From: Avi Kivity <avi@qumranet.com>
To: "David S. Ahern" <daahern@cisco.com>
Cc: kvm-devel <kvm-devel@lists.sourceforge.net>
Subject: Re: performance with guests running 2.4 kernels (specifically RHEL3)
Date: Sat, 26 Apr 2008 09:45:00 +0300 [thread overview]
Message-ID: <4812CF6C.3080903@qumranet.com> (raw)
In-Reply-To: <481215DE.3000302@cisco.com>
David S. Ahern wrote:
> David S. Ahern wrote:
>
>> Avi Kivity wrote:
>>
>>
>>> David S. Ahern wrote:
>>>
>>>
>>>> I added the traces and captured data over another apparent lockup of
>>>> the guest.
>>>> This seems to be representative of the sequence (pid/vcpu removed).
>>>>
>>>> (+4776) VMEXIT [ exitcode = 0x00000000, rip = 0x00000000
>>>> c016127c ]
>>>> (+ 0) PAGE_FAULT [ errorcode = 0x00000003, virt = 0x00000000
>>>> c0009db4 ]
>>>> (+3632) VMENTRY
>>>> (+4552) VMEXIT [ exitcode = 0x00000000, rip = 0x00000000
>>>> c016104a ]
>>>> (+ 0) PAGE_FAULT [ errorcode = 0x0000000b, virt = 0x00000000
>>>> fffb61c8 ]
>>>> (+ 54928) VMENTRY
>>>>
>>>>
>>> Can you oprofile the host to see where the 54K cycles are spent?
>>>
>>>
>
> Most of the cycles (~80% of that 54k+) are spent in paging64_prefetch_page():
>
> for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
> gpa_t pte_gpa = gfn_to_gpa(sp->gfn);
> pte_gpa += (i+offset) * sizeof(pt_element_t);
>
> r = kvm_read_guest_atomic(vcpu->kvm, pte_gpa, &pt,
> sizeof(pt_element_t));
> if (r || is_present_pte(pt))
> sp->spt[i] = shadow_trap_nonpresent_pte;
> else
> sp->spt[i] = shadow_notrap_nonpresent_pte;
> }
>
> This loop is run 512 times and takes a total of ~45k cycles, or ~88 cycles per
> loop.
>
> This function gets run >20,000/sec during some of the kscand loops.
>
>
We really ought to optimize it. That's second order however. The real
fix is making sure it isn't called so often.
--
Do not meddle in the internals of kernels, for they are subtle and quick to panic.
-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
next prev parent reply other threads:[~2008-04-26 6:45 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-16 0:15 performance with guests running 2.4 kernels (specifically RHEL3) David S. Ahern
2008-04-16 8:46 ` Avi Kivity
2008-04-17 21:12 ` David S. Ahern
2008-04-18 7:57 ` Avi Kivity
2008-04-21 4:31 ` David S. Ahern
2008-04-21 9:19 ` Avi Kivity
2008-04-21 17:07 ` David S. Ahern
2008-04-22 20:23 ` David S. Ahern
2008-04-23 8:04 ` Avi Kivity
2008-04-23 15:23 ` David S. Ahern
2008-04-23 15:53 ` Avi Kivity
2008-04-23 16:39 ` David S. Ahern
2008-04-24 17:25 ` David S. Ahern
2008-04-26 6:43 ` Avi Kivity
2008-04-26 6:20 ` Avi Kivity
2008-04-25 17:33 ` David S. Ahern
2008-04-26 6:45 ` Avi Kivity [this message]
2008-04-28 18:15 ` Marcelo Tosatti
2008-04-28 23:45 ` David S. Ahern
2008-04-30 4:18 ` David S. Ahern
2008-04-30 9:55 ` Avi Kivity
2008-04-30 13:39 ` David S. Ahern
2008-04-30 13:49 ` Avi Kivity
2008-05-11 12:32 ` Avi Kivity
2008-05-11 13:36 ` Avi Kivity
2008-05-13 3:49 ` David S. Ahern
2008-05-13 7:25 ` Avi Kivity
2008-05-14 20:35 ` David S. Ahern
2008-05-15 10:53 ` Avi Kivity
2008-05-17 4:31 ` David S. Ahern
[not found] ` <482FCEE1.5040306@qumranet.com>
[not found] ` <4830F90A.1020809@cisco.com>
2008-05-19 4:14 ` [kvm-devel] " David S. Ahern
2008-05-19 14:27 ` Avi Kivity
2008-05-19 16:25 ` David S. Ahern
2008-05-19 17:04 ` Avi Kivity
2008-05-20 14:19 ` Avi Kivity
2008-05-20 14:34 ` Avi Kivity
2008-05-22 22:08 ` David S. Ahern
2008-05-28 10:51 ` Avi Kivity
2008-05-28 14:13 ` David S. Ahern
2008-05-28 14:35 ` Avi Kivity
2008-05-28 19:49 ` David S. Ahern
2008-05-29 6:37 ` Avi Kivity
2008-05-28 14:48 ` Andrea Arcangeli
2008-05-28 14:57 ` Avi Kivity
2008-05-28 15:39 ` David S. Ahern
2008-05-29 11:49 ` Avi Kivity
2008-05-29 12:10 ` Avi Kivity
2008-05-29 13:49 ` David S. Ahern
2008-05-29 14:08 ` Avi Kivity
2008-05-28 15:58 ` Andrea Arcangeli
2008-05-28 15:37 ` Avi Kivity
2008-05-28 15:43 ` David S. Ahern
2008-05-28 17:04 ` Andrea Arcangeli
2008-05-28 17:24 ` David S. Ahern
2008-05-29 10:01 ` Avi Kivity
2008-05-29 14:27 ` Andrea Arcangeli
2008-05-29 15:11 ` David S. Ahern
2008-05-29 15:16 ` Avi Kivity
2008-05-30 13:12 ` Andrea Arcangeli
2008-05-31 7:39 ` Avi Kivity
2008-05-29 16:42 ` David S. Ahern
2008-05-31 8:16 ` Avi Kivity
2008-06-02 16:42 ` David S. Ahern
2008-06-05 8:37 ` Avi Kivity
2008-06-05 16:20 ` David S. Ahern
2008-06-06 16:40 ` Avi Kivity
2008-06-19 4:20 ` David S. Ahern
2008-06-22 6:34 ` Avi Kivity
2008-06-23 14:09 ` David S. Ahern
2008-06-25 9:51 ` Avi Kivity
2008-04-30 13:56 ` Daniel P. Berrange
2008-04-30 14:23 ` David S. Ahern
2008-04-23 8:03 ` Avi Kivity
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=4812CF6C.3080903@qumranet.com \
--to=avi@qumranet.com \
--cc=daahern@cisco.com \
--cc=kvm-devel@lists.sourceforge.net \
/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