re: S4882 support

[YhLu]

Eric,

After check 84/a4/c4, 88/a8/c8 of function 1 of K8 northbridge, the link
speed/width is not init properly.

Regards

YH

-----邮件原件-----
发件人: [e-mail] [mailto:[e-mail]] 
发送时间: 2004年3月23日 23:42
收件人: YhLu
抄送: Stefan Reinauer; ron minnich; [e-mail]
主题: Re: S4882 support

YhLu  writes:
> Eric,
>
> The ht reset in auto.c seems not thoroughly.
>
> And it will stuck at hard_reset in amdk8_scan_chains. ( at this time can

not
> access 8111 yet).

Correct.  The early hypertransport initialization code is not
complete in the 4P case.  That incompleteness triggers the later
fixups.

I could not test that case, and there were at least some small
difficulties in that it was not obvious how the 4P case should
be extended to work properly.

Why you are getting stuck in amdk8_scan_chains I do not currently
understand yet.  I suspect Stefan's suggestion of a delayed reset
is the appropriate course to implement.

Eric

S4882 support

[Li-Ta Lo]

On Wed, 2004-03-24 at 10:43, YhLu wrote:
> Eric,
>
> After check 84/a4/c4, 88/a8/c8 of function 1 of K8 northbridge, the link
> speed/width is not init properly.

Is it function 0 ?

Ollie
> Regards
>
> YH
>
> -----邮件原件-----
> 发件人: [e-mail] [mailto:[e-mail]] 
> 发送时间: 2004年3月23日 23:42
> 收件人: YhLu
> 抄送: Stefan Reinauer; ron minnich; [e-mail]
> 主题: Re: S4882 support
>
> YhLu  writes:
> > Eric,
> >
> > The ht reset in auto.c seems not thoroughly.
> >
> > And it will stuck at hard_reset in amdk8_scan_chains. ( at this time can
>
> not
> > access 8111 yet).
>
> Correct.  The early hypertransport initialization code is not
> complete in the 4P case.  That incompleteness triggers the later
> fixups.
>
> I could not test that case, and there were at least some small
> difficulties in that it was not obvious how the 4P case should
> be extended to work properly.
>
> Why you are getting stuck in amdk8_scan_chains I do not currently
> understand yet.  I suspect Stefan's suggestion of a delayed reset
> is the appropriate course to implement.
>
> Eric
> _______________________________________________
> Linuxbios mailing list
> [e-mail]
> http://www.clustermatic.org/mailman/listinfo/linuxbios

S4882 support

[Stefan Reinauer]

* YhLu  [040324 18:43]:
> After check 84/a4/c4, 88/a8/c8 of function 1 of K8 northbridge, the link
> speed/width is not init properly.

Between which devices?

It seems that the current HT code (before my dynamic changes) is doing a
wrong assumption when optimizing the connections, that is only right on
the HDAMA. I fixed this a while ago, but it seems, this drowned in the
last big merge of 1.1.6. 
The assumption is that CPU0 and CPU1 are interconnected using the same
link on both CPUs:
optimize_connection(NODE_HT(0), 0x80 + CONNECTION_0_1, NODE_HT(7), 0x80 + CONNECTION_0_1);
The second link specified in the function call has to be from CPU 1 to
0, not from 0 to 1 as the first link. This happens to not matter on
HDAMA, but basically on all other machines. My dynamic connections diff
should fix this (even though it is not optimizing link speed between the
other CPUs yet).

Stefan