Re: Build regressions/improvements in v4.4-rc7

Re: Build regressions/improvements in v4.4-rc7

[Geert Uytterhoeven]

On Mon, Dec 28, 2015 at 1:21 PM, Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
> JFYI, when comparing v4.4-rc7[1] to v4.4-rc6[3], the summaries are:
>   - build errors: +14/-3

  + /home/kisskb/slave/src/include/linux/kqueue.h: error:
dereferencing type-punned pointer will break strict-aliasing rules
[-Werror=strict-aliasing]:  => 186:2

mips-allmodconfig

Reported in September by kbuild test robot <lkp@intel.com>

  + /tmp/cc5veAi5.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 378, 49

mips-allnoconfig

  + /tmp/cc6DopGj.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 43
  + /tmp/cc7GEfIa.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 43
  + /tmp/ccmqbWag.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 43

bigsur_defconfig

  + /tmp/ccGsoCyO.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 49, 366

ip22_defconfig

  + /tmp/ccH1dW0f.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 43
  + /tmp/ccP7l3ug.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 43
  + /tmp/ccSlbKej.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 43

ip27_defconfig

  + /tmp/ccemCHRe.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 50, 326
  + /tmp/cckgYy8k.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 329, 50
  + /tmp/cclNaAGh.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 366, 49

ip32_defconfig

  + /tmp/cczzZNSV.s: Error: can't resolve `_start' {*UND* section} -
`L0^A' {.text section}:  => 403, 41

malta_defconfig

I guess the last-minute build fix for MIPS broke the build for other versions of
binutils?

> [1] http://kisskb.ellerman.id.au/kisskb/head/9745/ (258 out of 259 configs)
> [3] http://kisskb.ellerman.id.au/kisskb/head/9734/ (all 259 configs)

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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

Re: Build regressions/improvements in v4.4-rc7

[Tejun Heo]

On Mon, Dec 28, 2015 at 01:29:18PM +0100, Geert Uytterhoeven wrote:
> On Mon, Dec 28, 2015 at 1:21 PM, Geert Uytterhoeven
> <geert@linux-m68k.org> wrote:
> > JFYI, when comparing v4.4-rc7[1] to v4.4-rc6[3], the summaries are:
> >   - build errors: +14/-3
> 
>   + /home/kisskb/slave/src/include/linux/kqueue.h: error:
> dereferencing type-punned pointer will break strict-aliasing rules
> [-Werror=strict-aliasing]:  => 186:2

kqueue.h?

-- 
tejun

Re: Build regressions/improvements in v4.4-rc7

[Randy Dunlap]

On 12/28/15 08:18, Tejun Heo wrote:
> On Mon, Dec 28, 2015 at 01:29:18PM +0100, Geert Uytterhoeven wrote:
>> On Mon, Dec 28, 2015 at 1:21 PM, Geert Uytterhoeven
>> <geert@linux-m68k.org> wrote:
>>> JFYI, when comparing v4.4-rc7[1] to v4.4-rc6[3], the summaries are:
>>>   - build errors: +14/-3
>>
>>   + /home/kisskb/slave/src/include/linux/kqueue.h: error:
>> dereferencing type-punned pointer will break strict-aliasing rules
>> [-Werror=strict-aliasing]:  => 186:2
> 
> kqueue.h?
> 

  + /home/kisskb/slave/src/include/linux/workqueue.h: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]:  => 186:2



-- 
~Randy

Re: Build regressions/improvements in v4.4-rc7

[Tejun Heo]

Hello,

On Mon, Dec 28, 2015 at 08:22:20AM -0800, Randy Dunlap wrote:
> On 12/28/15 08:18, Tejun Heo wrote:
> > On Mon, Dec 28, 2015 at 01:29:18PM +0100, Geert Uytterhoeven wrote:
> >> On Mon, Dec 28, 2015 at 1:21 PM, Geert Uytterhoeven
> >> <geert@linux-m68k.org> wrote:
> >>> JFYI, when comparing v4.4-rc7[1] to v4.4-rc6[3], the summaries are:
> >>>   - build errors: +14/-3
> >>
> >>   + /home/kisskb/slave/src/include/linux/kqueue.h: error:
> >> dereferencing type-punned pointer will break strict-aliasing rules
> >> [-Werror=strict-aliasing]:  => 186:2
> > 
> > kqueue.h?
> > 
> 
>   + /home/kisskb/slave/src/include/linux/workqueue.h: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]:  => 186:2

Heh, yeah, that makes a lot more sense.  At the same time tho, the
code has been there forever and there are a lot of simliar derefs
throughout kernel/workqueue.c.  Is there something the code should be
doing?

Thanks.

-- 
tejun