Subject: Re: is netbsd resolver reentrant?
To: None <tech-userlevel@netbsd.org>
From: Christos Zoulas <christos@zoulas.com>
List: tech-userlevel
Date: 02/20/2004 22:18:09
In article <x7brnt77l8.fsf@capsicum.wsrcc.com>,
Wolfgang S. Rupprecht <wolfgang+gnus20040220T133734@dailyplanet.dontspam.wsrcc.com> wrote:
>
>christos@zoulas.com (Christos Zoulas) writes:
>> In article <x7wu6h7ccj.fsf@capsicum.wsrcc.com>,
>> Wolfgang S. Rupprecht
><wolfgang+gnus20040220T114511@dailyplanet.dontspam.wsrcc.com> wrote:
>>>
>>>Is the netbsd-current resolver
>>
>> no
>>
>> (and libc for that matter) reentrant?
>>
>> portions of it are, according to each functions spec.
>
>Is there something in the thread infrastructure that would prevent the
>non-reentrant functions from being called reentrantly or can threaded
>programs just be expected to blow up when a non-reentrant function
>gets called in a nested fashion from different threads?

They are expected to blow up. If a threaded program needs to call
a non-reentrant API it needs to make sure that it does not re-enter
it by protecting it with a mutex.

christos