Re: [PATCH v3 1/2] mm/highmem: make kmap cache coloring aware

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Fri, 25 Jul 2014 23:43:46 +0400 Max Filippov <jcmvbkbc@xxxxxxxxx> wrote:

> VIPT cache with way size larger than MMU page size may suffer from
> aliasing problem: a single physical address accessed via different
> virtual addresses may end up in multiple locations in the cache.
> Virtual mappings of a physical address that always get cached in
> different cache locations are said to have different colors.
> L1 caching hardware usually doesn't handle this situation leaving it
> up to software. Software must avoid this situation as it leads to
> data corruption.
> 
> One way to handle this is to flush and invalidate data cache every time
> page mapping changes color. The other way is to always map physical page
> at a virtual address with the same color. Low memory pages already have
> this property. Giving architecture a way to control color of high memory
> page mapping allows reusing of existing low memory cache alias handling
> code.
> 
> Provide hooks that allow architectures with aliasing cache to align
> mapping address of high pages according to their color. Such architectures
> may enforce similar coloring of low- and high-memory page mappings and
> reuse existing cache management functions to support highmem.
> 
> This code is based on the implementation of similar feature for MIPS by
> Leonid Yegoshin <Leonid.Yegoshin@xxxxxxxxxx>.
> 

It's worth mentioning that xtensa needs this.

What is (still) missing from these changelogs is a clear description of
the end-user visible effects.  Does it fix some bug?  If so what?  Is
it a performace optimisation?  If so how much?  This info is the
top-line reason for the patchset and should be presented as such.

> --- a/mm/highmem.c
> +++ b/mm/highmem.c
> @@ -28,6 +28,9 @@
>  #include <linux/highmem.h>
>  #include <linux/kgdb.h>
>  #include <asm/tlbflush.h>
> +#ifdef CONFIG_HIGHMEM
> +#include <asm/highmem.h>
> +#endif

Should be unneeded - the linux/highmem.h inclusion already did this.

Apart from that it all looks OK to me.  I'm assuming this is 3.17-rc1
material, but I am unsure because of the missing end-user-impact info. 
If it's needed in earlier kernels then we can tag it for -stable
backporting but again, the -stable team (ie: Greg) will want so see the
justification for that backport.


[Index of Archives]     [Linux MIPS Home]     [LKML Archive]     [Linux ARM Kernel]     [Linux ARM]     [Linux]     [Git]     [Yosemite News]     [Linux SCSI]     [Linux Hams]

  Powered by Linux