Re: [PATCH][v2] powerpc/mm: using two zones for freescale 64 bit kernel

On Sep 20, 2012, at 8:36 AM, Kumar Gala wrote:

=20
On Sep 20, 2012, at 5:14 AM, Xie Shaohui-B21989 wrote:
=20

quoted

quoted

quoted

On Thu, 2012-08-30 at 15:49 -0500, Kumar Gala wrote:

quoted

On Aug 24, 2012, at 5:50 AM, Shaohui Xie wrote:
=20

quoted

PowerPC platform only supports ZONE_DMA zone for 64bit kernel, so
all the memory will be put into this zone. If the memory size is
greater than the device's DMA capability and device uses
dma_alloc_coherent to allocate memory, it will get an address
which is over the device's DMA addressing, the device will fail.
=20
So we split the memory to two zones: zone ZONE_DMA32 &
ZONE_NORMAL, since we already allocate PCICSRBAR/PEXCSRBAR right
below the 4G boundary (if the lowest PCI address is above 4G), so
we constrain the DMA zone ZONE_DMA32 to 2GB, also, we clear flag
__GFP_DMA &
__GFP_DMA32 and set __GFP_DMA32 only if the device's dma_mask <
total memory size. By doing this, devices which cannot DMA all =

the

quoted

quoted

quoted

quoted

quoted

memory will be limited to ZONE_DMA32, but devices which can DMA
all

the memory will not be affected by this limitation.

quoted

quoted

=20
Signed-off-by: Shaohui Xie <redacted>
Signed-off-by: Mingkai Hu <redacted>
Signed-off-by: Chen Yuanquan <redacted>
---
changes for v2:
1. use a config option for using two zones (ZONE_DMA32 &
ZONE_NORMAL) in freescale 64 bit kernel.
=20

=20
There must have been a misunderstanding. I think this should be a
runtime choice, possibly by the platform code. Any reason that =

can't be

quoted

quoted

done ?

quoted

=20

[S.H] Do you mean this:
=20
phys_addr_t platform_dma_size (maybe a default value should be used, =

then

quoted

quoted

platform code will change it)
=20
if (top_of_ram > platform_dma_size)
	max_zone_pfns[ZONE_DMA] =3D platform_dma_size >> PAGE_SHIFT; =

else

quoted

quoted

	max_zone_pfns[ZONE_DMA] =3D top_of_ram >> PAGE_SHIFT;
=20
max_zone_pfns[ZONE_NORMAL] =3D top_of_ram >> PAGE_SHIFT;
=20

quoted

Also how does Intel do it ?

[S.H] below are codes in Intel:
=20
403 void __init zone_sizes_init(void)
404 {
405         unsigned long max_zone_pfns[MAX_NR_ZONES];
406
407         memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
408
409 #ifdef CONFIG_ZONE_DMA
410         max_zone_pfns[ZONE_DMA]         =3D MAX_DMA_PFN;
411 #endif
412 #ifdef CONFIG_ZONE_DMA32
413         max_zone_pfns[ZONE_DMA32]       =3D MAX_DMA32_PFN;
414 #endif
415         max_zone_pfns[ZONE_NORMAL]      =3D max_low_pfn;
416 #ifdef CONFIG_HIGHMEM
417         max_zone_pfns[ZONE_HIGHMEM]     =3D max_pfn;
418 #endif
419
=20
For x86_64, there is no CONFIG_HIGHMEM, so there will be three =

zones:

quoted

quoted

ZONE_DMA/ZONE_DMA32/ZONE_NORMAL.
=20

[S.H] Hello, Ben,
=20
I have some questions, though I'm still expecting your comments.
PPC does not have ZONE_DMA32 by default, if we want to use it, we =

need to add "config ZONE_DMA32" in Kconfig first.

quoted

If setting multiple zones without ZONE_DMA, kmalloc in =

"include/linux/slab_def.h" will fail if it uses flag GFP_DMA.

quoted

For the runtime choice in 64-bit kernel, what exactly multiple zones =

should be used?

quoted

"ZONE_DMA & ZONE_NORMAL" or "ZONE_DMA & ZONE_DMA32 & ZONE_NORMAL"?
Then what the size should be set for them respectively?
=20
Please comment, Thanks!

=20
I think Ben is saying that Kconfig would enable ZONE_DMA32 for all =

PPC64, but make it runtime/per platform how we setup the zone's such =
that either ZONE_DMA32 is set to MAX_DMA32_PFN or it set to same value =
as ZONE_DMA.

=20
However that's just a guess.

Ben,

Can you help clarify your thoughts here.

thanks

- k=