qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Alexander Graf <agraf@csgraf.de>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: qemu-devel@nongnu.org, qemu-arm@nongnu.org,
	"Yanan Wang" <wangyanan55@huawei.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	"Marcel Apfelbaum" <marcel.apfelbaum@gmail.com>,
	"Eduardo Habkost" <eduardo@habkost.net>,
	"Shashi Mallela" <shashi.mallela@linaro.org>,
	"Eric Auger" <eric.auger@redhat.com>,
	"Neil Armstrong" <narmstrong@baylibre.com>
Subject: Re: [PATCH 0/2] hw/intc/arm_gicv3: Bump ITT entry size to 16
Date: Fri, 10 Mar 2023 14:48:56 +0100	[thread overview]
Message-ID: <44c6459b-f17f-8cb7-ba2d-a9187f32cde0@csgraf.de> (raw)
In-Reply-To: <CAFEAcA-1PUCFZPAF25uy1VqjgciKXiYmfW-89q6QqKhf1io7Lw@mail.gmail.com>


On 03.01.23 18:41, Peter Maydell wrote:
> On Fri, 23 Dec 2022 at 08:50, Alexander Graf <agraf@csgraf.de> wrote:
>> While trying to make Windows work with GICv3 emulation, I stumbled over
>> the fact that it only supports ITT entry sizes that are power of 2 sized.
>>
>> While the spec allows arbitrary sizes, in practice hardware will always
>> expose power of 2 sizes and so this limitation is not really a problem
>> in real world scenarios. However, we only expose a 12 byte ITT entry size
>> which makes Windows blue screen on boot.
>>
>> The easy way to get around that problem is to bump the size to 16. That
>> is a power of 2, basically is what hardware would expose given the amount
>> of bits we need per entry and doesn't break any existing scenarios. To
>> play it safe, this patch set only bumps them on newer machine types.
> This is a Windows bug and should IMHO be fixed in that guest OS.
> Changing the ITT entry size of QEMU's implementation introduces
> an unnecessary incompatibility in migration and wastes memory
> (we're already a bit unnecessarily profligate with ITT entries
> compared to real hardware).


Follow-up on this: Microsoft has fixed the issue in Windows. That won't 
make older versions work, but the current should be fine with GICv3:

https://fosstodon.org/@itanium/109909281184181276


Alex




      parent reply	other threads:[~2023-03-10 13:50 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-23  8:50 [PATCH 0/2] hw/intc/arm_gicv3: Bump ITT entry size to 16 Alexander Graf
2022-12-23  8:50 ` [PATCH 1/2] hw/intc/arm_gicv3: Make ITT entry size configurable Alexander Graf
2023-03-10  4:55   ` Joelle van Dyne
2022-12-23  8:50 ` [PATCH 2/2] hw/intc/arm_gicv3: Bump ITT entry size to 16 Alexander Graf
2022-12-23 10:14 ` [PATCH 0/2] " Philippe Mathieu-Daudé
2023-01-03 17:41 ` Peter Maydell
2023-01-03 19:30   ` Alexander Graf
2023-03-10 13:48   ` Alexander Graf [this message]

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=44c6459b-f17f-8cb7-ba2d-a9187f32cde0@csgraf.de \
    --to=agraf@csgraf.de \
    --cc=eduardo@habkost.net \
    --cc=eric.auger@redhat.com \
    --cc=marcel.apfelbaum@gmail.com \
    --cc=narmstrong@baylibre.com \
    --cc=peter.maydell@linaro.org \
    --cc=philmd@linaro.org \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=shashi.mallela@linaro.org \
    --cc=wangyanan55@huawei.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).