From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Jan Beulich <jbeulich@suse.com>
Cc: "Stefano Stabellini" <sstabellini@kernel.org>,
"Julien Grall" <julien@xen.org>, "Wei Liu" <wl@xen.org>,
Xen-devel <xen-devel@lists.xenproject.org>,
"Volodymyr Babchuk" <Volodymyr_Babchuk@epam.com>,
"Roger Pau Monné" <roger.pau@citrix.com>
Subject: Re: [Xen-devel] [PATCH 4/6] xen/hypercall: Cope with -ERESTART on more hypercall paths
Date: Mon, 9 Dec 2019 17:43:29 +0000 [thread overview]
Message-ID: <5ac03858-2ed8-7c4b-82c9-a1ebafa164bb@citrix.com> (raw)
In-Reply-To: <69f66b76-7352-dba2-bd04-2ffe2840a30b@suse.com>
On 09/12/2019 16:29, Jan Beulich wrote:
> On 09.12.2019 17:25, Jan Beulich wrote:
>> On 05.12.2019 23:30, Andrew Cooper wrote:
>>> These hypercalls each use continue_hypercall_on_cpu(), whose API is about to
>>> switch to use -ERESTART. Update the soon-to-be affected paths to cope,
>>> folding existing contination logic where applicable.
>>>
>>> In addition:
>>> * For platform op and sysctl, insert a cpu_relax() into what is otherwise a
>>> tight spinlock loop, and make the continuation logic common at the
>>> epilogue.
>> Is this really needed with a hypercall_preempt_check() invocation
>> already in the bodies of these loops?
Yes. The reason you're supposed to pause is to stop having memory
traffic constantly trying to pull the spinlock's cacheline into shared
state.
Racing round a tight loop constantly reading 4 or 5 memory locations is
almost as bad.
> And if it's really to be added, shouldn't it be at the bottom
> of the loop bodies rather than at the top?
It doesn't matter.
~Andrew
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
next prev parent reply other threads:[~2019-12-09 17:43 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-05 22:30 [Xen-devel] [PATCH 0/6] xen: Support continuations from tasklets Andrew Cooper
2019-12-05 22:30 ` [Xen-devel] [PATCH 1/6] xen/tasklet: Fix return value truncation on arm64 Andrew Cooper
2019-12-08 11:57 ` Julien Grall
2019-12-05 22:30 ` [Xen-devel] [PATCH 2/6] xen/tasklet: Switch data parameter from unsigned long to void * Andrew Cooper
2019-12-08 12:02 ` Julien Grall
2019-12-09 15:28 ` Jan Beulich
2019-12-05 22:30 ` [Xen-devel] [PATCH 3/6] xen/domctl: Consolidate hypercall continuation handling at the top level Andrew Cooper
2019-12-08 12:18 ` Julien Grall
2019-12-09 17:20 ` Andrew Cooper
2019-12-09 16:19 ` Jan Beulich
2019-12-09 17:29 ` Andrew Cooper
2019-12-10 8:09 ` Jan Beulich
2019-12-05 22:30 ` [Xen-devel] [PATCH 4/6] xen/hypercall: Cope with -ERESTART on more hypercall paths Andrew Cooper
2019-12-08 12:57 ` Julien Grall
2019-12-09 17:37 ` Andrew Cooper
2019-12-11 12:01 ` Julien Grall
2019-12-09 16:25 ` Jan Beulich
2019-12-09 16:29 ` Jan Beulich
2019-12-09 17:43 ` Andrew Cooper [this message]
2019-12-10 8:27 ` Jan Beulich
2019-12-05 22:30 ` [Xen-devel] [PATCH 5/6] xen/tasklet: Return -ERESTART from continue_hypercall_on_cpu() Andrew Cooper
2019-12-09 16:52 ` Jan Beulich
2019-12-09 17:49 ` Andrew Cooper
2019-12-10 8:55 ` Jan Beulich
2019-12-10 17:55 ` Andrew Cooper
2019-12-11 7:41 ` Jan Beulich
2019-12-11 9:00 ` Andrew Cooper
2019-12-05 22:30 ` [Xen-devel] [PATCH 6/6] x86/smt: Don't use -EBUSY for smt_up_down_helper() continuations Andrew Cooper
2019-12-10 10:29 ` Jan Beulich
2019-12-06 9:58 ` [Xen-devel] [PATCH 0/6] xen: Support continuations from tasklets Jan Beulich
2019-12-06 10:14 ` Andrew Cooper
2019-12-06 10:18 ` Jan Beulich
2019-12-06 10:22 ` Andrew Cooper
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=5ac03858-2ed8-7c4b-82c9-a1ebafa164bb@citrix.com \
--to=andrew.cooper3@citrix.com \
--cc=Volodymyr_Babchuk@epam.com \
--cc=jbeulich@suse.com \
--cc=julien@xen.org \
--cc=roger.pau@citrix.com \
--cc=sstabellini@kernel.org \
--cc=wl@xen.org \
--cc=xen-devel@lists.xenproject.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.