interesting... I'll keep that in mind,

Anybody against that ugly (but probably working) trick ?

On Thu, Apr 14, 2011 at 7:09 PM, Allefant <[email protected]> wrote:
> On Thu, 2011-04-14 at 18:54 +0200, Allefant wrote:
>> On Thu, 2011-04-14 at 18:22 +0200, David Philippi wrote:
>> > Am Thursday 14 April 2011 schrieb Paul Ebermann:
>> > > This does not help on the short term, but might result in this feature
>> > > being implemented before Wesnoth 1.10 comes out.
>> >
>> > There should be a very simple workaround. Use a wrapper that sets the 
>> > variable
>> > then does an exec of the real wesnoth binary. Might be a shell script or a
>> > binary executable, whatever works best.
>> >
>>
>> If a binary can be used, then could the real wesnoth binary exec itself
>> with the modified environment (and a command-line argument preventing
>> execing itself again)?
>>
>
> Well, guess I can answer that myself:
>
> int main(int argc, char **argv) {
>    if (!getenv("OMP_WAIT_POLICY")) {
>        setenv("OMP_WAIT_POLICY", "PASSIVE", 1);
>        execv(argv[0], argv);
>    }
>    return 0;
> }
>
>
>
> _______________________________________________
> Wesnoth-dev mailing list
> [email protected]
> https://mail.gna.org/listinfo/wesnoth-dev
>

_______________________________________________
Wesnoth-dev mailing list
[email protected]
https://mail.gna.org/listinfo/wesnoth-dev

Reply via email to