Snapshot 20120214: fork: can't reserve memory for stack

[Jehan Bing]

On 2012-02-15 13:35, Corinna Vinschen wrote:
> On Feb 15 22:13, Corinna Vinschen wrote:
>> On Feb 15 14:01, Eric Blake wrote:
>>> On 02/15/2012 01:47 PM, Corinna Vinschen wrote:
>>>> On Feb 15 11:45, Jehan Bing wrote:
>>>>> I'm getting the following error starting with cygwin snapshot
>>>>> 20120214 when running a git command:
>>>>>
>>>>> $ git add bigfile-test.dll
>>>>>        0 [main] git 7440 D:\dev_apps\cygwin\bin\git.exe: *** fatal
>>>>> error in forked process - fork: can't reserve memory for stack
>>>>> 0xAD0000 - 0xBD0000, Win32 error 487
>>>>
>>>> I can't reproduce that (also on W7).  How big is bigfile-test.dll?
>>>>
>>>> Eric, does git use pthreads and fork from a thread, by any chance?
>>>
>>> /me rummages around in git.git
>>>
>>> Well, it certainly has both pthread_create and fork calls in the same
>>> files (at least run-command.c), although I'm not quire sure if it tries
>>> to actually fork from a pthread rather than from the main thread.
>>
>> Ok, thanks anyway.  I think I know what happens.  Two days ago I thought
>> it is a good idea to use the Windows-provided stack from CreateThread by
>> default (== if the application didn't provide its own).  Too bad that
>> Windows creates the stacks typically in the lower regions of the
>> processes VM.  I guess I have to revert that, so that pthread_create
>> always uses a self-created stack again.  Oh well.  So much for "good
>> ideas"...
>
> Just reverted in CVS.  Will show up in the next snapshot.

Snapshot 20120216 works.

Thanks,
	Jehan


--
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