
> 
> Leaked composite object at 445ee0 (tests/leak_test.c:19, sz=9, NORMAL)
> 
> Leaked composite object at 445f20 (tests/leak_test.c:19, sz=10, NORMAL)
> 
> Leaked composite object at 445f60 (tests/leak_test.c:19, sz=11, NORMAL)
> 
> Leaked composite object at 445fa0 (tests/leak_test.c:19, sz=12, NORMAL)
> 
> Leaked composite object at 444ef0 (tests/leak_test.c:19, sz=8, NORMAL)
> 
> Leaked composite object at 444f20 (tests/leak_test.c:19, sz=4, NORMAL)
> 
> Leaked composite object at 444f50 (tests/leak_test.c:19, sz=5, NORMAL)
> 
> Leaked composite object at 444f80 (tests/leak_test.c:19, sz=6, NORMAL)
> 
> Leaked composite object at 444fb0 (tests/leak_test.c:19, sz=7, NORMAL)
> 
> PASS: leaktest
> Final heap size is 131072
> PASS: middletest
> GC_check_heap_block: found smashed heap objects:
>  452fe8 in or near object at 452fc0 (tests/smash_test.c:22, sz=40)
> 
> GC_check_heap_block: found smashed heap objects:
>  452fe8 in or near object at 452fc0 (tests/smash_test.c:22, sz=40)
> 
> GC_check_heap_block: found smashed heap objects:
>  4c8a98 in or near object at 4c8a70 (tests/smash_test.c:22, sz=40)
>  452fe8 in or near object at 452fc0 (tests/smash_test.c:22, sz=40)
> 
> PASS: smashtest
> PASS: hugetest
> Heap size: 65536
> Heap size: 131072
> PASS: realloc_test
> PASS: staticrootstest
> Unthreaded disclaim test.
> PASS: disclaim_test
>                         fin. ratio       time/s    time/fin.
> regular finalization:       0.9922         8.38  2.01368e-06
>  finalize on reclaim:       0.9999         1.41  3.36188e-07
>      no finalization:       0.0000         0.56          N/A
> PASS: disclaim_bench
> ===================
> All 10 tests passed
> ===================
> gmake[2]: Leaving directory `/home/dam/tmp/libgc/bdwgc'
> gmake[1]: Leaving directory `/home/dam/tmp/libgc/bdwgc'

> As I recall there should be  a workaround for it (just disabling PROC_VDB). It would be good if someone find out what's wrong with it on Solaris 10 (I've failed to do it).

Anything I should specifically test?

How about a release with the much more functional version from GIT?

Best regards

  -- Dago

> Regards.
> 
> 16 01 2012, 01:03 Dagobert Michelsen <dam at opencsw.org>:
>> Hi,
>> 
>> I am currently working on a package for OpenCSW and got a failing test on
>> Solaris 10 amd64:
>> 
>>> gmake[2]: Entering directory `/home/dam/mgar/pkg/libgc/trunk/work/solaris9-i386/build-isa-amd64/gc-7.1'
>>> Switched to incremental mode
>>> Reading dirty bits from /proc
>>> /bin/bash: line 4: 10134 Segmentation Fault      (core dumped) ${dir}$tst
>>> FAIL: gctest
>> 
>> The stacktrace looks like this:
>> 
>>> unstable10x% pstack ./work/solaris9-i386/build-isa-amd64/gc-7.1/core
>>> core './work/solaris9-i386/build-isa-amd64/gc-7.1/core' of 10134:       /home/dam/mgar/pkg/libgc/trunk/work/solaris9-i386/build-isa-amd64/gc-7
>>> 0000000000402a5a check_ints () + 2a
>>> 0000000000402f4b reverse_test () + 26b
>>> 0000000000404284 run_one_test () + 584
>>> 00000000004046f0 main () + 40
>>> 000000000040242c ???????? ()
>> 
>> The testsuite runs fine on
>>  Solaris 9 Sparc 32 bit
>>  Solaris 9 Sparc 64 bit
>>  Solaris 9 i386 32 bit
>> Just Solaris 10 amd64 64 bit is failing.
>> 
>> Please let me know if you need more information. I can also provide an
>> account on the buildfarm for further tests.
>> 
>> Best regards
>> 
>>  -- Dago
>> 
>> --
>> "You don't become great by trying to be great, you become great by wanting to do something,
>> and then doing it so hard that you become great in the process." - xkcd #896
>> 
>> _______________________________________________
>> Gc mailing list
>> Gc at linux.hpl.hp.com
>> http://www.hpl.hp.com/hosted/linux/mail-archives/gc/
>> 
> 
> _______________________________________________
> Gc mailing list
> Gc at linux.hpl.hp.com
> http://www.hpl.hp.com/hosted/linux/mail-archives/gc/

-- 
"You don't become great by trying to be great, you become great by wanting to do something,
and then doing it so hard that you become great in the process." - xkcd #896