From: Sean Knox (seanobstacle9.com)
Date: Thu Jun 02 2005 - 00:47:43 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hey Bob,

Thanks for the info. I originally asked as I'm seeing between 80 and 90
percent interrupts on a gigabit firewall with some em(4) cards. I think
my issue may be expected given the scenario, so I'll pose that question
to the group in a different thread.

thanks,
sk

Bob Beck wrote:
> the idle loop problem will affect any driver that uses
> tsleep where stuff might need to be serviced from the idle loop.
>
> the bge booboo I found and fixed earlier with krw was
> that of it not testing correctly if interrupts were for itself
> in the shared interrupt case. totally different animal.
>
> -bob
>
>
> * Sean Knox <seanobstacle9.com> [2005-06-01 18:57]:
>
>>Marco Peereboom wrote:
>>
>>
>>>I remember that there was a boo boo in the bge interrupt handler.
>>>beck found it and I believe krw fixed it. If you can you should try
>>>something newer, like -current or whenever brad the latest releases
>>>3.7 errata that includes the "idle loop fix".
>>
>>Does this affect other NIC chipsets as well? Bob's message on kerneltrap
>> said the idle loop adjustment affects "everything on i386".
>>
>>sk

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]