From: Anthony Liguori <anthony@codemonkey.ws>
To: Jan Kiszka <jan.kiszka@web.de>
Cc: Marcelo Tosatti <mtosatti@redhat.com>,
qemu-devel@nongnu.org, kvm@vger.kernel.org,
Alexander Graf <agraf@suse.de>
Subject: Re: [Qemu-devel] Re: [PATCH 26/35] kvm: Eliminate KVMState arguments
Date: Mon, 10 Jan 2011 14:11:34 -0600 [thread overview]
Message-ID: <4D2B67F6.5030909@codemonkey.ws> (raw)
In-Reply-To: <4D282489.90506@web.de>
On 01/08/2011 02:47 AM, Jan Kiszka wrote:
> OK, but I don't want to argue about the ioeventfd API. So let's put this
> case aside. :)
>
I often reply too quickly without explaining myself. Let me use
ioeventfd as an example to highlight why KVMState is a good thing.
In real life, PIO and MMIO are never directly communicated to the device
from the processor. Instead, they go through a series of other
devices. In the case of something like an ISA device, a PIO first goes
to the chipset into the PCI complex, it will then go through a
PCI-to-ISA bridge via subtractive decoding, and then forward over the
ISA device where it will be interpreted by some device.
The path to the chipset may be shared among different processors but it
may also be unique. The APIC is the best example as there are historic
APICs that hung directly off of the CPUs such that the same MMIO access
across different CPUs did not go to the same device. This is why the
APIC emulation in QEMU is so weird because we don't model this behavior
correctly.
This means that a PIO operation needs to flow from a CPUState to a
DeviceState. It can then flow through to another DeviceState until it's
finally handled.
The first problem with ioeventfd is that it's a per-VM operation. It
should be per VCPU.
But even if this were the case, the path that a PIO operation takes
should not be impacted by ioeventfd. IOW, a device shouldn't be
allocating an eventfd() and handing it to a magical KVM call. Instead,
a device should register a callback for a particular port in the same
way it always does. *As an optimization*, we should have another
interface that says that these values are only valid for this IO port.
That would let us create eventfds and register things behind the scenes.
That means we can handle TCG, older KVM kernels, and newer KVM kernels
without any special support in the device model. It also means that the
device models never have to worry about KVMState because there's an
entirely different piece of code that's consulting the set of special
ports and then deciding how to handle it. The result is better, more
portable code that doesn't have KVM-isms.
If passing state around seems to be ugly, it's probably because we're
not abstracting things correctly. Removing the state and making it
implicit is the wrong solution. Fixing the abstraction is the right
solution (or living with the ugliness until someone else is motivated to
fix it properly).
Regards,
Anthony Liguori
next prev parent reply other threads:[~2011-01-10 20:12 UTC|newest]
Thread overview: 300+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-06 17:56 [PATCH 00/35] [PULL] qemu-kvm.git uq/master queue Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 01/35] kvm: Enable user space NMI injection for kvm guest Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 02/35] kvm: convert kvm_ioctl(KVM_CHECK_EXTENSION) to kvm_check_extension() Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 03/35] Clean up cpu_inject_x86_mce() Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 04/35] Add "broadcast" option for mce command Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-09 18:51 ` Jan Kiszka
2011-01-09 18:51 ` [Qemu-devel] " Jan Kiszka
2011-01-15 16:24 ` Jan Kiszka
2011-01-15 16:24 ` [Qemu-devel] " Jan Kiszka
2011-01-06 17:56 ` [PATCH 05/35] Add function for checking mca broadcast of CPU Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 06/35] kvm: introduce kvm_mce_in_progress Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 07/35] kvm: kvm_mce_inj_* subroutines for templated error injections Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 08/35] kvm: introduce kvm_inject_x86_mce_on Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 09/35] kvm: x86: Fix DPL write back of segment registers Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 10/35] kvm: x86: Remove obsolete SS.RPL/DPL aligment Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 11/35] kvm: x86: Prevent sign extension of DR7 in guest debugging mode Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 12/35] kvm: x86: Fix a few coding style violations Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 13/35] kvm: Fix " Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 14/35] kvm: Drop return value of kvm_cpu_exec Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-08 13:09 ` Jan Kiszka
2011-01-08 13:09 ` [Qemu-devel] " Jan Kiszka
2011-01-06 17:56 ` [PATCH 15/35] kvm: Stop on all fatal exit reasons Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 16/35] kvm: Improve reporting of fatal errors Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 17/35] x86: Optionally dump code bytes on cpu_dump_state Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 18/35] kvm: x86: Align kvm_arch_put_registers code with comment Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 19/35] kvm: x86: Prepare kvm_get_mp_state for in-kernel irqchip Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 20/35] kvm: x86: Remove redundant mp_state initialization Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 21/35] kvm: x86: Fix xcr0 reset mismerge Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 22/35] kvm: x86: Refactor msr_star/hsave_pa setup and checks Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 23/35] kvm: x86: Reset paravirtual MSRs Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 24/35] Synchronize VCPU states before reset Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 25/35] kvm: x86: Drop MCE MSRs write back restrictions Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 26/35] kvm: Eliminate KVMState arguments Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 19:24 ` Anthony Liguori
2011-01-06 19:24 ` [Qemu-devel] " Anthony Liguori
2011-01-07 9:03 ` Jan Kiszka
2011-01-07 9:03 ` [Qemu-devel] " Jan Kiszka
2011-01-07 23:27 ` Anthony Liguori
2011-01-07 23:27 ` [Qemu-devel] " Anthony Liguori
2011-01-08 8:47 ` Jan Kiszka
2011-01-08 8:47 ` [Qemu-devel] " Jan Kiszka
2011-01-10 19:59 ` Anthony Liguori
2011-01-10 19:59 ` [Qemu-devel] " Anthony Liguori
2011-01-10 20:12 ` Jan Kiszka
2011-01-10 20:12 ` [Qemu-devel] " Jan Kiszka
2011-01-10 20:23 ` Anthony Liguori
2011-01-10 20:34 ` Jan Kiszka
2011-01-11 9:01 ` Avi Kivity
2011-01-11 9:01 ` Avi Kivity
2011-01-11 14:00 ` Anthony Liguori
2011-01-11 14:00 ` Anthony Liguori
2011-01-11 14:06 ` Alexander Graf
2011-01-11 14:06 ` Alexander Graf
2011-01-11 14:09 ` Anthony Liguori
2011-01-11 14:09 ` Anthony Liguori
2011-01-11 14:22 ` Avi Kivity
2011-01-11 14:22 ` Avi Kivity
2011-01-11 14:36 ` Anthony Liguori
2011-01-11 14:36 ` Anthony Liguori
2011-01-11 14:56 ` Avi Kivity
2011-01-11 14:56 ` Avi Kivity
2011-01-11 15:12 ` Anthony Liguori
2011-01-11 15:12 ` Anthony Liguori
2011-01-11 15:17 ` Alexander Graf
2011-01-11 15:17 ` Alexander Graf
2011-01-11 15:37 ` Avi Kivity
2011-01-11 15:37 ` Avi Kivity
2011-01-11 15:55 ` Anthony Liguori
2011-01-11 15:55 ` Anthony Liguori
2011-01-11 16:03 ` Avi Kivity
2011-01-11 16:03 ` Avi Kivity
2011-01-11 16:26 ` Anthony Liguori
2011-01-11 16:26 ` Anthony Liguori
2011-01-11 17:05 ` Avi Kivity
2011-01-11 17:05 ` Avi Kivity
2011-01-11 14:24 ` Alexander Graf
2011-01-11 14:24 ` Alexander Graf
2011-01-11 14:18 ` Avi Kivity
2011-01-11 14:18 ` Avi Kivity
2011-01-11 14:28 ` Anthony Liguori
2011-01-11 14:28 ` Anthony Liguori
2011-01-11 14:52 ` Avi Kivity
2011-01-11 14:52 ` Avi Kivity
2011-01-10 20:11 ` Anthony Liguori [this message]
2011-01-10 20:15 ` Jan Kiszka
2011-01-11 9:17 ` Avi Kivity
2011-01-11 9:17 ` Avi Kivity
2011-01-06 17:56 ` [PATCH 27/35] kvm: x86: Fix !CONFIG_KVM_PARA build Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 28/35] kvm: x86: Introduce kvmclock device to save/restore its state Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-10 20:31 ` Anthony Liguori
2011-01-10 20:31 ` Anthony Liguori
2011-01-10 21:06 ` Jan Kiszka
2011-01-10 21:06 ` Jan Kiszka
2011-01-10 22:21 ` Jan Kiszka
2011-01-10 22:21 ` Jan Kiszka
2011-01-10 23:02 ` Anthony Liguori
2011-01-10 23:02 ` Anthony Liguori
2011-01-11 5:54 ` Jan Kiszka
2011-01-11 5:54 ` Jan Kiszka
2011-01-11 8:00 ` Paolo Bonzini
2011-01-11 8:00 ` Paolo Bonzini
2011-01-11 8:53 ` Gerd Hoffmann
2011-01-11 8:53 ` Gerd Hoffmann
2011-01-11 17:13 ` Jan Kiszka
2011-01-11 17:13 ` Jan Kiszka
2011-01-11 9:31 ` Markus Armbruster
2011-01-11 9:31 ` Markus Armbruster
2011-01-11 13:54 ` Anthony Liguori
2011-01-11 13:54 ` Anthony Liguori
2011-01-12 10:22 ` Avi Kivity
2011-01-12 10:22 ` Avi Kivity
2011-01-12 10:31 ` Jan Kiszka
2011-01-12 10:31 ` Jan Kiszka
2011-01-18 14:28 ` Jan Kiszka
2011-01-18 14:28 ` Jan Kiszka
2011-01-18 15:04 ` Anthony Liguori
2011-01-18 15:04 ` Anthony Liguori
2011-01-18 15:43 ` Jan Kiszka
2011-01-18 15:43 ` Jan Kiszka
2011-01-18 15:48 ` Anthony Liguori
2011-01-18 15:48 ` Anthony Liguori
2011-01-18 15:54 ` Jan Kiszka
2011-01-18 15:54 ` Jan Kiszka
2011-01-18 17:02 ` Alex Williamson
2011-01-18 17:02 ` Alex Williamson
2011-01-18 17:08 ` Jan Kiszka
2011-01-18 17:08 ` Jan Kiszka
2011-01-18 17:39 ` Alex Williamson
2011-01-18 17:39 ` Alex Williamson
2011-01-18 15:50 ` Anthony Liguori
2011-01-18 15:50 ` Anthony Liguori
2011-01-18 16:01 ` Jan Kiszka
2011-01-18 16:01 ` Jan Kiszka
2011-01-18 16:04 ` Anthony Liguori
2011-01-18 16:04 ` Anthony Liguori
2011-01-18 16:17 ` Jan Kiszka
2011-01-18 16:17 ` Jan Kiszka
2011-01-18 16:37 ` Anthony Liguori
2011-01-18 16:37 ` Anthony Liguori
2011-01-18 16:56 ` Jan Kiszka
2011-01-18 16:56 ` Jan Kiszka
2011-01-18 17:09 ` Anthony Liguori
2011-01-18 17:09 ` Anthony Liguori
2011-01-18 17:20 ` Jan Kiszka
2011-01-18 17:20 ` Jan Kiszka
2011-01-18 17:31 ` Anthony Liguori
2011-01-18 17:31 ` Anthony Liguori
2011-01-18 17:45 ` Jan Kiszka
2011-01-18 17:45 ` Jan Kiszka
2011-01-19 9:48 ` Gerd Hoffmann
2011-01-19 9:48 ` Gerd Hoffmann
2011-01-19 13:11 ` Markus Armbruster
2011-01-19 13:11 ` Markus Armbruster
2011-01-19 16:54 ` Anthony Liguori
2011-01-19 16:54 ` Anthony Liguori
2011-01-19 17:19 ` Daniel P. Berrange
2011-01-19 17:19 ` Daniel P. Berrange
2011-01-19 17:43 ` Anthony Liguori
2011-01-19 17:43 ` Anthony Liguori
2011-01-20 8:44 ` Gerd Hoffmann
2011-01-20 8:44 ` Gerd Hoffmann
2011-01-20 10:33 ` Daniel P. Berrange
2011-01-20 10:33 ` Daniel P. Berrange
2011-01-20 19:42 ` Anthony Liguori
2011-01-20 19:42 ` Anthony Liguori
2011-01-20 19:39 ` Anthony Liguori
2011-01-20 19:39 ` Anthony Liguori
2011-01-21 8:35 ` Gerd Hoffmann
2011-01-21 8:35 ` Gerd Hoffmann
2011-01-21 10:03 ` Markus Armbruster
2011-01-21 10:03 ` Markus Armbruster
2011-01-19 16:53 ` Anthony Liguori
2011-01-19 16:53 ` Anthony Liguori
2011-01-19 17:01 ` Daniel P. Berrange
2011-01-19 17:01 ` Daniel P. Berrange
2011-01-19 17:51 ` Anthony Liguori
2011-01-19 17:51 ` Anthony Liguori
2011-01-19 18:52 ` Daniel P. Berrange
2011-01-19 18:52 ` Daniel P. Berrange
2011-01-19 18:58 ` Anthony Liguori
2011-01-19 18:58 ` Anthony Liguori
2011-01-19 17:35 ` Daniel P. Berrange
2011-01-19 17:35 ` Daniel P. Berrange
2011-01-19 17:42 ` Anthony Liguori
2011-01-19 17:42 ` Anthony Liguori
2011-01-19 18:53 ` Daniel P. Berrange
2011-01-19 18:53 ` Daniel P. Berrange
2011-01-19 13:09 ` Markus Armbruster
2011-01-19 13:09 ` Markus Armbruster
2011-01-24 8:45 ` Gleb Natapov
2011-01-24 8:45 ` Gleb Natapov
2011-01-19 13:15 ` Markus Armbruster
2011-01-19 13:15 ` Markus Armbruster
2011-01-19 16:57 ` Anthony Liguori
2011-01-19 16:57 ` [Qemu-devel] " Anthony Liguori
2011-01-19 17:25 ` Jan Kiszka
2011-01-19 17:25 ` Jan Kiszka
2011-01-19 19:32 ` Blue Swirl
2011-01-19 19:32 ` Blue Swirl
2011-01-20 9:33 ` Jan Kiszka
2011-01-20 9:33 ` Jan Kiszka
2011-01-20 19:27 ` Blue Swirl
2011-01-20 19:27 ` Blue Swirl
2011-01-20 21:22 ` Jan Kiszka
2011-01-20 21:22 ` Jan Kiszka
2011-01-20 21:40 ` Blue Swirl
2011-01-20 21:40 ` Blue Swirl
2011-01-20 21:53 ` Jan Kiszka
2011-01-20 21:53 ` Jan Kiszka
2011-01-25 11:10 ` Avi Kivity
2011-01-25 11:10 ` Avi Kivity
2011-01-21 8:46 ` Gerd Hoffmann
2011-01-21 8:46 ` Gerd Hoffmann
2011-01-21 10:05 ` Markus Armbruster
2011-01-21 10:05 ` Markus Armbruster
2011-01-21 16:37 ` Blue Swirl
2011-01-21 16:37 ` Blue Swirl
2011-01-21 17:21 ` Jan Kiszka
2011-01-21 17:21 ` Jan Kiszka
2011-01-21 18:04 ` Blue Swirl
2011-01-21 18:04 ` Blue Swirl
2011-01-21 18:17 ` Jan Kiszka
2011-01-21 18:17 ` Jan Kiszka
2011-01-21 18:49 ` Blue Swirl
2011-01-21 18:49 ` Blue Swirl
2011-01-24 14:08 ` Jan Kiszka
2011-01-24 14:08 ` Jan Kiszka
2011-01-24 21:35 ` Blue Swirl
2011-01-24 21:35 ` Blue Swirl
2011-01-24 21:57 ` Jan Kiszka
2011-01-24 21:57 ` Jan Kiszka
2011-01-20 19:37 ` Anthony Liguori
2011-01-20 19:37 ` Anthony Liguori
2011-01-20 20:02 ` Blue Swirl
2011-01-20 20:02 ` Blue Swirl
2011-01-20 21:42 ` Jan Kiszka
2011-01-20 21:42 ` Jan Kiszka
2011-01-20 21:27 ` Jan Kiszka
2011-01-20 21:27 ` Jan Kiszka
2011-01-25 11:06 ` Avi Kivity
2011-01-25 11:06 ` Avi Kivity
2011-01-25 14:30 ` Anthony Liguori
2011-01-25 14:30 ` Anthony Liguori
2011-01-25 10:34 ` Avi Kivity
2011-01-25 10:34 ` Avi Kivity
2011-01-25 10:27 ` Avi Kivity
2011-01-25 10:27 ` Avi Kivity
2011-01-25 13:58 ` Anthony Liguori
2011-01-25 13:58 ` Anthony Liguori
2011-01-12 12:04 ` Markus Armbruster
2011-01-12 12:04 ` Markus Armbruster
2011-01-10 23:04 ` Anthony Liguori
2011-01-10 23:04 ` Anthony Liguori
2011-01-11 5:55 ` Jan Kiszka
2011-01-11 5:55 ` Jan Kiszka
2011-01-06 17:56 ` [PATCH 29/35] kvm: Drop smp_cpus argument from init functions Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 30/35] kvm: Consolidate must-have capability checks Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 31/35] kvm: x86: Rework identity map and TSS setup for larger BIOS sizes Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 32/35] kvm: Flush coalesced mmio buffer on IO window exits Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 33/35] kvm: Do not use qemu_fair_mutex Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 34/35] kvm: x86: Implicitly clear nmi_injected/pending on reset Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-06 17:56 ` [PATCH 35/35] kvm: x86: Only read/write MSR_KVM_ASYNC_PF_EN if supported Marcelo Tosatti
2011-01-06 17:56 ` [Qemu-devel] " Marcelo Tosatti
2011-01-27 14:39 ` [PATCH] kvm: x86: Fix build in absence of KVM_CAP_ASYNC_PF Jan Kiszka
2011-01-27 14:39 ` [Qemu-devel] " Jan Kiszka
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=4D2B67F6.5030909@codemonkey.ws \
--to=anthony@codemonkey.ws \
--cc=agraf@suse.de \
--cc=jan.kiszka@web.de \
--cc=kvm@vger.kernel.org \
--cc=mtosatti@redhat.com \
--cc=qemu-devel@nongnu.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.