* fork problem on latest cygwin CVS
@ 2011-11-14 14:40 marco atzeri
2011-11-14 15:31 ` Christopher Faylor
0 siblings, 1 reply; 9+ messages in thread
From: marco atzeri @ 2011-11-14 14:40 UTC (permalink / raw)
To: cygwin
Hi Cgf,
after updating cygwin CVS build from
http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00063.html
to
http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00064.html .. 66html
on CYGWIN_NT-6.1-WOW64
the shell have fork issues during make/build session.
1 [main] sh 6768 E:\cygwin\bin\sh.exe: *** fatal error in forked
process - fork: can't commit memory for stack 0x28C000(81920), Win32
error 487
Stack trace:
Frame Function Args
0028EB88 6102F6AC (FFFFFFFE, 0028EBA4, 0028EBEC, 768D0DAE)
0028EE78 6102F6AC (6119AF60, 00008000, 00000000, 6119CAEF)
0028FEA8 61005E2C (6119CB70, 0028FED4, 0028C000, 0028C000)
0028FEC8 61005E68 (6119CB70, 0028C000, 00014000, 00000040)
0028FEF8 6100610F (008F102E, 0028FF00, 00001F80, 037F037F)
0028FF28 6100616C (008F102E, 00000000, 00000000, 00000000)
0028FF48 610061EB (00401F10, 00000000, 0028FF88, 00401033)
0028FF58 00465672 (00401F10, 00000000, 00000000, 00000000)
0028FF88 00401033 (7EFDE000, 0028FFD4, 779A9ED2, 7EFDE000)
0028FF94 767333CA (7EFDE000, 76CB068D, 00000000, 00000000)
0028FFD4 779A9ED2 (00401000, 7EFDE000, 00000000, 00000000)
0028FFEC 779A9EA5 (00401000, 7EFDE000, 00000000, 78746341)
End of stack trace
100 [main] sh 6872 fork: child -1 - forked process died
unexpectedly, retry 10, exit code 256, errno 11
/bin/sh: fork: retry: Resource temporarily unavailable
rebaseall does not solve the problem.
Regards
Marco
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: fork problem on latest cygwin CVS
2011-11-14 14:40 fork problem on latest cygwin CVS marco atzeri
@ 2011-11-14 15:31 ` Christopher Faylor
2011-11-14 16:09 ` marco atzeri
0 siblings, 1 reply; 9+ messages in thread
From: Christopher Faylor @ 2011-11-14 15:31 UTC (permalink / raw)
To: cygwin
On Mon, Nov 14, 2011 at 03:40:32PM +0100, marco atzeri wrote:
>Hi Cgf,
>
>
>after updating cygwin CVS build from
> http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00063.html
>to
> http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00064.html .. 66html
>
>on CYGWIN_NT-6.1-WOW64
>
>the shell have fork issues during make/build session.
Come on, you know the drill by now. How about a simple test case?
cgf
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: fork problem on latest cygwin CVS
2011-11-14 15:31 ` Christopher Faylor
@ 2011-11-14 16:09 ` marco atzeri
2011-11-14 17:11 ` Christopher Faylor
0 siblings, 1 reply; 9+ messages in thread
From: marco atzeri @ 2011-11-14 16:09 UTC (permalink / raw)
To: cygwin
On 11/14/2011 4:31 PM, Christopher Faylor wrote:
> On Mon, Nov 14, 2011 at 03:40:32PM +0100, marco atzeri wrote:
>> Hi Cgf,
>>
>>
>> after updating cygwin CVS build from
>> http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00063.html
>> to
>> http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00064.html .. 66html
>>
>> on CYGWIN_NT-6.1-WOW64
>>
>> the shell have fork issues during make/build session.
>
> Come on, you know the drill by now. How about a simple test case?
>
> cgf
>
I know the drill, but I just updated from XP-32 to W7-64,
one disk crashed, so I am reinstalling everything from scratch and
taking a test case when building octave is never easy.
So I need time to give you a simple test case, and eventually
someone else will notice the same problem in advance.
FYI 2011-11-08 snapshot does not suffer such problem.
Thanks
Marco
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: fork problem on latest cygwin CVS
2011-11-14 16:09 ` marco atzeri
@ 2011-11-14 17:11 ` Christopher Faylor
2011-11-15 15:05 ` marco atzeri
0 siblings, 1 reply; 9+ messages in thread
From: Christopher Faylor @ 2011-11-14 17:11 UTC (permalink / raw)
To: cygwin
On Mon, Nov 14, 2011 at 05:09:07PM +0100, marco atzeri wrote:
>On 11/14/2011 4:31 PM, Christopher Faylor wrote:
>> On Mon, Nov 14, 2011 at 03:40:32PM +0100, marco atzeri wrote:
>>> Hi Cgf,
>>>
>>>
>>> after updating cygwin CVS build from
>>> http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00063.html
>>> to
>>> http://cygwin.com/ml/cygwin-cvs/2011-q4/msg00064.html .. 66html
>>>
>>> on CYGWIN_NT-6.1-WOW64
>>>
>>> the shell have fork issues during make/build session.
>>
>> Come on, you know the drill by now. How about a simple test case?
>
>I know the drill, but I just updated from XP-32 to W7-64,
>one disk crashed, so I am reinstalling everything from scratch and
>taking a test case when building octave is never easy.
>
>So I need time to give you a simple test case, and eventually
>someone else will notice the same problem in advance.
>
>FYI 2011-11-08 snapshot does not suffer such problem.
FYI, the last thing any developer wants to hear after a major code
change is a generic "It's broke" report with no details and no way to
duplicate the problem. A stack trace from a home-grown version of
cygwin1.dll is not a detail. It's meaningless unless the addresses are
decoded.
cgf
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: fork problem on latest cygwin CVS
2011-11-14 17:11 ` Christopher Faylor
@ 2011-11-15 15:05 ` marco atzeri
2011-11-15 19:22 ` Christopher Faylor
0 siblings, 1 reply; 9+ messages in thread
From: marco atzeri @ 2011-11-15 15:05 UTC (permalink / raw)
To: cygwin
On 11/14/2011 6:11 PM, Christopher Faylor wrote:
> FYI, the last thing any developer wants to hear after a major code
> change is a generic "It's broke" report with no details and no way to
> duplicate the problem. A stack trace from a home-grown version of
> cygwin1.dll is not a detail. It's meaningless unless the addresses are
> decoded.
>
> cgf
Ok,
rewinding to step 1
I simplified the test case to a short one, involving just 1 make call,
using your snapshot
$ uname -a
CYGWIN_NT-6.1-WOW64 MARCOATZERI 1.7.10s(0.255/5/3) 20111114 02:01:52
i686 Cygwin
please find it on:
http://matzeri.altervista.org/works/test_case/
calling
make oct-conf.h
produces:
--------------------------------------------------------------------
making oct-conf.h from oct-conf.h.in
0 [main] sh 4636 E:\cygwin\bin\sh.exe: *** fatal error in forked
process - fork: can't commit memory for stack 0x28C000(81920), Win32
error 487
Stack trace:
Frame Function Args
0028EBB8 6102B30B (0028EBB8, 00000000, 00000000, 00000000)
0028EEA8 6102B30B (61192F60, 00008000, 00000000, 61194A6C)
0028FED8 61005403 (61194FB0, 0028FEF4, 0028FF08, 61006BE1)
0028FEE8 61005438 (61194FB0, 0028C000, 00014000, 00000040)
0028FF08 61006BE1 (007F102E, 0028FF10, 00001F80, 037F037F)
0028FF38 61006C2D (007F102E, 00000000, 0028FF58, 00465672)
0028FF48 61006CAB (00401F10, 00000000, 0028FF88, 00401033)
0028FF58 00465672 (00401F10, 00000000, 00000000, 00000000)
0028FF88 00401033 (7EFDE000, 0028FFD4, 770A9ED2, 7EFDE000)
0028FF94 750E33CA (7EFDE000, 7660A4F7, 00000000, 00000000)
0028FFD4 770A9ED2 (00401000, 7EFDE000, 00000000, 00000000)
0028FFEC 770A9EA5 (00401000, 7EFDE000, 00000000, 78746341)
End of stack trace
0 [main] sh 300 fork: child -1 - forked process died
unexpectedly, retry 10, exit code 256, errno 11
/bin/sh: fork: retry: Resource temporarily unavailable
...
-----------------------------------------------------------------------
this does not happen with 20111108
Regards
Marco
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: fork problem on latest cygwin CVS
2011-11-15 15:05 ` marco atzeri
@ 2011-11-15 19:22 ` Christopher Faylor
2011-11-16 19:14 ` Christopher Faylor
0 siblings, 1 reply; 9+ messages in thread
From: Christopher Faylor @ 2011-11-15 19:22 UTC (permalink / raw)
To: cygwin
On Tue, Nov 15, 2011 at 04:04:16PM +0100, marco atzeri wrote:
>On 11/14/2011 6:11 PM, Christopher Faylor wrote:
>
>> FYI, the last thing any developer wants to hear after a major code
>> change is a generic "It's broke" report with no details and no way to
>> duplicate the problem. A stack trace from a home-grown version of
>> cygwin1.dll is not a detail. It's meaningless unless the addresses are
>> decoded.
>>
>> cgf
>
>Ok,
>rewinding to step 1
>
>I simplified the test case to a short one, involving just 1 make call,
>using your snapshot
Amazingly short. Thanks very much! I can duplicate this and will fix it.
cgf
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: fork problem on latest cygwin CVS
2011-11-15 19:22 ` Christopher Faylor
@ 2011-11-16 19:14 ` Christopher Faylor
2011-11-16 22:01 ` marco atzeri
0 siblings, 1 reply; 9+ messages in thread
From: Christopher Faylor @ 2011-11-16 19:14 UTC (permalink / raw)
To: cygwin
On Tue, Nov 15, 2011 at 02:22:02PM -0500, Christopher Faylor wrote:
>On Tue, Nov 15, 2011 at 04:04:16PM +0100, marco atzeri wrote:
>>On 11/14/2011 6:11 PM, Christopher Faylor wrote:
>>
>>> FYI, the last thing any developer wants to hear after a major code
>>> change is a generic "It's broke" report with no details and no way to
>>> duplicate the problem. A stack trace from a home-grown version of
>>> cygwin1.dll is not a detail. It's meaningless unless the addresses are
>>> decoded.
>>
>>Ok,
>>rewinding to step 1
>>
>>I simplified the test case to a short one, involving just 1 make call,
>>using your snapshot
>
>Amazingly short. Thanks very much! I can duplicate this and will fix
>it.
I've made some changes and can no longer duplicate the problem from your
test case. If you can confirm that the latest snapshot works for you I
would appreciate it.
As it turns out, this was a change I was planning on making "at some
point in the future". The future, apparently, is now.
cgf
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: fork problem on latest cygwin CVS
2011-11-16 19:14 ` Christopher Faylor
@ 2011-11-16 22:01 ` marco atzeri
2011-11-16 22:24 ` Christopher Faylor
0 siblings, 1 reply; 9+ messages in thread
From: marco atzeri @ 2011-11-16 22:01 UTC (permalink / raw)
To: cygwin
On 11/16/2011 8:14 PM, Christopher Faylor wrote:
> On Tue, Nov 15, 2011 at 02:22:02PM -0500, Christopher Faylor wrote:
>> On Tue, Nov 15, 2011 at 04:04:16PM +0100, marco atzeri wrote:
>>> On 11/14/2011 6:11 PM, Christopher Faylor wrote:
>>>
>>>> FYI, the last thing any developer wants to hear after a major code
>>>> change is a generic "It's broke" report with no details and no way to
>>>> duplicate the problem. A stack trace from a home-grown version of
>>>> cygwin1.dll is not a detail. It's meaningless unless the addresses are
>>>> decoded.
>>>
>>> Ok,
>>> rewinding to step 1
>>>
>>> I simplified the test case to a short one, involving just 1 make call,
>>> using your snapshot
>>
>> Amazingly short. Thanks very much! I can duplicate this and will fix
>> it.
>
> I've made some changes and can no longer duplicate the problem from your
> test case. If you can confirm that the latest snapshot works for you I
> would appreciate it.
>
> As it turns out, this was a change I was planning on making "at some
> point in the future". The future, apparently, is now.
>
> cgf
>
CYGWIN_NT-6.1-WOW64 1.7.10s(0.255/5/3) 20111116 04:41:29 i686 Cygwin
passed my crash test:
a full build of octave binary from scratch plus the make check.
Regards
Marco
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: fork problem on latest cygwin CVS
2011-11-16 22:01 ` marco atzeri
@ 2011-11-16 22:24 ` Christopher Faylor
0 siblings, 0 replies; 9+ messages in thread
From: Christopher Faylor @ 2011-11-16 22:24 UTC (permalink / raw)
To: cygwin
On Wed, Nov 16, 2011 at 11:01:16PM +0100, marco atzeri wrote:
>On 11/16/2011 8:14 PM, Christopher Faylor wrote:
>> On Tue, Nov 15, 2011 at 02:22:02PM -0500, Christopher Faylor wrote:
>>> On Tue, Nov 15, 2011 at 04:04:16PM +0100, marco atzeri wrote:
>>>> On 11/14/2011 6:11 PM, Christopher Faylor wrote:
>>>>
>>>>> FYI, the last thing any developer wants to hear after a major code
>>>>> change is a generic "It's broke" report with no details and no way to
>>>>> duplicate the problem. A stack trace from a home-grown version of
>>>>> cygwin1.dll is not a detail. It's meaningless unless the addresses are
>>>>> decoded.
>>>>
>>>> Ok,
>>>> rewinding to step 1
>>>>
>>>> I simplified the test case to a short one, involving just 1 make call,
>>>> using your snapshot
>>>
>>> Amazingly short. Thanks very much! I can duplicate this and will fix
>>> it.
>>
>> I've made some changes and can no longer duplicate the problem from your
>> test case. If you can confirm that the latest snapshot works for you I
>> would appreciate it.
>>
>> As it turns out, this was a change I was planning on making "at some
>> point in the future". The future, apparently, is now.
>>
>> cgf
>>
>
>CYGWIN_NT-6.1-WOW64 1.7.10s(0.255/5/3) 20111116 04:41:29 i686 Cygwin
>
>passed my crash test:
>a full build of octave binary from scratch plus the make check.
Phew.
Thanks for confirming.
cgf
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2011-11-16 22:24 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-11-14 14:40 fork problem on latest cygwin CVS marco atzeri
2011-11-14 15:31 ` Christopher Faylor
2011-11-14 16:09 ` marco atzeri
2011-11-14 17:11 ` Christopher Faylor
2011-11-15 15:05 ` marco atzeri
2011-11-15 19:22 ` Christopher Faylor
2011-11-16 19:14 ` Christopher Faylor
2011-11-16 22:01 ` marco atzeri
2011-11-16 22:24 ` Christopher Faylor
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).