From: Mike Holden <rockbox_at_mikeholden.uklinux.net>
Date: Sat, 12 Apr 2003 20:26:56 +0100 (BST)

Björn Stenberg said:
> Mike Holden wrote:
>> My remaining issue is "Silent Play".
>
> I believe silent play is an mpeg issue (misinterpreting vbr headers etc)
> and not a disk reading issue. I also haven't seen any reports about this
> since Linus fixed some problems in the mpeg code a while ago. Feel free
> to brief us if this happens to you.

Yes, I have still been seeing silent play. It ALWAYS happens towards the
end of a track, and continues into the beginning of the next track.

I see that a fix has just gone in for byte swapping, so I will investigate
that, and report back if that doesn't fix it!

>> The code comment has 5 * nop, and the comments says 400ms. Where did
>> this figure come from? Is it the ATA spec? How long does a nop take,
>
> Yes, the 400ns is defined in the PIO protocol part of the ATA
> specification:
>
> "HPIOI1: Check_Status State:
> This state is entered when the host has written a PIO data-in command to
> the device and nIEN is set to one, or when INTRQ is asserted. When in
> this state, the host shall read the device Status register. When
> entering this state from the HI4 state, the host shall wait 400 ns
> before reading the Status register."
>
> A nop instruction takes one clock cycle, and since the cpu is clocked at
> 12 MHz each cycle is 83ns long.

Fair enough. Makes sense. Of course the FM has a slower CPU (11Mhz), so a
NOP is therefore more like 90ns! Can't see that making any difference
really.

How much leeway do you get on the 400ns? Is it any time after that period?
I assume it must be - it's just a minimum period to wait to give the
hardware time to perform the operation and populate the register.

-- 
Mike Holden

Received on 2003-04-12