Re: [PATCH 0/3] Modernize tasklet callback API

Re: [PATCH 0/3] Modernize tasklet callback API

[Kees Cook]

[heavily trimmed CC list because I think lkml is ignoring this
thread...]

On Thu, Jul 30, 2020 at 09:03:55AM +0200, Thomas Gleixner wrote:
> Kees,
> 
> Kees Cook <keescook@chromium.org> writes:
> > This is the infrastructure changes to prepare the tasklet API for
> > conversion to passing the tasklet struct as the callback argument instead
> > of an arbitrary unsigned long. The first patch details why this is useful
> > (it's the same rationale as the timer_struct changes from a bit ago:
> > less abuse during memory corruption attacks, more in line with existing
> > ways of doing things in the kernel, save a little space in struct,
> > etc). Notably, the existing tasklet API use is much less messy, so there
> > is less to clean up.
> >
> > It's not clear to me which tree this should go through... Greg since it
> > starts with a USB clean-up, -tip for timer or interrupt, or if I should
> > just carry it. I'm open to suggestions, but if I don't hear otherwise,
> > I'll just carry it.
> >
> > My goal is to have this merged for v5.9-rc1 so that during the v5.10
> > development cycle the new API will be available. The entire tree of
> > changes is here[1] currently, but to split it up by maintainer the
> > infrastructure changes need to be landed first.
> >
> > Review and Acks appreciated! :)
> 
> I'd rather see tasklets vanish from the planet completely, but that's
> going to be a daring feat. So, grudgingly:

Understood! I will update the comments near the tasklet API.

> Acked-by: Thomas Gleixner <tglx@linutronix.de>

Thanks!

-- 
Kees Cook

Re: [PATCH 0/3] Modernize tasklet callback API

[Allen]

Kees,

>
> [heavily trimmed CC list because I think lkml is ignoring this
> thread...]
>
> On Thu, Jul 30, 2020 at 09:03:55AM +0200, Thomas Gleixner wrote:
> > Kees,
> >
> > Kees Cook <keescook@chromium.org> writes:
> > > This is the infrastructure changes to prepare the tasklet API for
> > > conversion to passing the tasklet struct as the callback argument instead
> > > of an arbitrary unsigned long. The first patch details why this is useful
> > > (it's the same rationale as the timer_struct changes from a bit ago:
> > > less abuse during memory corruption attacks, more in line with existing
> > > ways of doing things in the kernel, save a little space in struct,
> > > etc). Notably, the existing tasklet API use is much less messy, so there
> > > is less to clean up.
> > >
> > > It's not clear to me which tree this should go through... Greg since it
> > > starts with a USB clean-up, -tip for timer or interrupt, or if I should
> > > just carry it. I'm open to suggestions, but if I don't hear otherwise,
> > > I'll just carry it.
> > >
> > > My goal is to have this merged for v5.9-rc1 so that during the v5.10
> > > development cycle the new API will be available. The entire tree of
> > > changes is here[1] currently, but to split it up by maintainer the
> > > infrastructure changes need to be landed first.
> > >
> > > Review and Acks appreciated! :)
> >
> > I'd rather see tasklets vanish from the planet completely, but that's
> > going to be a daring feat. So, grudgingly:
>
> Understood! I will update the comments near the tasklet API.
>
> > Acked-by: Thomas Gleixner <tglx@linutronix.de>
>

Here's the series re-based on top of 5.8
https://github.com/allenpais/tasklets/tree/V3

Let me know how you would want these to be reviewed.

Also, I was thinking if removing tasklets completely could be a task
on KSPP wiki. If yes, I did like to take ownership of that task. I have a
couple of ideas in mind, which could be discussed in a separate email.

Thanks.

-- 
       - Allen

Re: [PATCH 0/3] Modernize tasklet callback API

[Allen]

Kees,

> >
>
> Here's the series re-based on top of 5.8
> https://github.com/allenpais/tasklets/tree/V3
>
> Let me know how you would want these to be reviewed.
>

  I see the first set of infrastructure patches for tasklets have
landed in Linus's tree. Good time to send out the ~200 patches?

- Allen

Re: [PATCH 0/3] Modernize tasklet callback API

[Kees Cook]

On Mon, Aug 03, 2020 at 02:16:15PM +0530, Allen wrote:
> Here's the series re-based on top of 5.8
> https://github.com/allenpais/tasklets/tree/V3

Great!

> Let me know how you would want these to be reviewed.

Was a Coccinelle script used for any of these conversions? I wonder if
it'd be easier to do a single treewide patch for the more mechanical
changes.

And, actually, I still think the "prepare" patches should just be
collapsed into the actual "covert" patches -- there are only a few.

After those, yeah, I think getting these sent to their respective
maintainers is the next step.

> Also, I was thinking if removing tasklets completely could be a task
> on KSPP wiki. If yes, I did like to take ownership of that task. I have a
> couple of ideas in mind, which could be discussed in a separate email.

Sure! I will add it to the tracker. Here's for the refactoring:
https://github.com/KSPP/linux/issues/30

and here's for the removal:
https://github.com/KSPP/linux/issues/94

if you can added details/examples of how they should be removed, that'd
help other folks too, if they wanted to jump in. :)

-Kees

-- 
Kees Cook

Re: [PATCH 0/3] Modernize tasklet callback API

[Takashi Iwai]

On Tue, 11 Aug 2020 23:33:13 +0200,
Kees Cook wrote:
> 
> On Mon, Aug 03, 2020 at 02:16:15PM +0530, Allen wrote:
> > Here's the series re-based on top of 5.8
> > https://github.com/allenpais/tasklets/tree/V3
> 
> Great!
> 
> > Let me know how you would want these to be reviewed.
> 
> Was a Coccinelle script used for any of these conversions? I wonder if
> it'd be easier to do a single treewide patch for the more mechanical
> changes.
> 
> And, actually, I still think the "prepare" patches should just be
> collapsed into the actual "covert" patches -- there are only a few.
> 
> After those, yeah, I think getting these sent to their respective
> maintainers is the next step.
> 
> > Also, I was thinking if removing tasklets completely could be a task
> > on KSPP wiki. If yes, I did like to take ownership of that task. I have a
> > couple of ideas in mind, which could be discussed in a separate email.
> 
> Sure! I will add it to the tracker. Here's for the refactoring:
> https://github.com/KSPP/linux/issues/30
> 
> and here's for the removal:
> https://github.com/KSPP/linux/issues/94
> 
> if you can added details/examples of how they should be removed, that'd
> help other folks too, if they wanted to jump in. :)

I have a patch set to convert the remaining tasklet usage in sound
drivers to either the threaded IRQ or the work, but it wasn't
submitted / merged for 5.8 due to the obvious conflict with your API
changes.
Each conversion is rather simple, but it's always a question of the
nature of each tasklet usage which alternative is the best fit.

FWIW, the current version is found in test/kill-tasklet branch of
sound git tree
  git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound.git


thanks,

Takashi

Re: [PATCH 0/3] Modernize tasklet callback API

[Allen]

>
> I have a patch set to convert the remaining tasklet usage in sound
> drivers to either the threaded IRQ or the work, but it wasn't
> submitted / merged for 5.8 due to the obvious conflict with your API
> changes.
> Each conversion is rather simple, but it's always a question of the
> nature of each tasklet usage which alternative is the best fit.
>
> FWIW, the current version is found in test/kill-tasklet branch of
> sound git tree
>   git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound.git

Great. Currently my tree has these converted to use the new
tasklet_setup() api. I will add these to my threaded IRQ/work tree
(which is still wip).

Thanks.


-- 
       - Allen

Re: [PATCH 0/3] Modernize tasklet callback API

[Allen]

Kees,

> Was a Coccinelle script used for any of these conversions? I wonder if
> it'd be easier to do a single treewide patch for the more mechanical
> changes.

No, I should have written one. Will do it.

> And, actually, I still think the "prepare" patches should just be
> collapsed into the actual "covert" patches -- there are only a few.

Okay. It's been done and pushed to:
https://github.com/allenpais/tasklets/tree/V4

> After those, yeah, I think getting these sent to their respective
> maintainers is the next step.

 Please look at the above branch, if it looks fine, let me know
if I can add your ACK on the patches.
>
> Sure! I will add it to the tracker. Here's for the refactoring:
> https://github.com/KSPP/linux/issues/30
>
> and here's for the removal:
> https://github.com/KSPP/linux/issues/94
>
> if you can added details/examples of how they should be removed, that'd
> help other folks too, if they wanted to jump in. :)

Sure, Thank you.

- Allen