[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20181119173904.GC2131@hirez.programming.kicks-ass.net>
Date: Mon, 19 Nov 2018 18:39:04 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: Aubrey Li <aubrey.li@...el.com>
Cc: tglx@...utronix.de, mingo@...hat.com, hpa@...or.com,
ak@...ux.intel.com, tim.c.chen@...ux.intel.com,
dave.hansen@...el.com, arjan@...ux.intel.com,
linux-kernel@...r.kernel.org, Aubrey Li <aubrey.li@...ux.intel.com>
Subject: Re: [PATCH v3 2/2] proc: add /proc/<pid>/arch_state
On Thu, Nov 15, 2018 at 07:00:07AM +0800, Aubrey Li wrote:
> Add a /proc/<pid>/arch_state interface to expose per-task cpu specific
> state values.
>
> Exposing AVX-512 Hi16_ZMM registers usage is for the user space job
> scheduler to cluster AVX-512 using tasks together, because these tasks
> could cause core turbo frequency drop.
I still don't much like the name; how about arch_simd_state ? Also,
since we're printing an integer, I still prefer we go print the turbo
license level. I know level 1 isn't too interesting atm, but consider
future hardware widening the thing again and us growing level 3 or
something.
Also; you were going to shop around with the other architectures to see
what they want/need for this interface. I see nothing on that.
Powered by blists - more mailing lists