From: stsp <stsp2@yandex.ru>
To: Sean Christopherson <seanjc@google.com>,
Jim Mattson <jmattson@google.com>
Cc: kvm list <kvm@vger.kernel.org>
Subject: Re: guest/host mem out of sync on core2duo?
Date: Sat, 19 Jun 2021 03:11:42 +0300 [thread overview]
Message-ID: <4834cc76-72d5-4d23-7a56-63e455683db5@yandex.ru> (raw)
In-Reply-To: <YM0fBtqYe+VyPME7@google.com>
19.06.2021 01:32, Sean Christopherson пишет:
> Argh! Check out this gem:
>
> /*
> * Fix the "Accessed" bit in AR field of segment registers for older
> * qemu binaries.
> * IA32 arch specifies that at the time of processor reset the
> * "Accessed" bit in the AR field of segment registers is 1. And qemu
> * is setting it to 0 in the userland code. This causes invalid guest
> * state vmexit when "unrestricted guest" mode is turned on.
> * Fix for this setup issue in cpu_reset is being pushed in the qemu
> * tree. Newer qemu binaries with that qemu fix would not need this
> * kvm hack.
> */
> if (is_unrestricted_guest(vcpu) && (seg != VCPU_SREG_LDTR))
> var->type |= 0x1; /* Accessed */
>
>
> KVM fixes up segs when unrestricted guest is enabled, but otherwise leaves 'em
> be, presumably because it has the emulator to fall back on for invalid state.
> Guess what's missing in the invalid state check...
>
> I think this should do it:
Until when will it run on an
emulator in this case?
Will it be too slow without a
slightest hint to the user?
If it is indeed the performance
penalty for no good reason,
then my preference would be
to get an error right from
KVM_SET_SREGS instead,
or maybe from KVM_RUN,
but not run everything on
an emulator.
next prev parent reply other threads:[~2021-06-19 0:11 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-12 22:49 guest/host mem out of sync on core2duo? stsp
2021-06-13 12:36 ` stsp
2021-06-14 17:06 ` Sean Christopherson
2021-06-14 17:32 ` stsp
2021-06-17 14:42 ` Sean Christopherson
2021-06-18 15:59 ` stsp
2021-06-18 21:07 ` Jim Mattson
2021-06-18 21:55 ` stsp
2021-06-18 22:06 ` Jim Mattson
2021-06-18 22:26 ` stsp
2021-06-18 22:32 ` Sean Christopherson
2021-06-19 0:11 ` stsp [this message]
2021-06-19 0:54 ` Sean Christopherson
2021-06-19 9:18 ` stsp
2021-06-21 2:34 ` exception vs SIGALRM race (was: Re: guest/host mem out of sync on core2duo?) stsp
2021-06-21 22:33 ` Jim Mattson
2021-06-21 23:32 ` stsp
2021-06-22 0:27 ` stsp
2021-06-28 21:47 ` Jim Mattson
2021-06-28 21:50 ` stsp
2021-06-28 22:00 ` stsp
2021-06-28 22:27 ` Jim Mattson
2021-07-06 16:28 ` Paolo Bonzini
2021-07-06 22:22 ` stsp
2021-07-06 23:41 ` Paolo Bonzini
2021-06-23 23:38 ` exception vs SIGALRM race (with test-case now!) stsp
2021-06-24 0:11 ` stsp
2021-06-24 0:25 ` stsp
2021-06-24 18:05 ` exception vs SIGALRM race on core2 CPUs (with qemu-based test-case this time!) stsp
2021-06-24 18:07 ` stsp
2021-06-25 23:35 ` exception vs SIGALRM race on core2 CPUs (with fix!) stsp
2021-06-26 0:15 ` Jim Mattson
2021-06-26 0:35 ` stsp
2021-06-26 21:50 ` stsp
2021-06-27 12:13 ` stsp
2021-06-26 14:03 ` exception vs SIGALRM race (another patch) stsp
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=4834cc76-72d5-4d23-7a56-63e455683db5@yandex.ru \
--to=stsp2@yandex.ru \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=seanjc@google.com \
/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 NNTP newsgroup(s).