Re: Question on KPI / NKE : sys/select.h : struct selinfo ?
Re: Question on KPI / NKE : sys/select.h : struct selinfo ?
- Subject: Re: Question on KPI / NKE : sys/select.h : struct selinfo ?
- From: Josh Graessley <email@hidden>
- Date: Thu, 5 May 2005 09:42:18 -0700
On May 5, 2005, at 8:48 AM, chas williams - CONTRACTOR wrote:
In message <email@hidden>,Jim
Magee writes:
Now that we have discovered a couple of drivers that could otherwise
be KPI-conforming if not for this, we will be creating a KPI to
allocate a selinfo structure dynamically. In the meantime, you can
just allocate memory large to be safe. Future versions of selinfo
will be smaller than the current one. But to be safe. at least
doubling the size of the current selinfo struct should suffice (to
say 64 bytes). You should have no reason to look inside the
structure.
how about access to kernel_flock? i suspect one will need to get
the kernel funnel if you are going to issue a selwakeup().
The kernel_flock was part of the old split funnel. The funnels are
gone. You are responsible for locking your own data structures. The
kernel uses a variety of locks to protect it's own internal data
structures.
any plans for making a kpi version of net_add_domain and
net_add_proto?
There are no plans at this time for net_add_domain and net_add_proto.
We are interested in hearing about what you might need such
functionality for.
-josh
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-kernel mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden