Re: [PATCH v1] kernel: add a simple timer based software watchpoint

[Feng Tang]

Hi Thomas,

Thanks for reviewing this!

On Wed, Jun 24, 2026 at 11:04:26AM +0200, Thomas Gleixner wrote:
> On Tue, Jun 23 2026 at 16:26, Feng Tang wrote:
> > On Mon, Jun 22, 2026 at 04:13:37PM +0200, David Hildenbrand (Arm) wrote:
> > As discussed in RFC patch review, this debug feature is similar to
> > soft/hard lockup detector and task-hung detector, should I make the control
> 
> How is this very specialized ad hoc debug magic in any way similar to
> generally useful and just working debug mechanism like the lockup or
> hung detector? Those are just turned on, do not need a boatload of
> command line parameters and are generally useful.

That's right. They are very useful and easy to use, as a big part
of my time is dealing with all kinds of lockup/task-hung bugs :)


> Your debug magic is a workaround for a disfunctional hardware debugger,
> which means it's going to be used by three people twice a year if at
> all. Seriously?

The HW debugger is a Lautebach TRACE32 one, and may not be disfunctional,
as it works well for watchpoint virtual address and other daily job. My
own guess is that it can only see virtual address and doesn't have the
ability to do the virtual to physical address translation instantly to
watch a _physical_ address. So I guess, not able to watchpoint a physical
address may be common for HW debuggers (I could be very wrong).

As in https://lore.kernel.org/lkml/ajkuf08Cj0Se4P_0@U-2FWC9VHC-2323.local/,
we also used this method to solve one issue that BIOS runtime service
corrupting ACPI_ENABLE register issue.

Then I tried to recall some old memory corruption issues I've met before,
and think about if there is some that could be captured by this method,
one example was a static global array overflow issue, which corrupted 
some other global variables which was next to it in kernel bss segment.

But yes, as you pointed out, the frequency is low (all of the 3 happened
in the past 6 months) for myself. And my wild guess is there could be
other developers that meet similar issues :)

Thanks,
Feng

> Thanks,
> 
>         tglx
> 
>