On 15.08.20 20:24, Julien Grall wrote:
Hi Oleksandr,
Hi Julien.
On 03/08/2020 19:21, Oleksandr Tyshchenko wrote:
From: Oleksandr Tyshchenko <oleksandr_tyshche...@epam.com>
Hello all.
The purpose of this patch series is to add IOREQ/DM support to Xen on
Arm.
You can find an initial discussion at [1]. Xen on Arm requires some
implementation
to forward guest MMIO access to a device model in order to implement
virtio-mmio
backend or even mediator outside of hypervisor. As Xen on x86 already
contains
required support this patch series tries to make it common and
introduce Arm
specific bits plus some new functionality. Patch series is based on
Julien's
PoC "xen/arm: Add support for Guest IO forwarding to a device emulator".
Besides splitting existing IOREQ/DM support and introducing Arm side,
the patch series also includes virtio-mmio related changes (toolstack)
for the reviewers to be able to see how the whole picture could look
like.
For a non-RFC, the IOREQ/DM and virtio-mmio support will be sent
separately.
According to the initial discussion there are a few open
questions/concerns
regarding security, performance in VirtIO solution:
1. virtio-mmio vs virtio-pci, SPI vs MSI, different use-cases require
different
transport...
2. virtio backend is able to access all guest memory, some kind of
protection
is needed: 'virtio-iommu in Xen' vs 'pre-shared-memory & memcpys
in guest'
3. interface between toolstack and 'out-of-qemu' virtio backend,
avoid using
Xenstore in virtio backend if possible.
4. a lot of 'foreing mapping' could lead to the memory exhaustion,
Julien
has some idea regarding that.
Looks like all of them are valid and worth considering, but the first
thing
which we need on Arm is a mechanism to forward guest IO to a device
emulator,
so let's focus on it in the first place.
***
Patch series [2] was rebased on Xen v4.14 release and tested on
Renesas Salvator-X
board + H3 ES3.0 SoC (Arm64) with virtio-mmio disk backend (we will
share it later)
running in driver domain and unmodified Linux Guest running on existing
virtio-blk driver (frontend). No issues were observed. Guest domain
'reboot/destroy'
use-cases work properly. Patch series was only build-tested on x86.
Please note, build-test passed for the following modes:
1. x86: CONFIG_HVM=y / CONFIG_IOREQ_SERVER=y (default)
2. x86: #CONFIG_HVM is not set / #CONFIG_IOREQ_SERVER is not set
3. Arm64: CONFIG_HVM=y / CONFIG_IOREQ_SERVER=y (default)
4. Arm64: CONFIG_HVM=y / #CONFIG_IOREQ_SERVER is not set
5. Arm32: CONFIG_HVM=y / #CONFIG_IOREQ_SERVER is not set
Build-test didn't pass for Arm32 mode with 'CONFIG_IOREQ_SERVER=y'
due to the lack of
cmpxchg_64 support on Arm32. See cmpxchg usage in
hvm_send_buffered_ioreq()).
I have sent a patch to implement cmpxchg64() and guest_cmpxchg64()
(see [1]).
Cheers,
[1]
https://lore.kernel.org/xen-devel/20200815172143.1327-1-jul...@xen.org/T/#u
Thank you! I have already build-tested it. No issues). I will update
corresponding patch to select IOREQ_SERVER for "config ARM" instead of
"config ARM64".
--
Regards,
Oleksandr Tyshchenko
