Re: [patch v2 04/14] x86/irq: Make irqstats array based

[Florian Fainelli]

On 3/26/26 03:29, Thomas Gleixner wrote:
> Florian!
>
> On Wed, Mar 25 2026 at 15:54, Florian Fainelli wrote:
> > On 3/25/26 15:52, Thomas Gleixner wrote:
> > > Sigh, yes. This GDB python insanity is a pain, debugging it even more so.
> >
> > All matter of perspective but kudos to you for trying to fix it, that is
> > much appreciated.
>
> What bothers me most on the kernel side is that there is neither
> annotation, which tells something is used in a gdb script, nor a compile
> time check.
>
> So if Radu hadn't noticed and pointed it out, I wouldn't have known
> about it at all.

Yes that's a fair point and it has been a complaint of mine whenever I 
had to use the GDB scripts because they get out of sync so easily with 
the code they are trying to parse, at some point it's not even fun.

This should still apply if you want to take it:

https://lore.kernel.org/all/20250625231053.1134589-6-florian.fainelli@xxxxxxxxxxxx/

As far as ensuring that the C/Rust code stays in sync with GDB at 
compile time, did not have any bright idea about how to do that.

> The insanity snark was just my annoyance about the python integration
> into gdb. It's amazing that a debugger provides an extension interface
> which is undebuggable. But admittedly it's at least consistent with the
> rest of GDB's usability.

Oh yes, that's something that was learned the hard way on my side too, 
not sure why that's not the default, much hair pulling resulted from 
having incomplete backtraces.
--
Florian