[PATCH] IP28: fix MC GIOPAR setting

[PATCH] IP28: fix MC GIOPAR setting

[peter fuerst]

We must not omit the MASTERGFX setting, unless we want Impact DMA to hang ;-)


Signed-off-by: peter fuerst <post@pfrst.de>


--- a/arch/mips/sgi-ip22/ip22-mc.c	Tue Jan 29 10:14:58 2008
+++ b/arch/mips/sgi-ip22/ip22-mc.c	Fri Mar 21 13:59:42 2008
@@ -180,7 +180,9 @@
 	/* First the basic invariants across all GIO64 implementations. */
 	tmp = SGIMC_GIOPAR_HPC64;	/* All 1st HPC's interface at 64bits */
 	tmp |= SGIMC_GIOPAR_ONEBUS;	/* Only one physical GIO bus exists */
-
+#ifdef CONFIG_SGI_IP28
+	tmp |= SGIMC_GIOPAR_MASTERGFX;	/* GFX can act as a bus master */
+#endif
 	if (ip22_is_fullhouse()) {
 		/* Fullhouse specific settings. */
 		if (SGIOC_SYSID_BOARDREV(sgioc->sysid) < 2) {

Re: [PATCH] IP28: fix MC GIOPAR setting

[Thomas Bogendoerfer]

On Fri, Mar 21, 2008 at 04:03:16PM +0100, peter fuerst wrote:
> We must not omit the MASTERGFX setting, unless we want Impact DMA to hang ;-)

is there a reason to restrict this to IP28 only ?

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessary a
good idea.                                                [ RFC1925, 2.3 ]

Re: [PATCH] IP28: fix MC GIOPAR setting

[peter fuerst]

On Fri, 21 Mar 2008, Thomas Bogendoerfer wrote:

> Date: Fri, 21 Mar 2008 20:47:37 +0100
> From: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
> To: peter fuerst <post@pfrst.de>
> Cc: linux-mips@linux-mips.org, ralf@linux-mips.org
> Subject: Re: [PATCH] IP28: fix MC GIOPAR setting
>
> On Fri, Mar 21, 2008 at 04:03:16PM +0100, peter fuerst wrote:
> > We must not omit the MASTERGFX setting, unless we want Impact DMA to hang ;-)
>
> is there a reason to restrict this to IP28 only ?

Would indeed be most surprising, if this isn't appropriate for any Indigo2-
Impact, but don't know for sure.  And can't check, whether it at least doesn't
hurt Non-Impact Indigo2.  Of course, being able to avoid '#ifdef' at all would
be the prettiest alternative.

>
> Thomas.
>
> --
> Crap can work. Given enough thrust pigs will fly, but it's not necessary a
> good idea.                                                [ RFC1925, 2.3 ]
>
>

Re: [PATCH] IP28: fix MC GIOPAR setting

[Thomas Bogendoerfer]

On Fri, Mar 21, 2008 at 10:07:38PM +0100, peter fuerst wrote:
> > is there a reason to restrict this to IP28 only ?
> 
> Would indeed be most surprising, if this isn't appropriate for any Indigo2-
> Impact, but don't know for sure.  And can't check, whether it at least doesn't
> hurt Non-Impact Indigo2.  Of course, being able to avoid '#ifdef' at all would
> be the prettiest alternative.

I'll check my IP22 machines, if they are ok with that change. Another
solution could be to have gio_set_master() similair to pci_set_master().
That way we only enable master, if it is requested by a driver.

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessary a
good idea.                                                [ RFC1925, 2.3 ]

Re: [PATCH] IP28: fix MC GIOPAR setting

[Ralf Baechle]

On Fri, Mar 21, 2008 at 10:32:33PM +0100, Thomas Bogendoerfer wrote:

> > Would indeed be most surprising, if this isn't appropriate for any Indigo2-
> > Impact, but don't know for sure.  And can't check, whether it at least doesn't
> > hurt Non-Impact Indigo2.  Of course, being able to avoid '#ifdef' at all would
> > be the prettiest alternative.
> 
> I'll check my IP22 machines, if they are ok with that change. Another
> solution could be to have gio_set_master() similair to pci_set_master().
> That way we only enable master, if it is requested by a driver.

That sounds like a clean solution.  Anyway where are we standing with this?
I assume it's release critical for IP28?

  Ralf

Re: [PATCH] IP28: fix MC GIOPAR setting

[Markus Gothe]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Thanks to O'Reilly in Germany, I've got an I^2, non-impact... I'll set
up an test enviroment for mips-linux there,
when I've got the time.

//Markus

Ralf Baechle wrote:
> On Fri, Mar 21, 2008 at 10:32:33PM +0100, Thomas Bogendoerfer
> wrote:
>
>>> Would indeed be most surprising, if this isn't appropriate for
>>> any Indigo2- Impact, but don't know for sure.  And can't check,
>>> whether it at least doesn't hurt Non-Impact Indigo2.  Of
>>> course, being able to avoid '#ifdef' at all would be the
>>> prettiest alternative.
>> I'll check my IP22 machines, if they are ok with that change.
>> Another solution could be to have gio_set_master() similair to
>> pci_set_master(). That way we only enable master, if it is
>> requested by a driver.
>
> That sounds like a clean solution.  Anyway where are we standing
> with this? I assume it's release critical for IP28?
>
> Ralf
>


- --
_______________________________________

Mr Markus Gothe
Software Engineer

Phone: +46 (0)13 21 81 20 (ext. 1046)
Fax: +46 (0)13 21 21 15
Mobile: +46 (0)70 348 44 35
Diskettgatan 11, SE-583 35 Linköping, Sweden
www.27m.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFH8me16I0XmJx2NrwRCG5JAJ9vCHZhAVXD9GGwtVC8RBgaTun+rQCePUnE
uDmjdB3ii0qRasWST4pqaJc=
=LCvE
-----END PGP SIGNATURE-----