by Leon Romanovsky

[permalink] [raw]

Subject: Re: [GIT] Networking

On Mon, Feb 24, 2020 at 01:47:32PM +0100, Greg KH wrote:
> On Mon, Feb 24, 2020 at 11:01:09AM +0100, Geert Uytterhoeven wrote:
> > Hi Linus,
> >
> > On Thu, Jun 25, 2015 at 1:38 AM Linus Torvalds
> > <[email protected]> wrote:
> > > On Wed, Jun 24, 2015 at 6:39 AM, David Miller <[email protected]> wrote:
> > > > git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git master
> > >
> > > On the *other* side of the same conflict, I find an even more
> > > offensive commit, namely commit 4cd7c9479aff ("IB/mad: Add support for
> > > additional MAD info to/from drivers") which adds a BUG_ON() for a
> > > sanity check, rather than just returning -EINVAL or something sane
> > > like that.
> > >
> > > I'm getting *real* tired of that BUG_ON() shit. I realize that
> > > infiniband is a niche market, and those "commercial grade" niche
> > > markets are more-than-used-to crap code and horrible hacks, but this
> > > is still the kernel. We don't add random machine-killing debug checks
> > > when it is *so* simple to just do
> > >
> > > if (WARN_ON_ONCE(..))
> > > return -EINVAL;
> > >
> > > instead.
> >
> > And if we follow that advice, friendly Greg will respond with:
> > "We really do not want WARN_ON() anywhere, as that causes systems with
> > panic-on-warn to reboot."
> > https://lore.kernel.org/lkml/[email protected]/
>
> Yes, we should not have any WARN_ON calls for something that userspace
> can trigger, because then syzbot will trigger it and we will get an
> annoying report saying to fix it :)

Impressive backlog :)
Geert, you replied on original discussion from 2015.

Thanks

>
> thanks,
>
> greg k-h

2020-02-24 16:37:06

by Doug Ledford

[permalink] [raw]

Subject: Re: [GIT] Networking

> On Feb 24, 2020, at 11:33 AM, Leon Romanovsky <[email protected]> wrote:
>
> On Mon, Feb 24, 2020 at 01:47:32PM +0100, Greg KH wrote:
>> On Mon, Feb 24, 2020 at 11:01:09AM +0100, Geert Uytterhoeven wrote:
>>> Hi Linus,
>>>
>>> On Thu, Jun 25, 2015 at 1:38 AM Linus Torvalds
>>> <[email protected]> wrote:
>>>> On Wed, Jun 24, 2015 at 6:39 AM, David Miller <[email protected]> wrote:
>>>>> git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git master
>>>>
>>>> On the *other* side of the same conflict, I find an even more
>>>> offensive commit, namely commit 4cd7c9479aff ("IB/mad: Add support for
>>>> additional MAD info to/from drivers") which adds a BUG_ON() for a
>>>> sanity check, rather than just returning -EINVAL or something sane
>>>> like that.
>>>>
>>>> I'm getting *real* tired of that BUG_ON() shit. I realize that
>>>> infiniband is a niche market, and those "commercial grade" niche
>>>> markets are more-than-used-to crap code and horrible hacks, but this
>>>> is still the kernel. We don't add random machine-killing debug checks
>>>> when it is *so* simple to just do
>>>>
>>>> if (WARN_ON_ONCE(..))
>>>> return -EINVAL;
>>>>
>>>> instead.
>>>
>>> And if we follow that advice, friendly Greg will respond with:
>>> "We really do not want WARN_ON() anywhere, as that causes systems with
>>> panic-on-warn to reboot."
>>> https://lore.kernel.org/lkml/[email protected]/
>>
>> Yes, we should not have any WARN_ON calls for something that userspace
>> can trigger, because then syzbot will trigger it and we will get an
>> annoying report saying to fix it :)
>
> Impressive backlog :)
> Geert, you replied on original discussion from 2015.

Yeah, that threw me for a loop too ;-). Took several double takes on that one just to make sure none of the IB comments from Linus were related to anything current!

--
Doug Ledford <[email protected]>
GPG KeyID: B826A3330E572FDD
Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD

Attachments:

signature.asc (849.00 B)
Message signed with OpenPGP

2020-02-24 16:54:12

by Geert Uytterhoeven

[permalink] [raw]

Subject: Re: [GIT] Networking

Hi Doug et al,

On Mon, Feb 24, 2020 at 5:36 PM Doug Ledford <[email protected]> wrote:
> > On Feb 24, 2020, at 11:33 AM, Leon Romanovsky <[email protected]> wrote:
> > On Mon, Feb 24, 2020 at 01:47:32PM +0100, Greg KH wrote:
> >> On Mon, Feb 24, 2020 at 11:01:09AM +0100, Geert Uytterhoeven wrote:
> >>> On Thu, Jun 25, 2015 at 1:38 AM Linus Torvalds
> >>> <[email protected]> wrote:
> >>>> On Wed, Jun 24, 2015 at 6:39 AM, David Miller <[email protected]> wrote:
> >>>>> git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git master
> >>>>
> >>>> On the *other* side of the same conflict, I find an even more
> >>>> offensive commit, namely commit 4cd7c9479aff ("IB/mad: Add support for
> >>>> additional MAD info to/from drivers") which adds a BUG_ON() for a
> >>>> sanity check, rather than just returning -EINVAL or something sane
> >>>> like that.
> >>>>
> >>>> I'm getting *real* tired of that BUG_ON() shit. I realize that
> >>>> infiniband is a niche market, and those "commercial grade" niche
> >>>> markets are more-than-used-to crap code and horrible hacks, but this
> >>>> is still the kernel. We don't add random machine-killing debug checks
> >>>> when it is *so* simple to just do
> >>>>
> >>>> if (WARN_ON_ONCE(..))
> >>>> return -EINVAL;
> >>>>
> >>>> instead.
> >>>
> >>> And if we follow that advice, friendly Greg will respond with:
> >>> "We really do not want WARN_ON() anywhere, as that causes systems with
> >>> panic-on-warn to reboot."
> >>> https://lore.kernel.org/lkml/[email protected]/
> >>
> >> Yes, we should not have any WARN_ON calls for something that userspace
> >> can trigger, because then syzbot will trigger it and we will get an
> >> annoying report saying to fix it :)
> >
> > Impressive backlog :)
> > Geert, you replied on original discussion from 2015.

Oops.
I was looking up a recent net commit that was part of Dave's last pull
request, couldn't find what I was looking for, and must have suddenly
ended up in the email for an old pull request instead (they're all called
"[GIT] Networking")...

> Yeah, that threw me for a loop too ;-). Took several double takes on that one just to make sure none of the IB comments from Linus were related to anything current!

Sorry for that. I hope I didn't cause any lost heartbeats.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds