lkml.org 
[lkml]   [1998]   [Sep]   [27]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: PCI_LATENCY_TIMER


I did some mistakes in my previous calculations that need some fixes:
Sorry for the bandwitch.

On Sun, 27 Sep 1998, Gerard Roudier wrote:

>
> On Sun, 27 Sep 1998, Mike Black wrote:
>
> > I found the discussions concerning this to be quite interesting but not very
> > illuminating.
>
> What kind of illumination were you expecting?
>
> What about the following?
>
> > Here's my question: I've got a network server that primarily uses SCSI,
> > IDE, and a 100BaseT Network Card. All other PCI devices (e.g VGA) are not
> > used much. How do I optimize this setup for the SCSI/Network IDE/Network
> > throughput with the understanding that generally only SCSI or IDE access is
> > happening at one time??
> >
> > I see that my SCSI, IDE and Netcard report:
> > SCSI Latency: 8 min, 8 max, 64 set, cache line size 08
> > IDE Latency: 64 set
> > Network Latency: 8 min, 28 max, 64 set
>
> Let me translate this output into PCI language:
>
> SCSI:
> MIN_GNT = 8 --> 8 x 0.25 = 4 micro-seconds
2

> MAX_LAT = 8 --> 8 x 0.25 = 4 micro-seconds
2

> LATENCY_TIMER = 64 --> 64x0.030 = 1.92 micro-seconds
>
> IDE:
> LATENCY_TIMER = 64 --> 64x0.030 = 1.92 micro-seconds
>
> Network:
> MIN_GNT = 8 --> 8 x 0.25 = 4 micro-seconds
2

> MAX_LAT = 28 --> 28x 0.25 = 7 micro-seconds
> LATENCY_TIMER = 64 --> 64x0.030 = 1.92 micro-seconds
>
> If we only take into account these 3 devices, the predictable PCI BUS
> latency is 2*1.92 = 3.84 micro-seconds that fits the MAX_LAT requirement
> of the SCSI device that is the lowest value for MAX_LAT.

Changed to:
In order to achieve the SCSI device desired maximum bus latency, the PCI
system software could assign a LATENCY TIMER of 33 to PCI devices, which
is probaly a too short value at least for the SCSI device. But, the
MAX_LAT requirement of the SCSI device looks like a bogus value and
assuming the right value could be at least twice the MIN_GNT seems a lot
more realistic.

> My comments:
>
> 1 - The system software that chose a latency timer of 64 for all
> devices has not been able to fit the MAX_GNT value due to the SCSI
> controller providing it very probably _wrong_ informations, but the
> MAX_LAT requirement of all devices has been achieved.

Changed to:
The MAX_LAT value provided by the SCSI device is pain in the ass for the
PCI software.

> 2 - A device that desires to be granted 4 us for a BUS transaction and
> that want the maximum BUS latency to be at most that 4 us is kind of
> shit-maker for PCI BIOSes and PCI drivers that want to make things
> fine, unless it is required to be the unique device on a PCI BUS.

(Same remark for 2 us)

>
> > 00:0b.0 SCSI storage controller: Adaptec AIC-7880U
>
> What a great illumination I have had 4 years ago to go with Symbios
> controllers rather than Adaptec ones. ;-)

No change needed, there. ;-)

>
>
> Regards,
> Gerard.
>
>


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

\
 
 \ /
  Last update: 2005-03-22 13:44    [from the cache]
©2003-2011 Jasper Spaans