From: Akira Yokosawa <akiyks@gmail.com>
To: Alan Huang <mmpgouride@gmail.com>
Cc: paulmck@kernel.org, perfbook@vger.kernel.org
Subject: Re: [PATCH] CodeSamples/count: add necessary memory barrier
Date: Mon, 27 Mar 2023 10:46:21 +0900 [thread overview]
Message-ID: <0a381404-e40d-268d-da91-f8c48753e39e@gmail.com> (raw)
In-Reply-To: <20230326162828.165472-1-mmpgouride@gmail.com>
Hello Alan,
On Sun, 26 Mar 2023 12:28:28 -0400, Alan Huang wrote:
> This patch add two necessary memory barriers, the first one added in
> flush_local_count makes sure the reading from theftp[t] happens before
> the reading from counterp[t]. The litmus testing below represents the
> pattern, and the result is "Sometimes":
Nice catch!
>
> C counter_sig
>
> {}
>
> P0(int *theft, int *counter)
> {
> int r0;
> int r1;
>
> r0 = READ_ONCE(*theft);
> r1 = READ_ONCE(*counter);
> }
>
> P1(int *theft, int *counter)
> {
> WRITE_ONCE(*counter, 1);
> smp_mb();
> WRITE_ONCE(*theft, 1);
> }
>
> exists (0:r0=1 /\ 0:r1=0)
To address this, instead of heavy-weight smp_mb(),
P0() can use smp_load_acquire():
P0(int *theft, int *counter)
{
int r0;
int r1;
r0 = smp_load_acquire(theft);
r1 = READ_ONCE(*counter);
}
>
> Second memory barrier is added to make sure that setting counterp[t]
> happens before the setting theftp[p] to THEFT_IDLE. Here is the litmus
> testing, The result is "Sometimes":
>
> C counter_sig_2
>
> {
> int theft = 1;
> int counter = 1;
> }
>
> P0(int *theft, int *counter)
> {
> WRITE_ONCE(*counter, 0);
> WRITE_ONCE(*theft, 0);
> }
>
> P1(int *theft, int *counter)
> {
> if (READ_ONCE(*theft) == 0)
> {
> smp_mb();
> WRITE_ONCE(*counter, READ_ONCE(*counter)+1);
> }
> }
>
> P2(int *counter)
> {
> int r0;
>
> r0 = READ_ONCE(*counter);
> }
>
> exists (2:r0=2)
Similarly, P0() can use smp_store_release():
P0(int *theft, int *counter)
{
WRITE_ONCE(*counter, 0);
smp_store_release(theft, 0);
}
Thanks, Akira
>
> Note that I added one smp_mb() in P1's "if" statement,
> because in add_count/sub_count's fast path, there is a
> control dependency.
>
> Signed-off-by: Alan Huang <mmpgouride@gmail.com>
> ---
> CodeSamples/count/count_lim_sig.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/CodeSamples/count/count_lim_sig.c b/CodeSamples/count/count_lim_sig.c
> index 023d6215..7bfeb004 100644
> --- a/CodeSamples/count/count_lim_sig.c
> +++ b/CodeSamples/count/count_lim_sig.c
> @@ -86,10 +86,12 @@ static void flush_local_count(void) //\lnlbl{flush:b}
> if (READ_ONCE(*theftp[t]) == THEFT_REQ)//\lnlbl{flush:check:REQ}
> pthread_kill(tid, SIGUSR1);//\lnlbl{flush:signal2}
> } //\lnlbl{flush:loop3:e}
> + smp_mb();
> globalcount += *counterp[t]; //\lnlbl{flush:thiev:b}
> *counterp[t] = 0;
> globalreserve -= *countermaxp[t];
> *countermaxp[t] = 0; //\lnlbl{flush:thiev:e}
> + smp_mb();
> WRITE_ONCE(*theftp[t], THEFT_IDLE); //\lnlbl{flush:IDLE}
> } //\lnlbl{flush:loop2:e}
> } //\lnlbl{flush:e}
prev parent reply other threads:[~2023-03-27 1:46 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-26 16:28 [PATCH] CodeSamples/count: add necessary memory barrier Alan Huang
2023-03-27 1:46 ` Akira Yokosawa [this message]
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=0a381404-e40d-268d-da91-f8c48753e39e@gmail.com \
--to=akiyks@gmail.com \
--cc=mmpgouride@gmail.com \
--cc=paulmck@kernel.org \
--cc=perfbook@vger.kernel.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 read-only IMAP folder(s) and NNTP newsgroup(s).