On Sun, Dec 23, 2001 at 09:57 -0800, Matthew Dillon wrote:
> [ the sio.c FIFO commit ]
>     Ha!  My original commit was correct.  My bruce-backout commit was
>     flawed.  Wonderful.  What do we do now?

Correct the patch to switch between 14 and 8 bytes trigger level
while leaving the default at 14 (the previous behaviour) and
providing a sysctl to lower it to 8 -- and maybe point to the
sysctl in the overflow printf.

This would leave the cannary for increased latency intact.  It
will continue to work if latency is (again) where it's meant to
be.  It will cause as little interrupt load as necessary for
the regular case.  Plus it will make those setups bitten by
the overflow work without recompiling a kernel (read "even
problematic cases work out of the box with a one line addition
to /etc/sysctl.conf").

What else could one want?  I understand that the extra cost of
a sysctl applies only when querying or setting it.  Referencing
the value is just a normal variable access.  Please tell me if
I'm wrong.

virtually yours   82D1 9B9C 01DC 4FB4 D7B4  61BE 3F49 4F77 72DE DA76
Gerhard Sittig   true | mail -s "get gpg key" [EMAIL PROTECTED]
     If you don't understand or are scared by any of the above
             ask your parents or an adult to help you.

To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message

Reply via email to