From: "Edwin Török" <edvin.torok@citrix.com>
To: <xen-devel@lists.xenproject.org>
Cc: "Andrew Cooper" <andrew.cooper3@citrix.com>,
"Wei Liu" <wei.liu2@citrix.com>,
"Edwin Török" <edvin.torok@citrix.com>,
"Jan Beulich" <jbeulich@suse.com>,
"Roger Pau Monné" <roger.pau@citrix.com>
Subject: [Xen-devel] [PATCH 0/1] Avoiding RDTSC emulation due to host clock drift
Date: Mon, 2 Sep 2019 19:27:06 +0100 [thread overview]
Message-ID: <cover.1567448405.git.edvin.torok@citrix.com> (raw)
I noticed that RDTSC emulation got turned on for a VM after a
suspend/host-reboot/resume cycle.
Xen currently expects an exact match between host CPU and saved guest CPU
frequency in KHz, otherwise it turns on RDTSC emulation if the CPU doesn't
support TSC scaling.
An exact match would require ~0.4 ppm accuracy, and even on physical hardware
the platform timer used for calibration is not that accurate. The best
accuracy I could find that datasheets/specifications require is 100 ppm, so let
Xen accept a 100 ppm difference in clock frequency as "the same" and do not
turn on RDTSC emulation due to that.
So far I have manually tested this on Intel(R) Xeon(R) CPU E5-2697 v3 and a
Debian 9 guest, more tests pending.
See the commit for more details.
Edwin Török (1):
x86/arch: VM resume: avoid RDTSC emulation due to host clock drift
xen/arch/x86/time.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
--
2.20.1
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
next reply other threads:[~2019-09-02 18:27 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-02 18:27 Edwin Török [this message]
2019-09-02 18:27 ` [Xen-devel] [PATCH 1/1] x86/arch: VM resume: avoid RDTSC emulation due to host clock drift Edwin Török
2019-09-03 7:54 ` Jan Beulich
2019-09-03 9:26 ` Edwin Török
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=cover.1567448405.git.edvin.torok@citrix.com \
--to=edvin.torok@citrix.com \
--cc=andrew.cooper3@citrix.com \
--cc=jbeulich@suse.com \
--cc=roger.pau@citrix.com \
--cc=wei.liu2@citrix.com \
--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.