[PATCH 1/2] arm64: Reduce add_map() complexity

[Marc Zyngier]

On Sat, 09 Mar 2024 12:29:10 +0000,
Fabio Estevam <festevam at gmail.com> wrote:
> 
> Hi Marc,
> 
> On Sat, Mar 9, 2024 at 6:53 AM Marc Zyngier <maz at kernel.org> wrote:
> 
> > It would be good to narrow down which access is generating this. It is
> > an asynchronous error, so the code above won't help.
> >
> > Alternatively, and if you are sure that this is due to this change,
> > dumping the page tables and comparing them before and after would
> > help.
> 
> Yes, I am sure the error is due to this change. It is 100% reproducible.

Can you figure out what memory access is triggering it? Even at
narrowing it to the subsystem level would be a good indication.

> I am not familiar with this part of the code, so I would appreciate
> it if you could tell me how to dump the page tables so I can compare
> them before and after.

You could just dump the entries as they are written. The order may not
be the same, but for a given VA you should observe the same entries
being written.

My hunch is that the new code is a lot more picky about the alignment
of things, and that could result in something being similarly
unaligned. But without access to the platform nor an idea of what gets
mapped, it's a bit hard to have a clue.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.