Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1430092 > unrolled thread

Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix Reza Arbab <arbab@linux.vnet.ibm.com> - 2016-06-23 21:40 +0200
    Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix Reza Arbab <arbab@linux.vnet.ibm.com> - 2016-06-23 22:00 +0200
    Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix Michael Ellerman <mpe@ellerman.id.au> - 2016-06-28 13:30 +0200
      Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com> - 2016-06-28 16:10 +0200
      Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix Reza Arbab <arbab@linux.vnet.ibm.com> - 2016-06-29 17:50 +0200

#1430092 — Re: [PATCH] powerpc/mm: update arch_{add,remove}_memory() for radix

On Thu, Jun 23, 2016 at 10:47:20PM +0530, Aneesh Kumar K.V wrote:
>Reza Arbab <arbab@linux.vnet.ibm.com> writes:
>> These functions are making direct calls to the hash table APIs,
>> leading to a BUG() on systems using radix.
>>
>> Switch them to the vmemmap_{create,remove}_mapping() wrappers, and
>> move to the __meminit section.
>
>They are really not the same. They can possibly end up using different
>base page size. Also vmemmap is available only with SPARSEMEM_VMEMMAP
>enabled. Does hotplug depend on sparsemem vmemmap ?

I'm not sure. Maybe it's best if I back up a step and explain what lead 
me to this patch. During hotplug, you get

...
	arch_add_memory
		create_section_mapping
			htab_bolt_mapping
				BUG_ON(!ppc_md.hpte_insert);

So it seemed to me that I needed a radix equivalent of 
create_section_mapping().

After some digging, I found hash__vmemmap_create_mapping() and 
radix__vmemmap_create_mapping() did what I needed. I did not notice the 
#ifdef SPARSEMEM_VMEMMAP around them.

Hotplug/remove are now working for me as far as I can tell, so I think
the functional change in itself was correct. Hotplug may not depend on 
sparsemem vmemmap, but we seem to need a radix create_section_mapping() 
regardless.

Could it be that the functions just need to be renamed 
hash__create_mapping()/radix__create_mapping() and moved out of #ifdef 
SPARSEMEM_VMEMMAP?

-- 
Reza Arbab

[toc] | [next] | [standalone]

#1430119

On Thu, Jun 23, 2016 at 02:37:39PM -0500, Reza Arbab wrote:
>Could it be that the functions just need to be renamed 
>hash__create_mapping()/radix__create_mapping() and moved out of #ifdef 
>SPARSEMEM_VMEMMAP?

Or vice-versa, maybe the callers should have been wrapped in the first 
place:

arch_add_memory() {
	...
	if (IS_ENABLED(CONFIG_SPARSEMEM_VMEMMAP))
		vmemmap_create_mapping()
	...
}

arch_remove_memory() {
	...
	if (IS_ENABLED(CONFIG_SPARSEMEM_VMEMMAP))
		vmemmap_remove_mapping()
	...
}

-- 
Reza Arbab

[toc] | [prev] | [next] | [standalone]

#1432877

On Thu, 2016-06-23 at 14:37 -0500, Reza Arbab wrote:
> On Thu, Jun 23, 2016 at 10:47:20PM +0530, Aneesh Kumar K.V wrote:
> > Reza Arbab <arbab@linux.vnet.ibm.com> writes:
> > > These functions are making direct calls to the hash table APIs,
> > > leading to a BUG() on systems using radix.
> > > 
> > > Switch them to the vmemmap_{create,remove}_mapping() wrappers, and
> > > move to the __meminit section.
> > 
> > They are really not the same. They can possibly end up using different
> > base page size. Also vmemmap is available only with SPARSEMEM_VMEMMAP
> > enabled. Does hotplug depend on sparsemem vmemmap ?
> 
> I'm not sure. Maybe it's best if I back up a step and explain what lead 
> me to this patch. During hotplug, you get
> 
> ...
> 	arch_add_memory
> 		create_section_mapping
> 			htab_bolt_mapping
> 				BUG_ON(!ppc_md.hpte_insert);
> 
> So it seemed to me that I needed a radix equivalent of 
> create_section_mapping().
> 
> After some digging, I found hash__vmemmap_create_mapping() and 
> radix__vmemmap_create_mapping() did what I needed. I did not notice the 
> #ifdef SPARSEMEM_VMEMMAP around them.

I think that's more by luck than design. The vmemmap routines use
mmu_vmemmap_psize which is probably but not definitely the same as
mmu_linear_psize.

> Could it be that the functions just need to be renamed 
> hash__create_mapping()/radix__create_mapping() and moved out of #ifdef 
> SPARSEMEM_VMEMMAP?

No, you need to use mmu_linear_psize for the hotplug case.

But you can probably factor out a common routine that both cases use, and hide
the hash vs radix check in that.

And probably send me a patch to make MEMORY_HOTPLUG depend on !RADIX for v4.7?

cheers

[toc] | [prev] | [next] | [standalone]

#1433013

Michael Ellerman <mpe@ellerman.id.au> writes:

> On Thu, 2016-06-23 at 14:37 -0500, Reza Arbab wrote:
>> On Thu, Jun 23, 2016 at 10:47:20PM +0530, Aneesh Kumar K.V wrote:
>> > Reza Arbab <arbab@linux.vnet.ibm.com> writes:
>> > > These functions are making direct calls to the hash table APIs,
>> > > leading to a BUG() on systems using radix.
>> > > 
>> > > Switch them to the vmemmap_{create,remove}_mapping() wrappers, and
>> > > move to the __meminit section.
>> > 
>> > They are really not the same. They can possibly end up using different
>> > base page size. Also vmemmap is available only with SPARSEMEM_VMEMMAP
>> > enabled. Does hotplug depend on sparsemem vmemmap ?
>> 
>> I'm not sure. Maybe it's best if I back up a step and explain what lead 
>> me to this patch. During hotplug, you get
>> 
>> ...
>> 	arch_add_memory
>> 		create_section_mapping
>> 			htab_bolt_mapping
>> 				BUG_ON(!ppc_md.hpte_insert);
>> 
>> So it seemed to me that I needed a radix equivalent of 
>> create_section_mapping().
>> 
>> After some digging, I found hash__vmemmap_create_mapping() and 
>> radix__vmemmap_create_mapping() did what I needed. I did not notice the 
>> #ifdef SPARSEMEM_VMEMMAP around them.
>
> I think that's more by luck than design. The vmemmap routines use
> mmu_vmemmap_psize which is probably but not definitely the same as
> mmu_linear_psize.
>
>> Could it be that the functions just need to be renamed 
>> hash__create_mapping()/radix__create_mapping() and moved out of #ifdef 
>> SPARSEMEM_VMEMMAP?
>
> No, you need to use mmu_linear_psize for the hotplug case.
>
> But you can probably factor out a common routine that both cases use, and hide
> the hash vs radix check in that.
>
> And probably send me a patch to make MEMORY_HOTPLUG depend on !RADIX for v4.7?

Few other stuff we need to still look from Radix point
1) machine check handling/memory errors
2) kexec

-aneesh

[toc] | [prev] | [next] | [standalone]

#1433864

On Tue, Jun 28, 2016 at 09:21:05PM +1000, Michael Ellerman wrote:
>No, you need to use mmu_linear_psize for the hotplug case.
>
>But you can probably factor out a common routine that both cases use, and hide
>the hash vs radix check in that.

Okay, I'm trying to refactor {create,remove}_section_mapping() into 
hash__ and radix__ variants. This lead to a couple of questions.

Pseudocode for radix__create_section_mapping(start, end):

	page_size = 1 << mmu_psize_defs[mmu_linear_psize].shift;
	start = _ALIGN_DOWN(start, page_size);

	for (; start < end; start += page_size) {
		radix__map_kernel_page(start, __pa(start),
				       PAGE_KERNEL, page_size);
	}

Should the above use PAGE_KERNEL, like the the hash table bolt, or 
(_PAGE_PRESENT | _PAGE_ACCESSED | _PAGE_KERNEL_RW), like in the radix 
vmemmap creation?

The other question is what radix__remove_section_mapping() should do.
I don't know offhand what the opposite of map_kernel_page() is. As 
Aneesh mentioned, radix vmemmap removal is currently stubbed as a FIXME 
so I couldn't use that as a reference.

-- 
Reza Arbab

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web