From: Carsten Otte <carsteno@de.ibm.com>
To: Avi Kivity <avi@redhat.com>
Cc: carsteno@linux.vnet.ibm.com, Alexander Graf <agraf@suse.de>,
qemu-devel <qemu-devel@nongnu.org>,
uli@suse.de, Carsten Otte <cotte@de.ibm.com>,
hare@suse.de, KVM list <kvm@vger.kernel.org>
Subject: Re: [PATCH 0/9] S390x KVM support
Date: Thu, 22 Oct 2009 12:43:54 +0200 [thread overview]
Message-ID: <4AE0376A.4070803@de.ibm.com> (raw)
In-Reply-To: <4AE033E2.5090902@redhat.com>
Avi Kivity wrote:
> On x86 we avoid emulating instructions in userspace. Instead the kernel
> requests userspace to do something (triggered by the instruction), and
> the kernel does anything which might be implied by the instruction (like
> copying the result into a register, or updating pc).
>
> An example is port I/O. instead of userspace reading %edx to query the
> port number and setting %eax to indicate the result, userspace reads a
> port number struct field and writes an I/O result struct field. Only
> the kernel accesses registers.
>
> I don't know whether that model makes sense or not for s390, but please
> consider it.
We do the same for many instructions (arch/s390/kvm/instruction.c). User
exits
are only performed for isntructions that cannot be handled in kernel.
Also, we do exit with requests to userspace, see the s390_reset exit
reason. I think in this regard, our implementation is very similar to
x86. Btw: this is something I did copycat from your implementation on
integration into kvm. The original zlive code did handle all
instructions in userland.
WARNING: multiple messages have this Message-ID (diff)
From: Carsten Otte <carsteno@de.ibm.com>
To: Avi Kivity <avi@redhat.com>
Cc: Carsten Otte <cotte@de.ibm.com>,
carsteno@linux.vnet.ibm.com, KVM list <kvm@vger.kernel.org>,
Alexander Graf <agraf@suse.de>,
qemu-devel <qemu-devel@nongnu.org>,
hare@suse.de
Subject: [Qemu-devel] Re: [PATCH 0/9] S390x KVM support
Date: Thu, 22 Oct 2009 12:43:54 +0200 [thread overview]
Message-ID: <4AE0376A.4070803@de.ibm.com> (raw)
In-Reply-To: <4AE033E2.5090902@redhat.com>
Avi Kivity wrote:
> On x86 we avoid emulating instructions in userspace. Instead the kernel
> requests userspace to do something (triggered by the instruction), and
> the kernel does anything which might be implied by the instruction (like
> copying the result into a register, or updating pc).
>
> An example is port I/O. instead of userspace reading %edx to query the
> port number and setting %eax to indicate the result, userspace reads a
> port number struct field and writes an I/O result struct field. Only
> the kernel accesses registers.
>
> I don't know whether that model makes sense or not for s390, but please
> consider it.
We do the same for many instructions (arch/s390/kvm/instruction.c). User
exits
are only performed for isntructions that cannot be handled in kernel.
Also, we do exit with requests to userspace, see the s390_reset exit
reason. I think in this regard, our implementation is very similar to
x86. Btw: this is something I did copycat from your implementation on
integration into kvm. The original zlive code did handle all
instructions in userland.
next prev parent reply other threads:[~2009-10-22 10:43 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-19 14:37 [Qemu-devel] [PATCH 0/9] S390x KVM support Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 1/9] Export function for VA defined ram allocation Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 2/9] Add KVM support for S390x Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 3/9] Add support for S390x system emulation Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 4/9] Add S390x virtio machine bus Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 5/9] Add S390x virtio machine description Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 6/9] S390 GDB stub Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 7/9] Implement early printk in virtio-console Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 8/9] Set default console to virtio on S390x Alexander Graf
2009-10-19 14:37 ` [Qemu-devel] [PATCH 9/9] Move mp_state to CPU_COMMON Alexander Graf
2009-10-19 19:48 ` [Qemu-devel] [PATCH 8/9] Set default console to virtio on S390x Gerd Hoffmann
2009-10-19 19:34 ` [Qemu-devel] [PATCH 4/9] Add S390x virtio machine bus Gerd Hoffmann
2009-10-19 19:40 ` Alexander Graf
2009-10-19 20:10 ` Gerd Hoffmann
2009-10-20 8:36 ` [Qemu-devel] Re: [PATCH 2/9] Add KVM support for S390x Carsten Otte
2009-10-20 8:41 ` Alexander Graf
2009-10-20 8:02 ` [Qemu-devel] Re: [PATCH 1/9] Export function for VA defined ram allocation Carsten Otte
2009-10-19 19:24 ` [Qemu-devel] [PATCH 0/9] S390x KVM support Gerd Hoffmann
2009-10-19 19:32 ` Alexander Graf
2009-10-20 5:05 ` Avi Kivity
2009-10-20 16:40 ` [Qemu-devel] " Carsten Otte
2009-10-21 7:20 ` Alexander Graf
2009-10-21 7:24 ` Alexander Graf
2009-10-21 8:18 ` Carsten Otte
2009-10-21 8:23 ` Alexander Graf
2009-10-21 7:26 ` Alexander Graf
2009-10-22 9:08 ` Avi Kivity
2009-10-22 9:08 ` [Qemu-devel] " Avi Kivity
2009-10-22 9:11 ` Alexander Graf
2009-10-22 9:11 ` [Qemu-devel] " Alexander Graf
2009-10-22 9:53 ` Avi Kivity
2009-10-22 9:53 ` [Qemu-devel] " Avi Kivity
2009-10-22 9:55 ` Alexander Graf
2009-10-22 9:55 ` [Qemu-devel] " Alexander Graf
2009-10-22 9:58 ` Alexander Graf
2009-10-22 9:58 ` [Qemu-devel] " Alexander Graf
2009-10-22 10:03 ` Avi Kivity
2009-10-22 10:03 ` [Qemu-devel] " Avi Kivity
2009-10-22 10:13 ` Alexander Graf
2009-10-22 10:13 ` [Qemu-devel] " Alexander Graf
2009-10-22 10:22 ` Carsten Otte
2009-10-22 10:22 ` [Qemu-devel] " Carsten Otte
2009-10-22 10:28 ` Avi Kivity
2009-10-22 10:28 ` [Qemu-devel] " Avi Kivity
2009-10-22 10:43 ` Carsten Otte [this message]
2009-10-22 10:43 ` Carsten Otte
2009-10-22 10:49 ` Avi Kivity
2009-10-22 10:49 ` [Qemu-devel] " Avi Kivity
2009-10-22 11:10 ` Carsten Otte
2009-10-22 11:10 ` [Qemu-devel] " Carsten Otte
2009-11-02 20:23 ` Alexander Graf
2009-11-02 20:23 ` [Qemu-devel] " Alexander Graf
2009-11-03 8:55 ` Avi Kivity
2009-11-03 8:55 ` [Qemu-devel] " Avi Kivity
2009-10-22 9:18 ` Carsten Otte
2009-10-22 9:18 ` [Qemu-devel] " Carsten Otte
2009-10-22 10:02 ` Avi Kivity
2009-10-22 10:02 ` [Qemu-devel] " Avi Kivity
2009-10-22 10:20 ` Carsten Otte
2009-10-22 10:20 ` [Qemu-devel] " Carsten Otte
2009-10-22 10:29 ` Avi Kivity
2009-10-22 10:29 ` [Qemu-devel] " 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=4AE0376A.4070803@de.ibm.com \
--to=carsteno@de.ibm.com \
--cc=agraf@suse.de \
--cc=avi@redhat.com \
--cc=carsteno@linux.vnet.ibm.com \
--cc=cotte@de.ibm.com \
--cc=hare@suse.de \
--cc=kvm@vger.kernel.org \
--cc=qemu-devel@nongnu.org \
--cc=uli@suse.de \
/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.