[RFC PATCH] arm64: fpsimd: improve stacking logic in non-interruptible context

On Tue, Dec 06, 2016 at 04:48:54PM +0000, Ard Biesheuvel wrote:

On 6 December 2016 at 16:43, Dave Martin [off-list ref] wrote:

quoted

On Tue, Dec 06, 2016 at 03:39:00PM +0000, Ard Biesheuvel wrote:

quoted

Currently, we allow kernel mode NEON in softirq or hardirq context by
stacking and unstacking a slice of the NEON register file for each call
to kernel_neon_begin() and kernel_neon_end(), respectively.

Given that
a) a CPU typically spends most of its time in userland, during which time
   no kernel mode NEON in process context is in progress,
b) a CPU spends most of its time in the kernel doing other things than
   kernel mode NEON when it gets interrupted to perform kernel mode NEON
   in softirq context

the stacking and subsequent unstacking is only necessary if we are
interrupting a thread while it is performing kernel mode NEON in process
context, which means that in all other cases, we can simply preserve the
userland FPSIMD state once, and only restore it upon return to userland,
even if we are being invoked from softirq or hardirq context.

So instead of checking whether we are running in interrupt context, keep
track of the level of nested kernel mode NEON calls in progress, and only
perform the eager stack/unstack of the level exceeds 1.

[...]

quoted

quoted

+     this_cpu_write(kernel_neon_nesting_level, level + 1);

Should we BUG_ON overflow/underflow of the nesting level?  That Should
Not Happen (tm), but we'll make a mess if it does.

True.

quoted

For the underflow case, perhaps DEBUG_PREEMPT is adequate for detecting
this via preempt count underflow.

I think it makes sense for the increment to check for overflow and the
decrement to check for underflow, regardless of whether DEBUG_PREEMPT
(or just PREEMPT) is enabled or not.

Fair enough

Cheers
---Dave

[...]