Well, another code fragment might be useful:
/*--+
| Set up iocb for aio_write() call |
+--*/
memset(iocb, 0, sizeof(struct
Sorry forgot something:
the Linux way of doing this is to fill in the si_band with information
on what has happened. This sound acceptable and there is no need to
be incompatible if the idea isn't too bad.
-jason
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hacker
>And now for a wish:
[ST_AIO stuff cut]
If I understand what you are trying to say, then
when real time signals are added, this will be unnecessary.
You can get the completion of an aio_* call from the signal
queue.
-jason
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe fr
Part of my patches is a new system call, aio_waitcomplete() which waits
for an aio job to complete, and returns a pointer to the userland aiocb
associated with the job. (This is a non-standard extension of my own
creation). The patches are available from http://tfeed.maxwell.syr.edu,
link is on
Hand up...
I have two machines running heavily hit aio based file server
mechanisms. If the patches will apply to a -current system about
2 weeks old I'll give it a try... (atleast without too much trouble).
And now for a wish:
/*-
Christopher Sedore scribbled this message on Sep 16:
> > Great, now do you want to tackle aio_cancel? ;^)
>
> I've been holding off to see whether I can get the other aio patches for
> improved socket io committed. I don't want to do two versions of
> aio_cancel (the changes for sockets alter th
Christopher Sedore scribbled this message on Sep 16:
> > Great, now do you want to tackle aio_cancel? ;^)
>
> I've been holding off to see whether I can get the other aio patches for
> improved socket io committed. I don't want to do two versions of
> aio_cancel (the changes for sockets alter t
Christopher Sedore wrote:
>
> On Thu, 16 Sep 1999, Wes Peters wrote:
>
> > Christopher Sedore wrote:
> > >
> > > On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> > > >
> > > > Do you by any change have an idea how to fix PR kern/13075
> > > > (signal is not posted for async I/O on raw devices)
> > >
Christopher Sedore wrote:
>
> On Thu, 16 Sep 1999, Wes Peters wrote:
>
> > Christopher Sedore wrote:
> > >
> > > On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> > > >
> > > > Do you by any change have an idea how to fix PR kern/13075
> > > > (signal is not posted for async I/O on raw devices)
> > >
On Thu, 16 Sep 1999, Wes Peters wrote:
> Christopher Sedore wrote:
> >
> > On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> > >
> > > Do you by any change have an idea how to fix PR kern/13075
> > > (signal is not posted for async I/O on raw devices)
> >
> > Yes. There is no code to post the sig
Christopher Sedore wrote:
>
> On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> >
> > Do you by any change have an idea how to fix PR kern/13075
> > (signal is not posted for async I/O on raw devices)
>
> Yes. There is no code to post the signal unless the job is of the lio
> variety. Writing a fix
On Thu, 16 Sep 1999, Wes Peters wrote:
> Christopher Sedore wrote:
> >
> > On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> > >
> > > Do you by any change have an idea how to fix PR kern/13075
> > > (signal is not posted for async I/O on raw devices)
> >
> > Yes. There is no code to post the si
Christopher Sedore wrote:
>
> On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> >
> > Do you by any change have an idea how to fix PR kern/13075
> > (signal is not posted for async I/O on raw devices)
>
> Yes. There is no code to post the signal unless the job is of the lio
> variety. Writing a fix
I filed a followup with a patch (against 4.x, but it will probably work
just as well against 3.x, but I don't have a handy way to try it).
-Chris
On Wed, 15 Sep 1999, Mike Smith wrote:
> >
> >
> > On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> >
> > > > The aio_* stuff (I use a custom patched
I filed a followup with a patch (against 4.x, but it will probably work
just as well against 3.x, but I don't have a handy way to try it).
-Chris
On Wed, 15 Sep 1999, Mike Smith wrote:
> >
> >
> > On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> >
> > > > The aio_* stuff (I use a custom patched
>
>
> On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
>
> > > The aio_* stuff (I use a custom patched version in 4.x) offers performance
> > > advantages over select() with large numbers of descriptors. In terms of
> > > efficiency, I don't have any trouble saturating full-duplex 100mbit link
> > >
On Tue, 14 Sep 1999, Nik Clayton wrote:
> What we need is people like yourself, who are having to go through the
> learning curve, to document stuff as you're finding it out. It doesn't
> matter if your notes are a bit rough and ready, because the next person
> to use them can improve on them,
>
>
> On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
>
> > > The aio_* stuff (I use a custom patched version in 4.x) offers performance
> > > advantages over select() with large numbers of descriptors. In terms of
> > > efficiency, I don't have any trouble saturating full-duplex 100mbit link
> > >
On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> > The aio_* stuff (I use a custom patched version in 4.x) offers performance
> > advantages over select() with large numbers of descriptors. In terms of
> > efficiency, I don't have any trouble saturating full-duplex 100mbit link
> > with aio routine
On Wed, 15 Sep 1999, Ruslan Ermilov wrote:
> > The aio_* stuff (I use a custom patched version in 4.x) offers performance
> > advantages over select() with large numbers of descriptors. In terms of
> > efficiency, I don't have any trouble saturating full-duplex 100mbit link
> > with aio routin
On Tue, 14 Sep 1999, Nik Clayton wrote:
> What we need is people like yourself, who are having to go through the
> learning curve, to document stuff as you're finding it out. It doesn't
> matter if your notes are a bit rough and ready, because the next person
> to use them can improve on them, a
On Tue, Sep 14, 1999 at 04:40:37PM -0400, Christopher Sedore wrote:
>
>
> On Mon, 13 Sep 1999, Jayson Nordwick wrote:
>
> > While reading through (at least trying to... I wish there was some sort of
> > kernel documentation available, the entry fee is very high) the aio_* calls,
> > I had a few
On Tue, Sep 14, 1999 at 04:40:37PM -0400, Christopher Sedore wrote:
>
>
> On Mon, 13 Sep 1999, Jayson Nordwick wrote:
>
> > While reading through (at least trying to... I wish there was some sort of
> > kernel documentation available, the entry fee is very high) the aio_* calls,
> > I had a few
On Mon, 13 Sep 1999, Jayson Nordwick wrote:
> While reading through (at least trying to... I wish there was some sort of
> kernel documentation available, the entry fee is very high) the aio_* calls,
> I had a few questions to clear up my understanding:
>
> 1) Do they only work on files? The
On Mon, 13 Sep 1999, Jayson Nordwick wrote:
> While reading through (at least trying to... I wish there was some sort of
> kernel documentation available, the entry fee is very high) the aio_* calls,
> I had a few questions to clear up my understanding:
>
> 1) Do they only work on files? The
On Mon, Sep 13, 1999 at 10:02:40PM -0700, Jayson Nordwick wrote:
> While reading through (at least trying to... I wish there was some sort of
> kernel documentation available, the entry fee is very high)
What we need is people like yourself, who are having to go through the
learning curve, to d
On Mon, Sep 13, 1999 at 10:02:40PM -0700, Jayson Nordwick wrote:
> While reading through (at least trying to... I wish there was some sort of
> kernel documentation available, the entry fee is very high)
What we need is people like yourself, who are having to go through the
learning curve, to do
[EMAIL PROTECTED] (Jayson Nordwick) writes:
> While reading through (at least trying to... I wish there was some sort of
> kernel documentation available, the entry fee is very high) the aio_* calls,
> I had a few questions to clear up my understanding:
> 1) Do they only work on files? The on
nordw...@scam.xcf.berkeley.edu (Jayson Nordwick) writes:
> While reading through (at least trying to... I wish there was some sort of
> kernel documentation available, the entry fee is very high) the aio_* calls,
> I had a few questions to clear up my understanding:
> 1) Do they only work on fi
29 matches
Mail list logo
