[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v2] hvf: arm: Ignore cache operations on MMIO

From: Richard Henderson
Subject: Re: [PATCH v2] hvf: arm: Ignore cache operations on MMIO
Date: Tue, 2 Nov 2021 14:19:52 -0400
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.13.0

On 11/2/21 7:15 AM, Richard Henderson wrote:
On 11/2/21 6:42 AM, Peter Maydell wrote:
On Tue, 2 Nov 2021 at 10:01, Richard Henderson
<richard.henderson@linaro.org> wrote:

On 11/1/21 3:35 PM, Peter Maydell wrote:
On Mon, 1 Nov 2021 at 19:28, Richard Henderson
<richard.henderson@linaro.org> wrote:

On 10/26/21 3:12 AM, Alexander Graf wrote:
Apple's Hypervisor.Framework forwards cache operations as MMIO traps
into user space. For MMIO however, these have no meaning: There is no
cache attached to them.

So let's just treat cache data exits as nops.

This fixes OpenBSD booting as guest.

Signed-off-by: Alexander Graf <agraf@csgraf.de>
Reported-by: AJ Barris <AwlsomeAlex@github.com>
Reference: https://github.com/utmapp/UTM/issues/3197
    target/arm/hvf/hvf.c | 7 +++++++
    1 file changed, 7 insertions(+)

Thanks, applied to target-arm.next

...did you see my email saying I think we also need
to test S1PTW ?

That arrived afterward.

Thinking it over later, I wouldn't be opposed to taking this
patch now and adding the S1PTW second -- I think we're
currently going to do the wrong thing for the "page tables
not in RAM case anyway", so you could regard it as a
separate bug fix.

Heh.  I was thinking the something similar, after I dequeued the patch.  That this fixes a problem emulating a well-behaved guest, and that the s1ptw hole is "merely" a problem vs an ill-behaved guest hitting the following assert.

Well, I've sent off the PR without this included, so I guess this should be the first cab off the rank for the next set.

I'm quite happy to hand the target-arm.next baton back to you.  ;-)

Since I need to re-spin the target-arm.next PR anyway, I've re-queued this.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]