Re: [PATCH] x86_64: (NEW) Dynamically allocate arch specific system vectors

[ebiederm@xmission.com (Eric W. Biederman)]

Ingo Molnar <mingo@elte.hu> writes:

> * Eric W. Biederman <ebiederm@xmission.com> wrote:
>
>> Alan Mayer <ajm@sgi.com> writes:
>> 
>> > Okay, here it is as an attachment.  I think my email client is munging it.
>> > I haven't been able to fix it, apparently.
>> >
>> > I, too, would like to know what Eric thinks.
>> 
>> I think arch/x86 is about to fall over from accidental complexity of 
>> the irq handling. [...]
>
> it was in that state for many years already ;-) Unification, cleanups of 
> other historic messes and the constant push for new hw support just made 
> it stand out more visibly. IRQ and APIC code unification is definitely 
> the final (and by far hardest) major step of x86 unification.

Well I think we are looking at the final couple of straws...

Yes unification (aka porting the infrastructure improvements
to x86_32) and then sharing the code is a piece of it.

>> [...]  Looking at your problem and the problem of killing NR_IRQS I 
>> spent way to much time playing with it this weekend then I should 
>> have, but I think I have found a path that works and is fairly easily 
>> verifiable.
>
> cool :-)
>
>> I have a patch series that gets me 90% of the way there, and the rest 
>> appears easy but I don't have any time to mess with it right now.  I 
>> will try and post it something in the next couple of days.
>
> i'm very interested in it, even if it's incomplete and wont build/boot 
> at all. So please consider posting your existing incomplete series as an 
> RFC right now, maybe we can help finish it sooner than you will find the 
> time? We can put it into a new tip/x86/irq-unification branch, without 
> merging it into tip/master just yet.

Sure. Mostly I just need to allocate a few minutes to post it, later today
or tomorrow hopefully.

Eric