chicken-hackers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Chicken-hackers] NetBSD w/ 4.8.0rc3


From: Timo Myyrä
Subject: Re: [Chicken-hackers] NetBSD w/ 4.8.0rc3
Date: Sat, 15 Sep 2012 09:31:15 +0300

Hi,

I tested this on my OpenBSD 5.2-current (amd64).
I run the symbolgc-tests as follows:
perl -e "system('./symbolgc-tests -:w > /dev/null') for (1 .. 100);"

And got segfaul on roughly each run except I also got 3 of these:

[panic] out of memory - cannot allocate vector (heap resizing
disabled) - execution terminated

...more...
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: get-output-string
symbolgc-tests.scm:17: string->symbol
symbolgc-tests.scm:17: open-output-string
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: get-output-string
symbolgc-tests.scm:17: string->symbol
symbolgc-tests.scm:17: open-output-string
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: display
symbolgc-tests.scm:17: get-output-string
symbolgc-tests.scm:17: string->symbol
symbolgc-tests.scm:17: open-output-string       <--

Timo

On Sat, Sep 15, 2012 at 1:04 AM, Derrell Piper <address@hidden> wrote:
> Jim Ursetto <address@hidden> writes:
>
> You're right.  I'm seeing the same thing on 10.8.1 though even more
> frequently (note that's out of 100, not 1000):
>
> fluffy533% clang -v
> Apple clang version 4.0 (tags/Apple/clang-421.0.60) (based on LLVM 3.1svn)
> Target: x86_64-apple-darwin12.1.0
> Thread model: posix
>
> fluffy532% for i in $(seq 1 100); do ./symbolgc-tests -:w; done > /dev/null
> Segmentation fault: 11
> Segmentation fault: 11
> Segmentation fault: 11
> Segmentation fault: 11
> Segmentation fault: 11
> Segmentation fault: 11
> Segmentation fault: 11
> Segmentation fault: 11
>
> Seems to be anywhere from 2-8 out of 100 here.
>
> FYI,
>
> NetBSD x86-64's fine w/ clang 3.1.  No failures over 1000 tests.
>
> Derrell
>
>> Try running the test a few thousand times with your existing chicken.
>> Just compile symbolgc-tests.scm with
>>
>>   csc symbolgc-tests.scm
>>
>> and run with something like:
>>
>>   for i in $(seq 1 1000); do ./symbolgc-tests -:w; done > /dev/null
>>
>> I'm getting this problem now with clang 4.0 (on OS X 64-bit), but only
>> 1-2 times per 1,000 runs.
>>
>> On Sep 14, 2012, at 2:48 PM, Derrell Piper wrote:
>>
>>> Mario Domenech Goulart <address@hidden> writes:
>>>
>>>> Hi Derrell
>>>>
>>>> On Fri, 14 Sep 2012 11:59:19 -0700 Derrell Piper <address@hidden> wrote:
>>>>
>>>>> FYI,
>>>>>
>>>>> I also have NetBSD systems.  I decided to check 4.8.0rc3 and gcc builds
>>>>> and completes 'make check'.  Unfortunately, clang doesn't.  It builds
>>>>> and installs fine, but 'make check' gives me this:
>>>>>
>>>>> clang version 3.0 (tags/RELEASE_30/final)
>>>>> Target: x86_64--netbsd
>>>>> Thread model: posix
>>>>>
>>>>> ======================================== symbol-GC tests ...
>>>>> ../chicken symbolgc-tests.scm -output-file a.c -verbose -include-path ..
>>>>> clang a.c -o a.o -c -fno-strict-aliasing -fwrapv -DHAVE_CHICKEN_CONFIG_H
>>>>> -DC_ENABLE_PTABLES -Os -fomit-frame-pointer
>>>>> -I.. -I"/usr/local/include/chicken"
>>>>> rm a.c
>>>>> clang a.o -o a.out -L.. -L"/usr/local/lib"  -Wl,-R"/usr/local/lib" 
>>>>> -lchicken -lm -lpthread
>>>>> rm a.o
>>>>> starting with 1490 symbols
>>>>> interning 10000 symbols ...
>>>>> [1]   Segmentation fault (core dumped) ./a.out -:w
>>>>> gmake[1]: *** [check] Error 139
>>>>> gmake[1]: Leaving directory `/home/ddp/src/chicken-4.8.0rc3'
>>>>> gmake: *** [check] Error 2
>>>>>
>>>>> I should probably try to find a newer version fo clang.  3.0 isn't
>>>>> current.
>>>>
>>>> I could build 4.8.0rc3 on Linux with clang 3.0.  It is possible that you
>>>> hit #724 (http://bugs.call-cc.org/ticket/724).  Could you please run
>>>> "make check" a couple more times to check if the behavior is
>>>> intermittent?
>>>>
>>>> Best wishes.
>>>> Mario
>>>
>>> I will if I can figure out how to downgrade clang.  I sup'd the pkgsrc
>>> already.  Sure looks the same from the ticket description.  Maybe it'd
>>> be more useful to re-run 'check' some more on clang 3.1 though.  3.0
>>> *is* pretty old.  I also should have added that this is on x86-64.
>>>
>>> Derrell
>>>
>>>
>>> _______________________________________________
>>> Chicken-hackers mailing list
>>> address@hidden
>>> https://lists.nongnu.org/mailman/listinfo/chicken-hackers
>
>
> _______________________________________________
> Chicken-hackers mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/chicken-hackers



reply via email to

[Prev in Thread] Current Thread [Next in Thread]