In article <20070521145331.900A4D2B88F@mail.connecty.lt>,
Mindaugas R. <rmind@NetBSD.org> wrote:
>Scott Ellis <scotte@warped.com> wrote:
>> Is this an oversight in NetBSD, or does si_value (specifically 
>> si_value.sival_int) map to something else?
>Well, it is not a problem of AIO support, but signaling part, which is
>actively used with AIO. Yes, there should be si_value member in siginfo.
>
>I will take a look at this, when time permits.

We call it si_sigval... I fixed it.

christos