Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

[Arnd Bergmann]

On Wed, Jul 19, 2017 at 11:50 PM,  <akpm@linux-foundation.org> wrote:
> ------------------------------------------------------
> From: Arnd Bergmann <arnd@arndb.de>
> Subject: kbuild: disable -Wformat-truncation warnings by default
>
> With x86 allmodconfig, we currently get 233 -Wformat-truncation warnings,
> which makes the entire warnings rather useless.
>
> This turns off the warning by default, unless we specify W=1 or higher
>
> Link: http://lkml.kernel.org/r/20170714120720.906842-2-arnd@arndb.de
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
> ---
>
>  scripts/Makefile.extrawarn |    3 +++
>  1 file changed, 3 insertions(+)
>
> diff -puN scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default scripts/Makefile.extrawarn
> --- a/scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default
> +++ a/scripts/Makefile.extrawarn
> @@ -67,5 +67,8 @@ KBUILD_CFLAGS += $(call cc-disable-warni
>  KBUILD_CFLAGS += $(call cc-disable-warning, sign-compare)
>  KBUILD_CFLAGS += $(call cc-disable-warning, format-zero-length)
>  KBUILD_CFLAGS += $(call cc-disable-warning, uninitialized)
> +else
> +# noisy gcc-7 warnings
> +KBUILD_CFLAGS += $(call cc-option,-Wformat-truncation=0)
>  endif
>  endif

Hi Andrew, Linus, Greg,

I noticed that Linus has made a similar patch in 4.13-rc1, commit bd664f6b3e37
("disable new gcc-7.1.1 warnings for now"), and it completely disables three
warnings (format-truncation, format-overflow and int-in-bool-context).

Obviously there is no point in having both, so let's talk about what we
want for 4.13, stable-backports and for future kernels, I'll then send those
patches. Here is my first suggestion:

- enable all three warnings with "make W=1" in 4.13, but leave them
  disabled by default.
- backport Linus' patch, plus the follow-up for W=1 to stable kernels,
  to allow stable kernels to build cleanly
- backport the patches that address any other gcc-7 warnings, as
  well as those that are not obvious false-positives to stable kernels
- In 4.14+, use my version above and address all int-in-bool-context
  and format-overflow warnings, but only use -Wformat-truncation
  with make W=1.

Any other suggestions?

        Arnd

Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

[gregkh]

On Thu, Jul 20, 2017 at 09:24:21AM +0200, Arnd Bergmann wrote:
> On Wed, Jul 19, 2017 at 11:50 PM,  <akpm@linux-foundation.org> wrote:
> > ------------------------------------------------------
> > From: Arnd Bergmann <arnd@arndb.de>
> > Subject: kbuild: disable -Wformat-truncation warnings by default
> >
> > With x86 allmodconfig, we currently get 233 -Wformat-truncation warnings,
> > which makes the entire warnings rather useless.
> >
> > This turns off the warning by default, unless we specify W=1 or higher
> >
> > Link: http://lkml.kernel.org/r/20170714120720.906842-2-arnd@arndb.de
> > Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> > Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
> > ---
> >
> >  scripts/Makefile.extrawarn |    3 +++
> >  1 file changed, 3 insertions(+)
> >
> > diff -puN scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default scripts/Makefile.extrawarn
> > --- a/scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default
> > +++ a/scripts/Makefile.extrawarn
> > @@ -67,5 +67,8 @@ KBUILD_CFLAGS += $(call cc-disable-warni
> >  KBUILD_CFLAGS += $(call cc-disable-warning, sign-compare)
> >  KBUILD_CFLAGS += $(call cc-disable-warning, format-zero-length)
> >  KBUILD_CFLAGS += $(call cc-disable-warning, uninitialized)
> > +else
> > +# noisy gcc-7 warnings
> > +KBUILD_CFLAGS += $(call cc-option,-Wformat-truncation=0)
> >  endif
> >  endif
> 
> Hi Andrew, Linus, Greg,
> 
> I noticed that Linus has made a similar patch in 4.13-rc1, commit bd664f6b3e37
> ("disable new gcc-7.1.1 warnings for now"), and it completely disables three
> warnings (format-truncation, format-overflow and int-in-bool-context).
> 
> Obviously there is no point in having both, so let's talk about what we
> want for 4.13, stable-backports and for future kernels, I'll then send those
> patches. Here is my first suggestion:
> 
> - enable all three warnings with "make W=1" in 4.13, but leave them
>   disabled by default.
> - backport Linus' patch, plus the follow-up for W=1 to stable kernels,
>   to allow stable kernels to build cleanly

I don't care which of these, I'll take whatever is in Linus's tree.
And I'll go backport that patch now as I'm getting annoyed by the
warnings at the moment...

> - backport the patches that address any other gcc-7 warnings, as
>   well as those that are not obvious false-positives to stable kernels

I'll do that as well, as I notice them go by.

> - In 4.14+, use my version above and address all int-in-bool-context
>   and format-overflow warnings, but only use -Wformat-truncation
>   with make W=1.

I don't really care about this, as long as we are not forced to do
"silly things" like some of the patches seemed to do :)

thanks,

greg k-h

Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

[Arnd Bergmann]

On Thu, Jul 20, 2017 at 10:17 AM, gregkh <gregkh@linuxfoundation.org> wrote:
> On Thu, Jul 20, 2017 at 09:24:21AM +0200, Arnd Bergmann wrote:
>> On Wed, Jul 19, 2017 at 11:50 PM,  <akpm@linux-foundation.org> wrote:
>> > ------------------------------------------------------
>> > From: Arnd Bergmann <arnd@arndb.de>
>> > Subject: kbuild: disable -Wformat-truncation warnings by default
>> >
>> > With x86 allmodconfig, we currently get 233 -Wformat-truncation warnings,
>> > which makes the entire warnings rather useless.
>> >
>> > This turns off the warning by default, unless we specify W=1 or higher
>> >
>> > Link: http://lkml.kernel.org/r/20170714120720.906842-2-arnd@arndb.de
>> > Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>> > Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
>> > ---
>> >
>> >  scripts/Makefile.extrawarn |    3 +++
>> >  1 file changed, 3 insertions(+)
>> >
>> > diff -puN scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default scripts/Makefile.extrawarn
>> > --- a/scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default
>> > +++ a/scripts/Makefile.extrawarn
>> > @@ -67,5 +67,8 @@ KBUILD_CFLAGS += $(call cc-disable-warni
>> >  KBUILD_CFLAGS += $(call cc-disable-warning, sign-compare)
>> >  KBUILD_CFLAGS += $(call cc-disable-warning, format-zero-length)
>> >  KBUILD_CFLAGS += $(call cc-disable-warning, uninitialized)
>> > +else
>> > +# noisy gcc-7 warnings
>> > +KBUILD_CFLAGS += $(call cc-option,-Wformat-truncation=0)
>> >  endif
>> >  endif
>>
>> Hi Andrew, Linus, Greg,
>>
>> I noticed that Linus has made a similar patch in 4.13-rc1, commit bd664f6b3e37
>> ("disable new gcc-7.1.1 warnings for now"), and it completely disables three
>> warnings (format-truncation, format-overflow and int-in-bool-context).
>>
>> Obviously there is no point in having both, so let's talk about what we
>> want for 4.13, stable-backports and for future kernels, I'll then send those
>> patches. Here is my first suggestion:
>>
>> - enable all three warnings with "make W=1" in 4.13, but leave them
>>   disabled by default.
>> - backport Linus' patch, plus the follow-up for W=1 to stable kernels,
>>   to allow stable kernels to build cleanly
>
> I don't care which of these, I'll take whatever is in Linus's tree.
>
> And I'll go backport that patch now as I'm getting annoyed by the
> warnings at the moment...

Yes, that's what I meant here, the follow-up obviously has to
be agreed on before that can be backported. Let me know what other
warnings you see on the stable kernels after backporting the
bd664f6b3e37 patch, I can help identify the fixes that went into
mainline to address those if you want.

>> - backport the patches that address any other gcc-7 warnings, as
>>   well as those that are not obvious false-positives to stable kernels
>
> I'll do that as well, as I notice them go by.
>
>> - In 4.14+, use my version above and address all int-in-bool-context
>>   and format-overflow warnings, but only use -Wformat-truncation
>>   with make W=1.
>
> I don't really care about this, as long as we are not forced to do
> "silly things" like some of the patches seemed to do :)

Sure, I understand that from the previous feedback, and for a couple
of patches I found better workarounds (that improve the code
while fixing the warning, rather than adding hacks). I also
reported two of them in gcc bugzilla already, one was decided
to be invalid:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81484

the other one looks valid, but possibly hard to fix:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81483

When I go through the remaining patches again, I'll have a
look at what others might be incorrect warnings that should
be fixed in gcc, or which ones can be fixed in a better way.
I just wanted to first make sure that there is no fundamental
objections to enabling the warnings in future kernels (at
whichever W= level) at all.

       Arnd

Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

[Masahiro Yamada]

Hi Arnd,


2017-07-20 16:24 GMT+09:00 Arnd Bergmann <arnd@arndb.de>:
> On Wed, Jul 19, 2017 at 11:50 PM,  <akpm@linux-foundation.org> wrote:
>> ------------------------------------------------------
>> From: Arnd Bergmann <arnd@arndb.de>
>> Subject: kbuild: disable -Wformat-truncation warnings by default
>>
>> With x86 allmodconfig, we currently get 233 -Wformat-truncation warnings,
>> which makes the entire warnings rather useless.
>>
>> This turns off the warning by default, unless we specify W=1 or higher
>>
>> Link: http://lkml.kernel.org/r/20170714120720.906842-2-arnd@arndb.de
>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
>> ---
>>
>>  scripts/Makefile.extrawarn |    3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff -puN scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default scripts/Makefile.extrawarn
>> --- a/scripts/Makefile.extrawarn~kbuild-disable-wformat-truncation-warnings-by-default
>> +++ a/scripts/Makefile.extrawarn
>> @@ -67,5 +67,8 @@ KBUILD_CFLAGS += $(call cc-disable-warni
>>  KBUILD_CFLAGS += $(call cc-disable-warning, sign-compare)
>>  KBUILD_CFLAGS += $(call cc-disable-warning, format-zero-length)
>>  KBUILD_CFLAGS += $(call cc-disable-warning, uninitialized)
>> +else
>> +# noisy gcc-7 warnings
>> +KBUILD_CFLAGS += $(call cc-option,-Wformat-truncation=0)
>>  endif
>>  endif
>
> Hi Andrew, Linus, Greg,
>
> I noticed that Linus has made a similar patch in 4.13-rc1, commit bd664f6b3e37
> ("disable new gcc-7.1.1 warnings for now"), and it completely disables three
> warnings (format-truncation, format-overflow and int-in-bool-context).
>
> Obviously there is no point in having both, so let's talk about what we
> want for 4.13, stable-backports and for future kernels, I'll then send those
> patches. Here is my first suggestion:
>
> - enable all three warnings with "make W=1" in 4.13, but leave them
>   disabled by default.
> - backport Linus' patch, plus the follow-up for W=1 to stable kernels,
>   to allow stable kernels to build cleanly
> - backport the patches that address any other gcc-7 warnings, as
>   well as those that are not obvious false-positives to stable kernels
> - In 4.14+, use my version above and address all int-in-bool-context
>   and format-overflow warnings, but only use -Wformat-truncation
>   with make W=1.
>

Talking about 4.14+, shall we move -Wformat-truncation
from the top Makefile (always disable) to
Makefile.extrawarn (enable with W=1) ?


-- 
Best Regards
Masahiro Yamada

Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

[Arnd Bergmann]

On Sun, Aug 20, 2017 at 3:19 PM, Masahiro Yamada
<yamada.masahiro@socionext.com> wrote:
> 2017-07-20 16:24 GMT+09:00 Arnd Bergmann <arnd@arndb.de>:
>> - enable all three warnings with "make W=1" in 4.13, but leave them
>>   disabled by default.
>> - backport Linus' patch, plus the follow-up for W=1 to stable kernels,
>>   to allow stable kernels to build cleanly
>> - backport the patches that address any other gcc-7 warnings, as
>>   well as those that are not obvious false-positives to stable kernels
>> - In 4.14+, use my version above and address all int-in-bool-context
>>   and format-overflow warnings, but only use -Wformat-truncation
>>   with make W=1.
>>
>
> Talking about 4.14+, shall we move -Wformat-truncation
> from the top Makefile (always disable) to
> Makefile.extrawarn (enable with W=1) ?

I dropped the ball on this one, sorry. I think we should do this for
all three warnings (format-overflow, format-truncation and
int-in-bool-context) for the time being.

In case of format-truncation, there are countless warnings,
most of them false-postives, so we simply can't enable them
by default.

For -Wformat-overflow, there is one patch that I need to
rewrite, all my other patches are pending for 4.14, see
https://patchwork.kernel.org/patch/9840801/ for the missing
one. This should be trivial to fix. However, enabling
CONFIG_UBSAN_SANITIZE_ALL results in seven additional
false positives. I created an patch for this in
https://pastebin.com/CD7nhRNp but can't submit that as it's
obviously bogus. I reported the gcc bug as
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81592

What we could do there is to disable the warning if
CONFIG_UBSAN_SANITIZE_ALL is turned on (like
we do for -Wmaybe-uninitialized in
CONFIG_UBSAN_SANITIZE_ALL) but leave it on otherwise.

I submitted patches for all -Wint-in-bool-context in arm/arm64/x86
randconfig builds, but there are still six known warnings for which
my patches did not get queued for 4.14.
I have to revisit those all to decide whether we can find an
acceptable workaround in the kernel and enable the warning again
by default, or leave it in W=1 until gcc improves enough.

      Arnd

Re: + kbuild-disable-wformat-truncation-warnings-by-default.patch added to -mm tree

[Masahiro Yamada]

Hi Arnd,

2017-08-21 4:51 GMT+09:00 Arnd Bergmann <arnd@arndb.de>:
> On Sun, Aug 20, 2017 at 3:19 PM, Masahiro Yamada
> <yamada.masahiro@socionext.com> wrote:
>> 2017-07-20 16:24 GMT+09:00 Arnd Bergmann <arnd@arndb.de>:
>>> - enable all three warnings with "make W=1" in 4.13, but leave them
>>>   disabled by default.
>>> - backport Linus' patch, plus the follow-up for W=1 to stable kernels,
>>>   to allow stable kernels to build cleanly
>>> - backport the patches that address any other gcc-7 warnings, as
>>>   well as those that are not obvious false-positives to stable kernels
>>> - In 4.14+, use my version above and address all int-in-bool-context
>>>   and format-overflow warnings, but only use -Wformat-truncation
>>>   with make W=1.
>>>
>>
>> Talking about 4.14+, shall we move -Wformat-truncation
>> from the top Makefile (always disable) to
>> Makefile.extrawarn (enable with W=1) ?
>
> I dropped the ball on this one, sorry. I think we should do this for
> all three warnings (format-overflow, format-truncation and
> int-in-bool-context) for the time being.
>
> In case of format-truncation, there are countless warnings,
> most of them false-postives, so we simply can't enable them
> by default.
>
> For -Wformat-overflow, there is one patch that I need to
> rewrite, all my other patches are pending for 4.14, see
> https://patchwork.kernel.org/patch/9840801/ for the missing
> one. This should be trivial to fix. However, enabling
> CONFIG_UBSAN_SANITIZE_ALL results in seven additional
> false positives. I created an patch for this in
> https://pastebin.com/CD7nhRNp but can't submit that as it's
> obviously bogus. I reported the gcc bug as
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81592
>
> What we could do there is to disable the warning if
> CONFIG_UBSAN_SANITIZE_ALL is turned on (like
> we do for -Wmaybe-uninitialized in
> CONFIG_UBSAN_SANITIZE_ALL) but leave it on otherwise.
>
> I submitted patches for all -Wint-in-bool-context in arm/arm64/x86
> randconfig builds, but there are still six known warnings for which
> my patches did not get queued for 4.14.
> I have to revisit those all to decide whether we can find an
> acceptable workaround in the kernel and enable the warning again
> by default, or leave it in W=1 until gcc improves enough.
>

I was just wondering how to handle your original patch.
I do not mean to press you.
We can take our time to make the right decision.  Thanks!



-- 
Best Regards
Masahiro Yamada