Re: [PATCH v2 2/3] target/arm/tcg/m_helper.c: Include the full helpers only with CONFIG_ARM_V7M

[Thomas Huth <thuth@redhat.com>]

On 08/03/2024 15.00, Peter Maydell wrote:
> On Fri, 8 Mar 2024 at 12:54, Thomas Huth <thuth@redhat.com> wrote:
>> I gave it a try, but then we end up again with the problem that I already
>> mentioned in the discussion about patch 1: CONFIG_ARM_V7M is not set for the
>> linux-user binaries, so m_helper.c would not get included there anymore and
>> we end up with lots of link failures.
>>
>> So if you don't like the current shape, I guess this needs a little bit more
>> pondering 'til it gets acceptable.
>>
>> But could you maybe at least pick up the first patch already? ... since it's
>> a patch with lots of code movement in it, this is quite ugly to rebase it
>> each time someone touches some lines of code in that area...
> 
> I don't object to taking the first patch, but... it doesn't apply
> so it needs rebasing :-/ If you can rebase and send it out this
> afternoon I can put it in a pullreq I'm working on.

That's what I meant with it's "ugly to rebase it" ;-) ... fortunately, it 
was only a rather simple conflict this time.

I sent out a v3 now, also including another try for the second patch (which 
now includes the stubs that assert() in the m_helper.c file itself instead 
of using a separate stub file). Not sure whether that looks much nicer, but 
at least it seems to work from a compilation/linking perspective. Anyway, if 
you could at least include patch 1 from that v3 in your next pull request, 
that would be great!

  Thanks,
   Thomas