Re: LSE Atomics

To: Tobias Nygren <tnn%NetBSD.org@localhost>
Subject: Re: LSE Atomics
From: Robert Swindells <rjs%fdy2.co.uk@localhost>
Date: Thu, 07 May 2020 16:34:42 +0100

Tobias Nygren <tnn%NetBSD.org@localhost> wrote:
>I don't think defaulting to the common denominator ISA is a good idea
>because big.LITTLE is implemented for battery savings and for that you
>do want to be on the core with hardware accelerated crypto when doing
>crypto.

Maybe we need a slightly higher level API to be able to ask "what
instructions can I use on the current CPU" than the current sysctl.

In the absense of a better API I think I will go with my current
patch to nss, it matches the pattern used elsewhere.

References:
- Re: LSE Atomics
  - From: Tobias Nygren

Prev by Date: Re: LSE Atomics
Next by Date: RPI3, serial consoles, gpio interrupts and -current
Previous by Thread: Re: LSE Atomics
Next by Thread: RPI3, serial consoles, gpio interrupts and -current
Indexes:

Home | Main Index | Thread Index | Old Index