RE: cciss updates for 2.6.6xxx [1/2]

[Miller, Mike (OS Dev)]

I like the idea of capping max commands based on the number of arrays. One problem is that we can add or remove a logical drive during runtime. How would Linux handle us reshuffling the max commands for each queue?

mikem

-----Original Message-----
From: Jeff Garzik [mailto:jgarzik@xxxxxxxxx]
Sent: Wednesday, April 07, 2004 11:34 AM
To: Miller, Mike (OS Dev)
Cc: alpm@xxxxxxxx; axboe@xxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
Subject: Re: cciss updates for 2.6.6xxx [1/2]


Miller, Mike (OS Dev) wrote:
> Yep, you're right. I just regurgitated the same code. I'll pull my head out and try again :(


The easiest thing to do may be to take your patch #1, and then add code 
to clamp the per-queue outstanding-command (tag) depth to
	1024 / n_arrays_found

at initialization time.  Or perhaps s/n_arrays_found/max_arrays_per_hba/

I bet that's just a few additional lines of code, and should work...

Regards,

	Jeff



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/