On Thu, May 21, 2020 at 09:38:58AM -0400, Prarit Bhargava wrote:
> On 5/21/20 9:37 AM, Neil Horman wrote:
> > On Thu, May 21, 2020 at 09:27:35AM -0400, Prarit Bhargava wrote:
> >> On 5/21/20 9:18 AM, Jiri Benc wrote:
> >>> On Thu, 21 May 2020 08:31:56 -0400, Prarit Bhargava wrote:
> >>>> We've used it (heavily might I add) to debug the kernel. It's a
> >>>> convenient
> >>>> place for a quick kernel module.
> >>>
> >>> Now I'm curious. How does such module help debugging the kernel?
> >>>
> >>> I'm not saying you're wrong. I'm just saying I can't think of anything.
> >>> Probably just me not being imaginative enough. More details would be
> >>> appreciated. It might come handy one time.
> >>>
> >>
> >> Well it's literally everything from dumping random values to confirming
> >> alignment of structs to debugging.
> >>
> >> What I'm doing today is attempting to break the Intel Split Lock detector
> >> by running
> >>
> >> for (cpu = 0; cpu < nr_cpu_ids; cpu++) {
> >> c = &cpu_data(cpu);
> >> printk("data for cpu %d\n", cpu);
> >> for (i = 0; i < (NCAPINTS + NBUGINTS * 32); i++) {
> >> cpu_has(c, i);
> >> }
> >> }
> >>
> >> in a tight loop. Unfortunately that doesn't seem to kick it. I suspect
> >> that
> >> the cpuinfo_x86 struct is magically aligned atm :(.
> >>
> >> I guess I could keep the dummy-module in it's own tree but that's just a
> >> PITA
> >> when I have to quickly debug something. Let's keep it in the tree where
> >> it's
> >> always been ;)
> >>
> > I remember using this module quite alot, and it was very helpful. I have
> > to ask
> > though, does systemtap offer the same level of functionality at this point
> > for
> > most of the debugging we need to do?
> >
>
> Yup! But I have had many cases when I'm digging for HW register values, etc.,
> that a module has been _invaluable_.
>
Good enough then, Ive got no problem with it.
Neil
> P.
>
> > Neil
> >
> >> P.
> >>
> >>> Thanks,
> >>>
> >>> Jiri
> >>>
> >> _______________________________________________
> >> kernel mailing list -- [email protected]
> >> To unsubscribe send an email to [email protected]
> >> Fedora Code of Conduct:
> >> https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> >> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> >> List Archives:
> >> https://lists.fedoraproject.org/archives/list/[email protected]
> >
>
_______________________________________________
kernel mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/[email protected]
