Re: [Gelato-technical] What's still broken?

Re: [Gelato-technical] What's still broken?

[Peter Chubb]

>>>>> "Ian" = Ian Wienand <ianw@gelato.unsw.edu.au> writes:

Ian> On Thu, Oct 09, 2003 at 03:43:22PM +1000, Peter Chubb wrote:
>> >>>>> "Stephane" = Stephane Eranian <eranian@hpl.hp.com> writes:
Stephane> What about the qla1280 driver? Last time I tried it would
Stephane> crash during boot.  I don't know if it has been fixed.

>> It has been fixed.

Ian> Well it still breaks a generic build as discussed in

Ian> http://www.gelato.unsw.edu.au/linux-ia64/0309/6661.html

That's not so much a qla1280 problem, as a generic build problem ---
it's half baked; it includes ALTIX-specific code in the qla1280
driver, but doesn't build the SN subtree.

A temporary fix would be to change

#if defined(CONFIG_IA64_GENERIC) || defined(CONFIG_IA64_SGI_SN2)

to 

#if defined(CONFIG_IA64_SGI_SN2)

in the QLA1280 driver, at least until the SN2 bits are included in the
generic kernel.

Jesse, is the SN2 stuff you're working on so hard going to be included
in generic builds?  Something like this (which doesn't yet work):

--- /tmp/geta1932	2003-10-09 20:13:51.000000000 +1000
+++ linux-2.5-test7/arch/ia64/Makefile		2003-10-09 20:11:04.000000000 +1000
@@ -57,7 +57,7 @@
 core-y				+= arch/ia64/kernel/ arch/ia64/mm/
 core-$(CONFIG_IA32_SUPPORT)	+= arch/ia64/ia32/
 core-$(CONFIG_IA64_DIG)	+= arch/ia64/dig/
-core-$(CONFIG_IA64_GENERIC)	+= arch/ia64/dig/
+core-$(CONFIG_IA64_GENERIC)	+= arch/ia64/dig/ arch/ia64/sn/
 core-$(CONFIG_IA64_HP_ZX1)	+= arch/ia64/dig/
 core-$(CONFIG_IA64_SGI_SN2)	+= arch/ia64/sn/

--
Dr Peter Chubb  http://www.gelato.unsw.edu.au  peterc AT gelato.unsw.edu.au
You are lost in a maze of BitKeeper repositories,   all slightly different.

Re: [Gelato-technical] What's still broken?

[Jesse Barnes]

On Thu, Oct 09, 2003 at 08:25:12PM +1000, Peter Chubb wrote:
> Jesse, is the SN2 stuff you're working on so hard going to be included

:)

> in generic builds?  Something like this (which doesn't yet work):

Yep, I've tested it thorougly here, and it works for me.  I don't think
adding the below patch will work until my virtual memmap + discontig
patch (the last one I posted) goes in though.

> --- /tmp/geta1932	2003-10-09 20:13:51.000000000 +1000
> +++ linux-2.5-test7/arch/ia64/Makefile		2003-10-09 20:11:04.000000000 +1000
> @@ -57,7 +57,7 @@
>  core-y				+= arch/ia64/kernel/ arch/ia64/mm/
>  core-$(CONFIG_IA32_SUPPORT)	+= arch/ia64/ia32/
>  core-$(CONFIG_IA64_DIG)	+= arch/ia64/dig/
> -core-$(CONFIG_IA64_GENERIC)	+= arch/ia64/dig/
> +core-$(CONFIG_IA64_GENERIC)	+= arch/ia64/dig/ arch/ia64/sn/
>  core-$(CONFIG_IA64_HP_ZX1)	+= arch/ia64/dig/
>  core-$(CONFIG_IA64_SGI_SN2)	+= arch/ia64/sn/

Thanks,
Jesse

Re: [Gelato-technical] What's still broken?

[David Mosberger]

I don't understand this entry:

 Peter> - AGP driver still doesn't build.

I build agpgart.ko, i460-gart.ko, and hp-agp.ko on a regular basis.
I'm not aware of any build issues.

	--david

Re: [Gelato-technical] What's still broken?

[Peter Chubb]

>>>>> "David" = David Mosberger <davidm@napali.hpl.hp.com> writes:

David> I don't understand this entry:
Peter> - AGP driver still doesn't build.

David> I build agpgart.ko, i460-gart.ko, and hp-agp.ko on a regular
David> basis.  I'm not aware of any build issues.

It seems to be fixed now.  At the time I tried it there were issues.
I've updated the Wiki.

Petr C