Re: [PATCH 2/2] arm/ptw: respect sctlr.{u}wxn in get_phys_addr_v6

[Pavel Skripkin <paskripkin@gmail.com>]

Hi Peter,

Peter Maydell <peter.maydell@linaro.org> says:
> On Thu, 14 Nov 2024 at 16:59, Pavel Skripkin <paskripkin@gmail.com> wrote:
>>
>> get_phys_addr_v6() is used for decoding armv7's short descriptor format.
>> Based on ARM ARM AArch32.S1SDHasPermissionsFault(), WXN should be
>> respected in !LPAE mode as well.
>>
>> Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
>> ---
>>  target/arm/ptw.c | 11 +++++++++++
>>  1 file changed, 11 insertions(+)
> 
> Instead of this, would it be possible to have get_phys_addr_v6()
> call get_S1prot() (replacing most of the existing open-coded
> handling of PAN, simple vs classing AP model, etc) ? I haven't
> looked at the fine detail, so we might need to tweak get_S1prot()
> if it's missing logic that only matters in the short-descriptor
> case, but I think that would be better than having two places
> that need updating for new architectural features.
> 

I also thought about that, but suspected there was a reason it was not 
used at the first place.

Will take a look, if it's possible.

> For instance, we are also missing the handling for SCR.SIF with
> short descriptors which you can see in the S1SDHasPermissionsFault()
> pseudocode and which we would get if we could make the
> short-descriptor code call get_S1Prot().
> 
> Couple of minor notes for v2:
>   * if you're sending a multi-patch patchset, please use a
>     cover-letter email. Some of our automated tooling gets
>     confused by multi-patch patches without cover letters
>     (and, conversely, by single-patches with cover letters)
>   * our coding style says all if() statements should have
>     braces, even for single-line bodies
> 


Thanks for guidance!



With regards,
Pavel Skripkin