[PATCH] Document x86-64 iommu kernel parameters

[PATCH] Document x86-64 iommu kernel parameters

[Jeff Garzik]

Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
---
After having to go figure out what some of these means, I figured I
would save others the trouble.

Some of these are "best guess" based on a quick scan of the code, so it
certainly needs a sanity review before going upstream.

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 4d175c7..8afea9b 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -763,6 +763,30 @@ and is between 256 and 4096 characters. It is defined in the file
 
 	inttest=	[IA64]
 
+	iommu=option[,option..]	[X86-64]
+		off		Disable IOMMU.
+		force		Unconditionally enable IOMMU.
+		noforce		Disable IOMMU and IOMMU merging, by default.
+		biomerge	Unconditionally enable IOMMU, IOMMU merging,
+				and set BIO IOMMU vmerge boundary to 4096.
+		panic		Panic on IOMMU overflow.
+		nopanic		Do not panic on IOMMU overflow.
+		merge		Unconditionally enable IOMMU, IOMMU merging.
+		nomerge		Disable IOMMU merging.
+		forcesac	Force single address cycle (SAC, 32-bit).
+		allowdac	Permit dual address cycle (DAC, 64-bit).
+		nodac		Forbid dual address cycle (DAC, 64-bit).
+		soft		Enable swiotlb.
+		calgary		Use Calgary IOMMU.
+
+			(GART-only options follow...)
+		<NNN>		Specify size of remapping area.
+		fullflush	Disable optimizing flushing strategy.
+		nofullflush	Enable optimizing flushing strategy.
+		noagp		Use entire aperture, AGP isn't using it.
+		noaperture	Disable aperture fixups / hole init.
+		memaper=<N>	malloc an aperture of order N.
+
 	io7=		[HW] IO7 for Marvel based alpha systems
 			See comment before marvel_specify_io7 in
 			arch/alpha/kernel/core_marvel.c.

Re: [PATCH] Document x86-64 iommu kernel parameters

[Randy Dunlap]

On Tue, 2 Oct 2007 21:34:13 -0400 Jeff Garzik wrote:

> 
> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
> ---
> After having to go figure out what some of these means, I figured I
> would save others the trouble.
> 
> Some of these are "best guess" based on a quick scan of the code, so it
> certainly needs a sanity review before going upstream.

"iommu" is listed in Documentation/x86_64/boot-options.txt
along with more x86_64-specific boot options.
A few other arches do something similar...


> diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
> index 4d175c7..8afea9b 100644
> --- a/Documentation/kernel-parameters.txt
> +++ b/Documentation/kernel-parameters.txt
> @@ -763,6 +763,30 @@ and is between 256 and 4096 characters. It is defined in the file
>  
>  	inttest=	[IA64]
>  
> +	iommu=option[,option..]	[X86-64]
> +		off		Disable IOMMU.
> +		force		Unconditionally enable IOMMU.
> +		noforce		Disable IOMMU and IOMMU merging, by default.
> +		biomerge	Unconditionally enable IOMMU, IOMMU merging,
> +				and set BIO IOMMU vmerge boundary to 4096.
> +		panic		Panic on IOMMU overflow.
> +		nopanic		Do not panic on IOMMU overflow.
> +		merge		Unconditionally enable IOMMU, IOMMU merging.
> +		nomerge		Disable IOMMU merging.
> +		forcesac	Force single address cycle (SAC, 32-bit).
> +		allowdac	Permit dual address cycle (DAC, 64-bit).
> +		nodac		Forbid dual address cycle (DAC, 64-bit).
> +		soft		Enable swiotlb.
> +		calgary		Use Calgary IOMMU.
> +
> +			(GART-only options follow...)
> +		<NNN>		Specify size of remapping area.
> +		fullflush	Disable optimizing flushing strategy.
> +		nofullflush	Enable optimizing flushing strategy.
> +		noagp		Use entire aperture, AGP isn't using it.
> +		noaperture	Disable aperture fixups / hole init.
> +		memaper=<N>	malloc an aperture of order N.
> +
>  	io7=		[HW] IO7 for Marvel based alpha systems
>  			See comment before marvel_specify_io7 in
>  			arch/alpha/kernel/core_marvel.c.
> -


---
~Randy

Re: [PATCH] Document x86-64 iommu kernel parameters

[Jeff Garzik]

Randy Dunlap wrote:
> On Tue, 2 Oct 2007 21:34:13 -0400 Jeff Garzik wrote:
> 
>> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
>> ---
>> After having to go figure out what some of these means, I figured I
>> would save others the trouble.
>>
>> Some of these are "best guess" based on a quick scan of the code, so it
>> certainly needs a sanity review before going upstream.
> 
> "iommu" is listed in Documentation/x86_64/boot-options.txt
> along with more x86_64-specific boot options.
> A few other arches do something similar...

Ah!  Well, seeing as how we already have a provision for arch-specific 
options in kernel-parameters.txt, and some less-obscure arch-specific 
options can be found there, I think an argument can be made for my patch :)

Nonethless, if the maintainer disagrees, they can drop this patch I suppose.

	Jeff

Re: [PATCH] Document x86-64 iommu kernel parameters

[Randy Dunlap]

Jeff Garzik wrote:
> Randy Dunlap wrote:
>> On Tue, 2 Oct 2007 21:34:13 -0400 Jeff Garzik wrote:
>>
>>> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
>>> ---
>>> After having to go figure out what some of these means, I figured I
>>> would save others the trouble.
>>>
>>> Some of these are "best guess" based on a quick scan of the code, so it
>>> certainly needs a sanity review before going upstream.
>>
>> "iommu" is listed in Documentation/x86_64/boot-options.txt
>> along with more x86_64-specific boot options.
>> A few other arches do something similar...
> 
> Ah!  Well, seeing as how we already have a provision for arch-specific 
> options in kernel-parameters.txt, and some less-obscure arch-specific 
> options can be found there, I think an argument can be made for my patch :)
> 
> Nonethless, if the maintainer disagrees, they can drop this patch I 
> suppose.

or maybe during the x86 merge, we can merge the docs also...

-- 
~Randy

Re: [PATCH] Document x86-64 iommu kernel parameters

[Randy Dunlap]

On Tue, 02 Oct 2007 22:30:31 -0400 Jeff Garzik wrote:

> Randy Dunlap wrote:
> > On Tue, 2 Oct 2007 21:34:13 -0400 Jeff Garzik wrote:
> > 
> >> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
> >> ---
> >> After having to go figure out what some of these means, I figured I
> >> would save others the trouble.
> >>
> >> Some of these are "best guess" based on a quick scan of the code, so it
> >> certainly needs a sanity review before going upstream.
> > 
> > "iommu" is listed in Documentation/x86_64/boot-options.txt
> > along with more x86_64-specific boot options.
> > A few other arches do something similar...
> 
> Ah!  Well, seeing as how we already have a provision for arch-specific 
> options in kernel-parameters.txt, and some less-obscure arch-specific 
> options can be found there, I think an argument can be made for my patch :)
> 
> Nonethless, if the maintainer disagrees, they can drop this patch I suppose.

[sorry if there be duplicates; I thought I sent this but can't find it
anywhere]


Maybe we can/should merge the doc files along with the x86 arch merge.

---
~Randy

Re: [PATCH] Document x86-64 iommu kernel parameters

[H. Peter Anvin]

Randy Dunlap wrote:
> 
> Maybe we can/should merge the doc files along with the x86 arch merge.
> 

Well, the x86 merge is pretty much mechanical.  It should be followed up 
with a lot of manual merging.

	-hpa

Re: [PATCH] Document x86-64 iommu kernel parameters

[Chuck Ebbert]

On 10/03/2007 12:10 AM, H. Peter Anvin wrote:
> Randy Dunlap wrote:
>>
>> Maybe we can/should merge the doc files along with the x86 arch merge.
>>
> 
> Well, the x86 merge is pretty much mechanical.  It should be followed up
> with a lot of manual merging.
> 

It would be really nice if identical kernel parameters used the same names,
so people didn't waste their time trying e.g. "nolapic_timer" on x86_64
because that option fixed their problems on i386 (it's "noapictimer" in
x86_64.)

Re: [PATCH] Document x86-64 iommu kernel parameters

[Andi Kleen]

On Wednesday 03 October 2007 18:58, Chuck Ebbert wrote:
> On 10/03/2007 12:10 AM, H. Peter Anvin wrote:
> > Randy Dunlap wrote:
> >> Maybe we can/should merge the doc files along with the x86 arch merge.
> >
> > Well, the x86 merge is pretty much mechanical.  It should be followed up
> > with a lot of manual merging.
>
> It would be really nice if identical kernel parameters used the same names,
> so people didn't waste their time trying e.g. "nolapic_timer" on x86_64
> because that option fixed their problems on i386 (it's "noapictimer" in
> x86_64.)

Well the x86-64 option was there first; you have to ask whoever
added the i386 option why they used a different name

But I normally added aliases as people reported them. Nobody did in 
this case.

One reason is probably that noapictimer is not very useful on x86-64 --
the nolapic_timer is really more a workaround that only started making
sense on HRT because it made "apicmaintimer" default. 
On current x86-64 which doesn't do that there is very little reason to use it
and you'll usually just break things.

So I guess the real reason is that they're not actually identical
when you look closely.

-Andi

Re: [PATCH] Document x86-64 iommu kernel parameters

[Chuck Ebbert]

On 10/04/2007 05:01 AM, Andi Kleen wrote:
> 
> Well the x86-64 option was there first; you have to ask whoever
> added the i386 option why they used a different name
> 
> But I normally added aliases as people reported them. Nobody did in 
> this case.
> 
> One reason is probably that noapictimer is not very useful on x86-64 --
> the nolapic_timer is really more a workaround that only started making
> sense on HRT because it made "apicmaintimer" default. 
> On current x86-64 which doesn't do that there is very little reason to use it
> and you'll usually just break things.
> 

I have HRT in my x86_64 kernel. I guess that patchset should add the alias.