Duration of Blocked Interrupts

[Srinivas.Narendula at libnet.com]

Mail Receiver wrote:
>
> On Tue, 28 Apr 1998, Doug Ledford wrote:
>
> > Justin T. Gibbs wrote:
> > > First of all, excuse my ignorance of Linux kernel internals.  I'm
> > > a FreeBSD hacker by trade...  But, I believe that Linux has what
> > > is called a "bottom half handler" which is somewhat similar to a
> > > software interrupt handler that can be preempted by other interrupts.
> >
> > There is one for the network devices already.  There isn't one for the
SCSI
> > drivers.  If you want one, you have to build it yourself.
>
> I've finally had a chance to investigate this a little more and I'm
> curious about this.  I've found a SCSI_BH and scsi_bottom_half_handler
(in > scsi.c) that look like they perform the exact function specified
(delayed > processing of completed commands).

There is in the 2.1.65 and later kernels.

> It appears as though simply setting use_new_eh_code to 1 in the AIC7XXX
> macro will cause the bottom half to be enabled.  All of this is from a
> cursory perusal of the code so I'd appreciate any further information
> anyone might have on this subject.

There's more to it than that.  Along with setting use_new_eh to 1, you also
need to assing some new functions (eh_abort, eh_device_reset, and
eh_bus_reset) and possibly one more new function (eh_strategy) plus some
other changes take place in the expected semantics of the other driver
entry points.  In short, it's a more drastic change than it looks like.
Additionally, prior to 2.1.103 or so, the new_eh code was fairly broken,
although now it appears to be shaping up.  So, it wasn't used in the past
for these reasons.  In the future, I will probably be moving over to the
new_eh code as it stabilizes since it already has a lot of the features I
either need from the mid level code or have to program in myself.

--

 Doug Ledford  <dledford at dialnet.net>
  Opinions expressed are my own, but
     they should be everybody's.

NOTE: I'm just now starting to recover from the Linux-Expo.  Most
      importantly, I'm just now getting around to answering a lot
      of my email.  So, if your message has lain around un-answered
      for many days, I apologize.

To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message



To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message