Re: [PATCH v2 10/11] lockdep: Only trace IRQ edges

[Thomas Gleixner]

On Fri, Aug 21 2020 at 10:47, Peter Zijlstra wrote:
> Now, given that the only reason to use the raw_* variant is because you don't
> want tracing, A) seems like a weird option (although it can be done), so we
> pick B) and declare any code that ends up doing:
>
> 	raw_local_irq_save()
> 	local_irq_save()
> 	lockdep_assert_irqs_disabled();
>
> broken. AFAICT this problem has existed forever, the only reason it came
> up is because I changed IRQ tracing vs lockdep recursion and the first

Who is 'I'? I know you made that change and you also rewrote the
changelog, but I only figured that out after scrolling further down. As
this patch is authored by Nick, the above is simply inconsistent. Can we
please have just a reference to the commit which changed that code
instead of a puzzle?

Other than that:

Reviewed-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>