* IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP
@ 2004-05-31 4:48 dann frazier
2004-05-31 14:32 ` Matthew Wilcox
` (3 more replies)
0 siblings, 4 replies; 5+ messages in thread
From: dann frazier @ 2004-05-31 4:48 UTC (permalink / raw)
To: linux-ia64
hey,
It was requested that I turn on DISCONTIGMEM in my builds for Debian, in order to support
Altix boxes. This option is not currently selectable w/ a 2.6.6 "Generic" kernel, but
it is with all other flavors.
I dug around to see if this is intentional, and found evidence[1] that it is not.
From what I can grok from Documentation/kbuild/kconfig-language.txt, select will force
these options off - maybe "depends on" is what was intended?
I tried creating a patch w/ this change, but found that adding thse deps creates a
recursive dependency tree that breaks Kconfig, since DISCONTIGMEM depends on
(IA64_DIG || IA64_SGI_SN2 || IA64_GENERIC).
What is the desired behavior? Should generic kernels require these options?
Should we force these options on for platforms that support them, and therefore
make them non-user configurable? Should we just turn them on by default, and
allow users to disable them?
[1]
|Age Author Annotate Comments
|8 months jbarnes@com[davidm] 1.48 ia64: force on appropriate generic options
|
|
|--- 1.47/arch/ia64/Kconfig Sun May 30 21:34:28 2004
|+++ 1.48/arch/ia64/Kconfig Sun May 30 21:34:28 2004
|@@ -57,6 +57,10 @@
|
| config IA64_GENERIC
| bool "generic"
|+ select NUMA
|+ select ACPI_NUMA
|+ select VIRTUAL_MEM_MAP
|+ select DISCONTIGMEM
| ---help---
| This selects the system type of your hardware. A "generic" kernel
| will run on any supported IA-64 system. However, if you configure
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP
2004-05-31 4:48 IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP dann frazier
@ 2004-05-31 14:32 ` Matthew Wilcox
2004-05-31 15:15 ` dann frazier
` (2 subsequent siblings)
3 siblings, 0 replies; 5+ messages in thread
From: Matthew Wilcox @ 2004-05-31 14:32 UTC (permalink / raw)
To: linux-ia64
On Sun, May 30, 2004 at 10:48:43PM -0600, dann frazier wrote:
> hey,
> It was requested that I turn on DISCONTIGMEM in my builds for Debian, in order to support
> Altix boxes. This option is not currently selectable w/ a 2.6.6 "Generic" kernel, but
> it is with all other flavors.
>
> I dug around to see if this is intentional, and found evidence[1] that it is not.
> >From what I can grok from Documentation/kbuild/kconfig-language.txt, select will force
> these options off - maybe "depends on" is what was intended?
No, select forces these options on. It's like Enhances: versus Suggests:
(if Enhances actually worked, which it doesn't). The reason you can't
turn DISCONTIGMEM on with a generic kernel is because it's forced on. I
suspect the bug submitter was smoking something good.
> | config IA64_GENERIC
> | bool "generic"
> |+ select NUMA
> |+ select ACPI_NUMA
> |+ select VIRTUAL_MEM_MAP
> |+ select DISCONTIGMEM
--
"Next the statesmen will invent cheap lies, putting the blame upon
the nation that is attacked, and every man will be glad of those
conscience-soothing falsities, and will diligently study them, and refuse
to examine any refutations of them; and thus he will by and by convince
himself that the war is just, and will thank God for the better sleep
he enjoys after this process of grotesque self-deception." -- Mark Twain
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP
2004-05-31 4:48 IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP dann frazier
2004-05-31 14:32 ` Matthew Wilcox
@ 2004-05-31 15:15 ` dann frazier
2004-05-31 15:42 ` Andreas Schwab
2004-06-01 0:22 ` dann frazier
3 siblings, 0 replies; 5+ messages in thread
From: dann frazier @ 2004-05-31 15:15 UTC (permalink / raw)
To: linux-ia64
On Mon, May 31, 2004 at 03:32:04PM +0100, Matthew Wilcox wrote:
> No, select forces these options on. It's like Enhances: versus Suggests:
> (if Enhances actually worked, which it doesn't). The reason you can't
> turn DISCONTIGMEM on with a generic kernel is because it's forced on. I
> suspect the bug submitter was smoking something good.
I suspected that initially, but this is what confused me:
$ make oldconfig > /dev/null && grep IA64_GENERIC .config && grep CONFIG_NUMA .config
CONFIG_IA64_GENERIC=y
# CONFIG_NUMA is not set
However, I can set it by editing .config by hand, and oldconfig doesn't drop
it, so I'm good to go.
--
---------------------------
dann frazier
Hewlett-Packard
Linux and Open Source Lab
dannf@hp.com
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP
2004-05-31 4:48 IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP dann frazier
2004-05-31 14:32 ` Matthew Wilcox
2004-05-31 15:15 ` dann frazier
@ 2004-05-31 15:42 ` Andreas Schwab
2004-06-01 0:22 ` dann frazier
3 siblings, 0 replies; 5+ messages in thread
From: Andreas Schwab @ 2004-05-31 15:42 UTC (permalink / raw)
To: linux-ia64
dann frazier <dannf@hp.com> writes:
> I suspected that initially, but this is what confused me:
> $ make oldconfig > /dev/null && grep IA64_GENERIC .config && grep CONFIG_NUMA .config
> CONFIG_IA64_GENERIC=y
> # CONFIG_NUMA is not set
Can't reproduce here. Whenever I select IA64_GENERIC I also get NUMA=y,
even if I start with a .config that has NUMA=n.
Andreas.
--
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Maxfeldstraße 5, 90409 Nürnberg, Germany
Key fingerprint = 58CA 54C7 6D53 942B 1756 01D3 44D5 214B 8276 4ED5
"And now for something completely different."
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP
2004-05-31 4:48 IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP dann frazier
` (2 preceding siblings ...)
2004-05-31 15:42 ` Andreas Schwab
@ 2004-06-01 0:22 ` dann frazier
3 siblings, 0 replies; 5+ messages in thread
From: dann frazier @ 2004-06-01 0:22 UTC (permalink / raw)
To: linux-ia64
On Mon, May 31, 2004 at 05:42:38PM +0200, Andreas Schwab wrote:
> dann frazier <dannf@hp.com> writes:
>
> > I suspected that initially, but this is what confused me:
> > $ make oldconfig > /dev/null && grep IA64_GENERIC .config && grep CONFIG_NUMA .config
> > CONFIG_IA64_GENERIC=y
> > # CONFIG_NUMA is not set
>
> Can't reproduce here. Whenever I select IA64_GENERIC I also get NUMA=y,
> even if I start with a .config that has NUMA=n.
Seems the machine I was doing my builds on has crashed, and I'm out of the country
at the moment - I can send my config when I regain access.
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2004-06-01 0:22 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-05-31 4:48 IA64_GENERIC vs. NUMA/DISCONTIG/VIRTUAL_MEM_MAP dann frazier
2004-05-31 14:32 ` Matthew Wilcox
2004-05-31 15:15 ` dann frazier
2004-05-31 15:42 ` Andreas Schwab
2004-06-01 0:22 ` dann frazier
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox