Re: new irq.c

Re: new irq.c

[Geert Uytterhoeven]

On Thu, 14 Jan 1999, Cort Dougan wrote:
> This is the latest version of the new irq.c
> 
> As it is, chrp needs to be improved a bit (it still uses the old
> irq_to_openpic() style of things) but that's not a problem right now.
> 
> Please try it out, if you have problems let me know.  If things look good
> I'll go ahead and commit it to VGER.

Works fine on my CHRP box.

Greetings,

						Geert

--
Geert Uytterhoeven                     Geert.Uytterhoeven@cs.kuleuven.ac.be
Wavelets, Linux/{m68k~Amiga,PPC~CHRP}  http://www.cs.kuleuven.ac.be/~geert/
Department of Computer Science -- Katholieke Universiteit Leuven -- Belgium


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

MCLK=100 breaks PowerBook

[Benjamin Herrenschmidt]

The default mclk setting of 100MHz for the Rage LT-G breaks on my
PowerBook G3 Series (garbage on screen, accelerated X hangs). I added
mclk:83 to my command line and it works fine again (and, progress from
pre-4, I have no more snow in X when moving things around).

>atyfb: 3D RAGE LT PRO [0x4c47 rev 0x80] 4M SGRAM, 230 MHz PLL, 83 Mhz MCLK

the chip ID 0x4c47 should be set to 83 MHz by default, not 100.

Note that I made a typo the first time I tested and I used 88Mhz. it
works fine too.

MacOS ATI driver creates a strange property in the device tree. It's name
is "MCLK,XCLK" and the value is 00001a38 00001a38. Geert, those values
have any meaning to you ? If you tell me how to do, I could calc the MCLK
used by MacOS by peeking appropriate registers from MacsBug and we could
use this value for the kernel with this chipID.


-- 
           E-Mail: <mailto:bh40@calva.net>
BenH.      Web   : <http://calvaweb.calvacom.fr/bh40/>




[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: MCLK=100 breaks PowerBook

[Geert Uytterhoeven]

On Mon, 18 Jan 1999, Benjamin Herrenschmidt wrote:
> The default mclk setting of 100MHz for the Rage LT-G breaks on my
> PowerBook G3 Series (garbage on screen, accelerated X hangs). I added
> mclk:83 to my command line and it works fine again (and, progress from
> pre-4, I have no more snow in X when moving things around).
> 
> >atyfb: 3D RAGE LT PRO [0x4c47 rev 0x80] 4M SGRAM, 230 MHz PLL, 83 Mhz MCLK
> 
> the chip ID 0x4c47 should be set to 83 MHz by default, not 100.
> 
> Note that I made a typo the first time I tested and I used 88Mhz. it
> works fine too.

I'm already aware that the parameters for the LT versions are not really
correct. On Cort's PowerBook, it's even worse: OF (or MacOS) seem to program it
to a mclk of only 40 MHz, with a PLL limit of 120 MHz. Unfortunately even with
these parameters it doesn't work :-(

> MacOS ATI driver creates a strange property in the device tree. It's name
> is "MCLK,XCLK" and the value is 00001a38 00001a38. Geert, those values

If you convert these to decimal, you get `6712', i.e. a mclk of 67 MHz.
Cort: what are the `MCLK,XCLK' properties for your PowerBook?

(my RAGE II+ doesn't have `MCLK,XCLK')

> have any meaning to you ? If you tell me how to do, I could calc the MCLK
> used by MacOS by peeking appropriate registers from MacsBug and we could
> use this value for the kernel with this chipID.

Just enable #define DEBUG in atyfb and feed the PLL values reported during boot
up to

    http://www.cs.kuleuven.ac.be/~geert/bin/atydebug.c.gz

Greetings,

						Geert

--
Geert Uytterhoeven                     Geert.Uytterhoeven@cs.kuleuven.ac.be
Wavelets, Linux/{m68k~Amiga,PPC~CHRP}  http://www.cs.kuleuven.ac.be/~geert/
Department of Computer Science -- Katholieke Universiteit Leuven -- Belgium


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: MCLK=100 breaks PowerBook

[Cort Dougan]

0x0000 0x1fe2 0x0000 0x3fc3

}If you convert these to decimal, you get `6712', i.e. a mclk of 67 MHz.
}Cort: what are the `MCLK,XCLK' properties for your PowerBook?


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: MCLK=100 breaks PowerBook

[Geert Uytterhoeven]

On Mon, 18 Jan 1999, Cort Dougan wrote:
> 0x0000 0x1fe2 0x0000 0x3fc3
> 
> }If you convert these to decimal, you get `6712', i.e. a mclk of 67 MHz.
> }Cort: what are the `MCLK,XCLK' properties for your PowerBook?

Hmm, this would indicate mclk (GUI engine clock) runs at 8 MHz, and xclk
(memory clock) runs at 16 MHz... very slow machine :-)

BTW, on most ATI chips mclk and xclk run at the same speed.

Greetings,

						Geert

--
Geert Uytterhoeven                     Geert.Uytterhoeven@cs.kuleuven.ac.be
Wavelets, Linux/{m68k~Amiga,PPC~CHRP}  http://www.cs.kuleuven.ac.be/~geert/
Department of Computer Science -- Katholieke Universiteit Leuven -- Belgium


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: MCLK=100 breaks PowerBook

[Paul Mackerras]

Geert Uytterhoeven <Geert.Uytterhoeven@cs.kuleuven.ac.be> wrote:

> On Mon, 18 Jan 1999, Cort Dougan wrote:
> > 0x0000 0x1fe2 0x0000 0x3fc3
> 
> Hmm, this would indicate mclk (GUI engine clock) runs at 8 MHz, and xclk
> (memory clock) runs at 16 MHz... very slow machine :-)

0x1fe2 == 8162, 0x3fc3 == 16323, so mclk = 81.62 MHz surely, not
8MHz??

Paul.

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: MCLK=100 breaks PowerBook

[Geert Uytterhoeven]

On Tue, 19 Jan 1999, Paul Mackerras wrote:
> Geert Uytterhoeven <Geert.Uytterhoeven@cs.kuleuven.ac.be> wrote:
> > On Mon, 18 Jan 1999, Cort Dougan wrote:
> > > 0x0000 0x1fe2 0x0000 0x3fc3
> > 
> > Hmm, this would indicate mclk (GUI engine clock) runs at 8 MHz, and xclk
> > (memory clock) runs at 16 MHz... very slow machine :-)
> 
> 0x1fe2 == 8162, 0x3fc3 == 16323, so mclk = 81.62 MHz surely, not
> 8MHz??

Oops, you're right.

Greetings,

						Geert

--
Geert Uytterhoeven                     Geert.Uytterhoeven@cs.kuleuven.ac.be
Wavelets, Linux/{m68k~Amiga,PPC~CHRP}  http://www.cs.kuleuven.ac.be/~geert/
Department of Computer Science -- Katholieke Universiteit Leuven -- Belgium


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Choosing an Ethernet controller...

[Benjamin Herrenschmidt]

I know this is a little bit out-of-topic, but I'm sure someone on this
list can answer and I didn't really know how to actually ask this...

For an embedded developement, I have to choose among 3 different PCI
ethernet controllers. A DEC/Intel 21143 (Tulip), an AMD79C971 (PCNet32)
or a SMC83C171 (Epic100).

The processor will not be a PPC but a StrongARM. (we proposed a PPC, but
the client wanted a StrongARM).
All three seems to be supported by Linux, I will use a 2.0.x kernel at
first, on a StrongARM, but we plan to move to 2.2 once it's stable enough.

Are there any known problems with one of those chips ? Is someone aware
of any specific problem with one of those beasts or it's linux driver ? 

Which one would you choose ?

Thanks !


-- 
           E-Mail: <mailto:bh40@calva.net>
BenH.      Web   : <http://calvaweb.calvacom.fr/bh40/>





[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]