aic7xxx / AHA2940 worries... anyone?

[Doug Ledford]

--------

> It is the same in Linux.  Every completed command is handed off to the
> higher-level SCSI layer to be processed.  But we do not do this until
> after we remove all the SCBs from the QOUTFIFO, so you wouldn't
> expect this problem on a single controller.

Except that the entire time we are handing off commands to the mid level 
scsi code the interrupt routine is blocked from re-entering.  So, in that 
case, the sequencer could be busy filling up the QOUTFIFO while we are 
handing code off to the mid level code (and with very high commands per lun, 
4 drives, and bonnie runs I saw this behaviour as well on a single 
controller, but not nearly so often as would make me worry, now I'm not so 
sure that wasn't related to my slow CPU speed or slower disk speeds instead 
of a lack of a problem).

>  But, if you had 3
> different controllers, all very active, this might be a problem.
> I think we should reintroduce this into the Linux driver also.

It wouldn't block any of the other controllers unless they shared IRQs (I 
take that back, a check of the code reveals that we are running inside of a 
cli() during the isr, so yes it would block other controllers).


-- 
*****************************************************************************
* Doug Ledford                      *   Unix, Novell, Dos, Windows 3.x,     *
* dledford at dialnet.net    873-DIAL  *     WfW, Windows 95 & NT Technician   *
*   PPP access $14.95/month         *****************************************
*   Springfield, MO and surrounding * Usenet news, e-mail and shell account.*
*   communities.  Sign-up online at * Web page creation and hosting, other  *
*   873-9000 V.34                   * services available, call for info.    *
*****************************************************************************