On Mon, Aug 24, 2020 at 11:38:53PM +0800, Luck, Tony wrote:
> Yes, that's what we suspected. And I just did another try to
> percpu mce structure aligned. And the regression seems to be gone (reduced
> from 14.1% to 2%), which further proved it.
I wonder whether it would be useful for bisection of performance issues
for you to change the global definition of DEFINE_PER_CPU() to make
all per CPU definitions aligned. Just like you switch compiler flags to make
all functions aligned.
Thanks for the hint! This will definitely help tracking strange performance
changes like this, as a general debug mthod.