From: "hch@infradead.org" <hch@infradead.org>
To: Atish Patra <atish.patra@wdc.com>
Cc: linux-kernel@vger.kernel.org, Albert Ou <aou@eecs.berkeley.edu>,
Allison Randal <allison@lohutok.net>,
Anup Patel <anup@brainfault.org>,
linux-riscv@lists.infradead.org,
Palmer Dabbelt <palmer@sifive.com>,
Paul Walmsley <paul.walmsley@sifive.com>,
Andreas Schwab <schwab@linux-m68k.org>,
"hch@infradead.org" <hch@infradead.org>
Subject: Re: [v2 PATCH] RISC-V: Optimize tlb flush path.
Date: Mon, 19 Aug 2019 20:06:41 -0700 [thread overview]
Message-ID: <20190820030641.GA24946@infradead.org> (raw)
In-Reply-To: <20190820004735.18518-1-atish.patra@wdc.com>
On Mon, Aug 19, 2019 at 05:47:35PM -0700, Atish Patra wrote:
> In RISC-V, tlb flush happens via SBI which is expensive.
> If the target cpumask contains a local hartid, some cost
> can be saved by issuing a local tlb flush as we do that
> in OpenSBI anyways. There is also no need of SBI call if
> cpumask is empty.
>
> Do a local flush first if current cpu is present in cpumask.
> Invoke SBI call only if target cpumask contains any cpus
> other than local cpu.
Btw, you can use up your 70-ish chars per line for commit logs..
> + if (cpumask_test_cpu(cpuid, cmask)) {
> + /* Save trap cost by issuing a local tlb flush here */
> + if ((start == 0 && size == -1) || (size > PAGE_SIZE))
> + local_flush_tlb_all();
> + else if (size == PAGE_SIZE)
> + local_flush_tlb_page(start);
> + }
This looks a little odd to m and assumes we never pass a size smaller
than PAGE_SIZE. Whule that is probably true, why not something like:
if (size < PAGE_SIZE && size != -1)
local_flush_tlb_page(start);
else
local_flush_tlb_all();
?
WARNING: multiple messages have this Message-ID (diff)
From: "hch@infradead.org" <hch@infradead.org>
To: Atish Patra <atish.patra@wdc.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>,
Anup Patel <anup@brainfault.org>,
Palmer Dabbelt <palmer@sifive.com>,
linux-kernel@vger.kernel.org,
"hch@infradead.org" <hch@infradead.org>,
Andreas Schwab <schwab@linux-m68k.org>,
Paul Walmsley <paul.walmsley@sifive.com>,
linux-riscv@lists.infradead.org,
Allison Randal <allison@lohutok.net>
Subject: Re: [v2 PATCH] RISC-V: Optimize tlb flush path.
Date: Mon, 19 Aug 2019 20:06:41 -0700 [thread overview]
Message-ID: <20190820030641.GA24946@infradead.org> (raw)
In-Reply-To: <20190820004735.18518-1-atish.patra@wdc.com>
On Mon, Aug 19, 2019 at 05:47:35PM -0700, Atish Patra wrote:
> In RISC-V, tlb flush happens via SBI which is expensive.
> If the target cpumask contains a local hartid, some cost
> can be saved by issuing a local tlb flush as we do that
> in OpenSBI anyways. There is also no need of SBI call if
> cpumask is empty.
>
> Do a local flush first if current cpu is present in cpumask.
> Invoke SBI call only if target cpumask contains any cpus
> other than local cpu.
Btw, you can use up your 70-ish chars per line for commit logs..
> + if (cpumask_test_cpu(cpuid, cmask)) {
> + /* Save trap cost by issuing a local tlb flush here */
> + if ((start == 0 && size == -1) || (size > PAGE_SIZE))
> + local_flush_tlb_all();
> + else if (size == PAGE_SIZE)
> + local_flush_tlb_page(start);
> + }
This looks a little odd to m and assumes we never pass a size smaller
than PAGE_SIZE. Whule that is probably true, why not something like:
if (size < PAGE_SIZE && size != -1)
local_flush_tlb_page(start);
else
local_flush_tlb_all();
?
_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv
next prev parent reply other threads:[~2019-08-20 3:06 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-20 0:47 [v2 PATCH] RISC-V: Optimize tlb flush path Atish Patra
2019-08-20 0:47 ` Atish Patra
2019-08-20 3:06 ` hch [this message]
2019-08-20 3:06 ` hch
2019-08-20 7:14 ` Andreas Schwab
2019-08-20 7:14 ` Andreas Schwab
2019-08-20 7:16 ` hch
2019-08-20 7:16 ` hch
2019-08-20 7:46 ` Andreas Schwab
2019-08-20 7:46 ` Andreas Schwab
2019-08-20 8:42 ` Atish Patra
2019-08-20 8:42 ` Atish Patra
2019-08-20 8:51 ` Andreas Schwab
2019-08-20 8:51 ` Andreas Schwab
2019-08-20 9:22 ` hch
2019-08-20 9:22 ` hch
2019-08-20 20:28 ` Atish Patra
2019-08-20 20:28 ` Atish Patra
2019-08-20 22:18 ` hch
2019-08-20 22:18 ` hch
2019-08-20 22:24 ` Atish Patra
2019-08-20 22:24 ` Atish Patra
2019-08-21 1:29 ` Alan Kao
2019-08-21 1:29 ` Alan Kao
2019-08-21 1:40 ` hch
2019-08-21 1:40 ` hch
2019-08-21 3:52 ` Anup Patel
2019-08-21 3:52 ` Anup Patel
2019-08-21 7:18 ` hch
2019-08-21 7:18 ` hch
2019-08-20 8:51 ` Anup Patel
2019-08-20 8:51 ` Anup Patel
2019-08-20 20:29 ` Atish Patra
2019-08-20 20:29 ` Atish Patra
2019-08-21 14:41 ` hch
2019-08-21 14:41 ` hch
2019-08-21 14:45 ` hch
2019-08-21 14:45 ` hch
2019-08-21 17:36 ` Atish Patra
2019-08-21 17:36 ` Atish Patra
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=20190820030641.GA24946@infradead.org \
--to=hch@infradead.org \
--cc=allison@lohutok.net \
--cc=anup@brainfault.org \
--cc=aou@eecs.berkeley.edu \
--cc=atish.patra@wdc.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-riscv@lists.infradead.org \
--cc=palmer@sifive.com \
--cc=paul.walmsley@sifive.com \
--cc=schwab@linux-m68k.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.