[Zope-dev] Zope/Python crashing

Myroslav Opyr myroslav@zope.net.ua
Tue, 21 May 2002 02:04:15 +0300


Matthew T. Kromer wrote:

> Myroslav Opyr wrote:
>
>> For FreeBSD there is another pthreads related bug in Python. And=20
>> quick-patch from=20
>> https://sourceforge.net/tracker/?func=3Ddetail&aid=3D554841&group_id=3D=
5470&atid=3D305470=20
>> with THREAD_STACK_SIZE=3D0x20000 solves the problem. But it is quick=20
>> and dirty hack. Will it be solved in next bugfix release of Python?=20
>> Whom/where I have to contact to have progress regarding the issue?
>>
>> I'll post comment into SF tracker because site which crashed=20
>> previously works fine now (for 5 days already)... Is that sufficient?=20
>> This dirty patch is not acceptable and clean solution have to take=20
>> place. Should I take part and try to develop one?
>
> Well, the way you start threading with pthreads requres a single=20
> initializer for the stack size; once you have initialized the=20
> threading environment you cannot change the stacksize later, to the=20
> best of my knowledge.=20

Ok. What general solution should be? One of the possible ways would be=20
to reserve big stack enough for all cases but only small portion of it=20
(generally used by apps) turn into phisical memory. It should be default=20
behavior for majority of OSes/Libs, shouldn't it?

> I suppose you could finesse a new python startup argument, or=20
> environment variable check.=20

+1 for startup argument
-1 for environment

> I think the SF tracker is the best way to approach that.=20

You mean discussion about the issue shoud take place in tracker?

m.
--=20
Myroslav Opyr
zope.net.ua <http://zope.net.ua/> =B0 Ukrainian Zope Hosting
e-mail: myroslav@zope.net.ua <mailto:myroslav@zope.net.ua>
cell: +380 50.3174578