Re: ARM64 notes: Re: [PATCH] CI: Extend eclair-*-allcode to enable as much as possible

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 05/01/2026 9:12 pm, Nicola Vetrini wrote:
> On 2026-01-05 19:14, Andrew Cooper wrote:
>>
>> llc-colouring.c uses binary literals.  These are safe to use now since
>> 4.21 with the updated toolchain baseline, but the Eclair config wants
>> updating to allow this language extension.
>>
>
> Yeah (though I don't see a strong reason to do so, for a single
> literal); I can write the patch.

A separate discussion happened about starting to use binary literals
more widely.  It's a capability we'd like to be able to use.

>
> Also xen/arch/arm/acpi/boot.c could use __func__ as almost everywhere
> else in xen/

Yeah (although I considered that not interesting enough to discuss.)

>
>> ipmmu-vmsa.c has a git:// url inside a block comment, which is
>> considered to be a Rule 3.1 violation.  In principle this ought to
>> fix it:
>>
>
> Indeed it should.
>
>> diff --git a/automation/eclair_analysis/ECLAIR/deviations.ecl
>> b/automation/eclair_analysis/ECLAIR/deviations.ecl
>> index 7dee4a488d45..8f5fc6c93bc5 100644
>> --- a/automation/eclair_analysis/ECLAIR/deviations.ecl
>> +++ b/automation/eclair_analysis/ECLAIR/deviations.ecl
>> @@ -60,7 +60,7 @@ removed by the compiler, the resulting slowdown is
>> negligible."
>>  
>>  -doc_begin="Comments starting with '/*' and containing hyperlinks
>> are safe as
>>  they are not instances of commented-out code."
>> --config=MC3A2.R3.1,reports+={safe, "first_area(text(^.*https?://.*$))"}
>> +-config=MC3A2.R3.1,reports+={safe,
>> "first_area(text(^.*(https?|git)://.*$))"}
>>  -doc_end
>>  
>>  #
>>
>>
>> but I've not tried it yet.
>>
>> There's a R8.4 violation against __stack_chk_guard.  I think this wants
>> deviating locally, because it's a fairly magic construct.
>>
>
> ack.

For the x86 side,
https://eclair-analysis-logs.xenproject.org/fs/space/verdesse0/XEN.ecdf/xen-project/hardware/xen-staging/ECLAIR_normal/andrew/eclair/X86_64/12595699289/PROJECT.ecd;/by_service/MC3A2.R8.4.html
shows that we've got the same problem with pvh_start_info_pa and
early_hypercall_insn.

Like __stack_chk_guard, these are external because they're accessed by
assembly, but don't need/want to be declared anywhere else in C.  What's
the recommended way of handling these issues?

~Andrew