Re: Build regressions/improvements in v6.4-rc1

Re: Build regressions/improvements in v6.4-rc1

[Geert Uytterhoeven]

On Mon, 8 May 2023, Geert Uytterhoeven wrote:
> Below is the list of build error/warning regressions/improvements in
> v6.4-rc1[1] compared to v6.3[2].
>
> Summarized:
>  - build errors: +9/-16
>  - build warnings: +1/-1439
>
> Happy fixing! ;-)
>
> Thanks to the linux-next team for providing the build service.
>
> [1] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/ac9a78681b921877518763ba0e89202254349d1b/ (all 152 configs)
> [2] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/457391b0380335d5e9a5babdec90ac53928b23b4/ (all 152 configs)
>
>
> *** ERRORS ***
>
> 9 error regressions:
>  + /kisskb/src/drivers/media/platform/nxp/imx8-isi/imx8-isi-core.c: error: 'mxc_isi_pm_resume' defined but not used [-Werror=unused-function]:  => 328:12
>  + /kisskb/src/drivers/media/platform/nxp/imx8-isi/imx8-isi-core.c: error: 'mxc_isi_pm_suspend' defined but not used [-Werror=unused-function]:  => 314:12

sparc64-gcc{5,11}/sparc-allmodconfig
(fix was available before the merge window)

>  + /kisskb/src/drivers/media/platform/nxp/imx8-isi/imx8-isi-hw.c: error: "CONFIG_ARCH_DMA_ADDR_T_64BIT" is not defined [-Werror=undef]:  => 66:5, 33:5, 51:5
>  + /kisskb/src/drivers/media/platform/nxp/imx8-isi/imx8-isi-hw.c: error: "CONFIG_ARCH_DMA_ADDR_T_64BIT" is not defined, evaluates to 0 [-Werror=undef]:  => 51:5, 66:5, 33:5

mips-gcc{8,11}/mips-allmodconfig
xtensa-gcc11/xtensa-allmodconfig
powerpc-gcc5/ppc32_allmodconfig
(fix was available before the merge window)

>  + /kisskb/src/drivers/mtd/spi-nor/spansion.c: error: 'op' is used uninitialized [-Werror=uninitialized]:  => 495:27, 364:27

um-x86_64-gcc12/um-allyesconfig

>  + /kisskb/src/fs/xfs/scrub/scrub.h: error: initializer element is not constant:  => 112:28

powerpc-gcc5/ppc64_book3e_allmodconfig
powerpc-gcc5/powerpc-allmodconfig
powerpc-gcc5/ppc32_allmodconfig
powerpc-gcc5/powerpc-allyesconfig
powerpc-gcc5/ppc64le_allmodconfig
arm64-gcc5/arm64-allmodconfig
(fix sent)

>  + error: modpost: "__floatunsidf" [drivers/phy/mediatek/phy-mtk-hdmi-drv.ko] undefined!:  => N/A
>  + error: modpost: "__gedf2" [drivers/phy/mediatek/phy-mtk-hdmi-drv.ko] undefined!:  => N/A
>  + error: modpost: "__ltdf2" [drivers/phy/mediatek/phy-mtk-hdmi-drv.ko] undefined!:  => N/A

sparc64-gcc5/sparc64-allmodconfig

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 v6.4-rc1

[Pratyush Yadav]

On Mon, May 08 2023, Geert Uytterhoeven wrote:

> On Mon, 8 May 2023, Geert Uytterhoeven wrote:
>> Below is the list of build error/warning regressions/improvements in
>> v6.4-rc1[1] compared to v6.3[2].
>>
>> Summarized:
>>  - build errors: +9/-16
>>  - build warnings: +1/-1439
>>
>> Happy fixing! ;-)
>>
>> Thanks to the linux-next team for providing the build service.
>>
>> [1] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/ac9a78681b921877518763ba0e89202254349d1b/ (all 152 configs)
>> [2] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/457391b0380335d5e9a5babdec90ac53928b23b4/ (all 152 configs)
>>
>>
[...]
>
>>  + /kisskb/src/drivers/mtd/spi-nor/spansion.c: error: 'op' is used uninitialized [-Werror=uninitialized]:  => 495:27, 364:27

Hmm, I don't get why we get this warning. Line 495 is in
s25fs256t_post_bfpt_fixup(). It declares 'op' and then it does

    op = (struct spi_mem_op)
        CYPRESS_NOR_RD_ANY_REG_OP(nor->params->addr_mode_nbytes,
                      SPINOR_REG_CYPRESS_ARCFN, 1,
                      nor->bouncebuf);
    ret = spi_nor_read_any_reg(nor, &op, nor->reg_proto);


which initializes 'op' before using it. Same with line 364 which is in
the function cypress_nor_set_addr_mode_nbytes().

Even the compiler warnings [0] don't seem to make much sense to me:

    /kisskb/src/drivers/mtd/spi-nor/spansion.c: In function 's25fs256t_post_bfpt_fixup':
    /kisskb/src/drivers/mtd/spi-nor/spansion.c:495:27: error: 'op' is used uninitialized [-Werror=uninitialized]
    495 |         struct spi_mem_op op;
        |                           ^~
    /kisskb/src/drivers/mtd/spi-nor/spansion.c:495:27: note: 'op' declared here
    495 |         struct spi_mem_op op;
        |                           ^~

[0] http://kisskb.ellerman.id.au/kisskb/buildresult/14922057/

>
> um-x86_64-gcc12/um-allyesconfig
>
[...]

-- 
Regards,
Pratyush Yadav

Re: Build regressions/improvements in v6.4-rc1

[Geert Uytterhoeven]

Hi Pratyush,

On Mon, May 8, 2023 at 5:19 PM Pratyush Yadav <pratyush@kernel.org> wrote:
> On Mon, May 08 2023, Geert Uytterhoeven wrote:
> > On Mon, 8 May 2023, Geert Uytterhoeven wrote:
> >> Below is the list of build error/warning regressions/improvements in
> >> v6.4-rc1[1] compared to v6.3[2].
> >>
> >> Summarized:
> >>  - build errors: +9/-16
> >>  - build warnings: +1/-1439
> >>
> >> Happy fixing! ;-)
> >>
> >> Thanks to the linux-next team for providing the build service.
> >>
> >> [1] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/ac9a78681b921877518763ba0e89202254349d1b/ (all 152 configs)
> >> [2] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/457391b0380335d5e9a5babdec90ac53928b23b4/ (all 152 configs)
> >>
> >>
> [...]
> >
> >>  + /kisskb/src/drivers/mtd/spi-nor/spansion.c: error: 'op' is used uninitialized [-Werror=uninitialized]:  => 495:27, 364:27
>
> Hmm, I don't get why we get this warning. Line 495 is in
> s25fs256t_post_bfpt_fixup(). It declares 'op' and then it does
>
>     op = (struct spi_mem_op)
>         CYPRESS_NOR_RD_ANY_REG_OP(nor->params->addr_mode_nbytes,
>                       SPINOR_REG_CYPRESS_ARCFN, 1,
>                       nor->bouncebuf);
>     ret = spi_nor_read_any_reg(nor, &op, nor->reg_proto);
>
>
> which initializes 'op' before using it. Same with line 364 which is in
> the function cypress_nor_set_addr_mode_nbytes().

I suspect gcc-12 became stricter, and not all members of the union
are initialized.

> Even the compiler warnings [0] don't seem to make much sense to me:
>
>     /kisskb/src/drivers/mtd/spi-nor/spansion.c: In function 's25fs256t_post_bfpt_fixup':
>     /kisskb/src/drivers/mtd/spi-nor/spansion.c:495:27: error: 'op' is used uninitialized [-Werror=uninitialized]
>     495 |         struct spi_mem_op op;
>         |                           ^~
>     /kisskb/src/drivers/mtd/spi-nor/spansion.c:495:27: note: 'op' declared here
>     495 |         struct spi_mem_op op;
>         |                           ^~
>
> [0] http://kisskb.ellerman.id.au/kisskb/buildresult/14922057/

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 v6.4-rc1

[Raghavendra, Vignesh]

On 5/8/2023 8:49 PM, Pratyush Yadav wrote:
> On Mon, May 08 2023, Geert Uytterhoeven wrote:
> 
>> On Mon, 8 May 2023, Geert Uytterhoeven wrote:
>>> Below is the list of build error/warning regressions/improvements in
>>> v6.4-rc1[1] compared to v6.3[2].
>>>
>>> Summarized:
>>>  - build errors: +9/-16
>>>  - build warnings: +1/-1439
>>>
>>> Happy fixing! ;-)
>>>
>>> Thanks to the linux-next team for providing the build service.
>>>
>>> [1] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/ac9a78681b921877518763ba0e89202254349d1b/ (all 152 configs)
>>> [2] http://kisskb.ellerman.id.au/kisskb/branch/linus/head/457391b0380335d5e9a5babdec90ac53928b23b4/ (all 152 configs)
>>>
>>>
> [...]
>>
>>>  + /kisskb/src/drivers/mtd/spi-nor/spansion.c: error: 'op' is used uninitialized [-Werror=uninitialized]:  => 495:27, 364:27
> 
> Hmm, I don't get why we get this warning. Line 495 is in
> s25fs256t_post_bfpt_fixup(). It declares 'op' and then it does
> 
>     op = (struct spi_mem_op)
>         CYPRESS_NOR_RD_ANY_REG_OP(nor->params->addr_mode_nbytes,
>                       SPINOR_REG_CYPRESS_ARCFN, 1,
>                       nor->bouncebuf);
>     ret = spi_nor_read_any_reg(nor, &op, nor->reg_proto);
> 

Not all the fields of struct get initialized if you assign later on in
the function vs at the time of declaration... For example cmd.dtr isn't
explicitly set to any value here and may have garbage?
> 
> which initializes 'op' before using it. Same with line 364 which is in
> the function cypress_nor_set_addr_mode_nbytes().
> 
> Even the compiler warnings [0] don't seem to make much sense to me:
> 
>     /kisskb/src/drivers/mtd/spi-nor/spansion.c: In function 's25fs256t_post_bfpt_fixup':
>     /kisskb/src/drivers/mtd/spi-nor/spansion.c:495:27: error: 'op' is used uninitialized [-Werror=uninitialized]
>     495 |         struct spi_mem_op op;
>         |                           ^~
>     /kisskb/src/drivers/mtd/spi-nor/spansion.c:495:27: note: 'op' declared here
>     495 |         struct spi_mem_op op;
>         |                           ^~
> 
> [0] http://kisskb.ellerman.id.au/kisskb/buildresult/14922057/
> 
>>
>> um-x86_64-gcc12/um-allyesconfig
>>
> [...]
>