linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
From: Nicholas Piggin <npiggin@gmail.com>
To: Christophe Leroy <christophe.leroy@c-s.fr>,
	linuxppc-dev@lists.ozlabs.org
Cc: "Aneesh Kumar K . V" <aneesh.kumar@linux.ibm.com>,
	Paul Mackerras <paulus@samba.org>
Subject: Re: [RFC PATCH] powerpc/64s/radix: introduce option to disable broadcast tlbie
Date: Thu, 01 Aug 2019 09:05:51 +1000	[thread overview]
Message-ID: <1564613600.hm3hrx4s31.astroid@bobo.none> (raw)
In-Reply-To: <188d2d1d-a254-d00c-eee4-dd71e01b443f@c-s.fr>

Christophe Leroy's on July 31, 2019 11:56 pm:
> 
> 
> Le 31/07/2019 à 14:32, Nicholas Piggin a écrit :
>> This is an initial hack of a quick option to disable use of the tlbie
>> instruction. This takes the simplest possible initial pass of just
>> replacing low level tlbie functions with IPIs. This means it's not as
>> performant as it could be if we spend some time optmizing it, but on
>> the other hand having a 1:1 replacement of tlbie is simple and can be
>> useful for comparisons so I think it's the right initial approach.
> 
> Can you explain why we want to optionnaly disable use of tlbie ?

It's something we've wanted to have more control of for a while.

One is for testing performance, especially on large systems it is not
always best to use tlbie. For example if you have threaded apps that
only run on a few CPUs (e.g., process * thread hybrid process model
server).

There is also concern about coherent accelerators may have high
latency to respond to tlbie, which can block the resource for others.

And we did have a hardware errata with early POWER9 it would have
been nice to disable it until the kernel could be updated with the
workaround.

Thanks,
Nick

  reply	other threads:[~2019-07-31 23:09 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-31 12:32 [RFC PATCH] powerpc/64s/radix: introduce option to disable broadcast tlbie Nicholas Piggin
2019-07-31 13:56 ` Christophe Leroy
2019-07-31 23:05   ` Nicholas Piggin [this message]
2019-08-14  2:06 ` Michael Ellerman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1564613600.hm3hrx4s31.astroid@bobo.none \
    --to=npiggin@gmail.com \
    --cc=aneesh.kumar@linux.ibm.com \
    --cc=christophe.leroy@c-s.fr \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=paulus@samba.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).