[PATCH v2 9/9] arm64: Work around systems with mismatched cache line sizes

Catalin Marinas catalin.marinas at arm.com
Fri Aug 26 10:00:34 PDT 2016


On Fri, Aug 26, 2016 at 05:16:27PM +0100, Will Deacon wrote:
> On Fri, Aug 26, 2016 at 02:08:01PM +0100, Suzuki K Poulose wrote:
> > On 26/08/16 14:04, Suzuki K Poulose wrote:
> > >On 26/08/16 12:03, Ard Biesheuvel wrote:
> > >>IMO, this is a pattern that we should avoid: you are introducing one
> > >>instance now, which will make it hard to say no to the next one in the
> > >>future. Isn't there a better way to organize the arm64_ftr_reg array
> > >>that allows us to reference entries directly? Ideally, a way that gets
> > >>rid of the runtime sorting, since I don't think that is a good
> > >>replacement for developer discipline anyway (although I should have
> > >>spoken up when that was first introduced) Or am I missing something
> > >>here?

I'm not sure we can have some simple direct access. Suzuki did some
grouping originally but I wouldn't call that a direct access either,
more like iterating through several groups.

A possibility would be to generate global variables for each
arm64_ftr_reg with the ARM64_FTR_REG macro extended to also place a
pointer in a dedicated section to be used as array. Assembly code would
access the global variable directly. But I'm not really sure it's worth
it.

> > >I had some form of direct access to the feature register in one of
> > >the versions [0], but was dropped based on Catalin's suggestion at [1].
> > 
> > Forgot to add, [0] wouldn't solve this issue cleanly either. It would simply
> > speed up the read_system_reg(). So we do need a call to read_system_reg()
> > from assembly code, which makes it a little bit tricky.
> 
> It might be worth looking to see if we can pass the ctr as an extra
> parameter to the assembly routines that need it. Then you can access it
> easily from C code, and if you pass it as 0 that could result in the asm
> code reading it from the h/w register, removing the need for the _raw
> stuff you add.

How often to we need to access a sanitised sysreg from assembly? AFAICT,
CTR_EL0 is the first. If we only need it to infer the minimum cache line
size, we could as well store the latter in a global variable and access
it directly. If we feel brave, we could patch a "mov \reg, #x"
instruction in the ?cache_line_size macros (starting with 32 by default,
though to make it less cumbersome we'd have to improve the run-time
patching code a bit).

-- 
Catalin



More information about the linux-arm-kernel mailing list