On Wed, Oct 9, 2019 at 3:25 PM Rong Chen <rong.a.chen(a)intel.com> wrote:
On 10/7/19 3:04 PM, Tzung-Bi Shih wrote:
> On Sun, Oct 6, 2019 at 12:54 AM kbuild test robot <lkp(a)intel.com> wrote:
>> url:
https://github.com/0day-ci/linux/commits/Tzung-Bi-Shih/ASoC-mediatek-mt81...
>> base:
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git
for-next
>> reproduce:
>> # apt-get install sparse
>> # sparse version: v0.6.1-rc1-42-g38eda53-dirty
>> make ARCH=x86_64 allmodconfig
>> make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
>>
>> If you fix the issue, kindly add following tag
>> Reported-by: kbuild test robot <lkp(a)intel.com>
>>
>>
>> sparse warnings: (new ones prefixed by >>)
>>
>>>> sound/soc/codecs/cros_ec_codec.c:430:39: sparse: sparse: incorrect type
in return expression (different address spaces) @@ expected void * @@ got void
[noderef] <asvoid * @@
>>>> sound/soc/codecs/cros_ec_codec.c:430:39: sparse: expected void *
>>>> sound/soc/codecs/cros_ec_codec.c:430:39: sparse: got void [noderef]
<asn:2> *
>>>> sound/soc/codecs/cros_ec_codec.c:549:69: sparse: sparse: incorrect type
in argument 2 (different address spaces) @@ expected void const volatile [noderef]
<asn:2> * @@ got latile [noderef] <asn:2> * @@
>>>> sound/soc/codecs/cros_ec_codec.c:549:69: sparse: expected void const
volatile [noderef] <asn:2> *
>>>> sound/soc/codecs/cros_ec_codec.c:549:69: sparse: got unsigned char
[usertype] *addr
>>>> sound/soc/codecs/cros_ec_codec.c:698:33: sparse: sparse: incorrect type
in argument 1 (different address spaces) @@ expected void volatile [noderef]
<asn:2> * @@ got latile [noderef] <asn:2> * @@
>>>> sound/soc/codecs/cros_ec_codec.c:698:33: sparse: expected void
volatile [noderef] <asn:2> *
>>>> sound/soc/codecs/cros_ec_codec.c:698:33: sparse: got unsigned char
[usertype] *wov_lang_shm_p
>> sound/soc/codecs/cros_ec_codec.c:699:48: sparse: sparse: incorrect type in
argument 1 (different address spaces) @@ expected void volatile [noderef] <asn:2>
* @@ got latile [noderef] <asn:2> * @@
>> sound/soc/codecs/cros_ec_codec.c:699:48: sparse: expected void volatile
[noderef] <asn:2> *
>>>> sound/soc/codecs/cros_ec_codec.c:699:48: sparse: got unsigned char
[usertype] *
> I cannot reproduce the same sparse errors.
>
> My commit stack: apply my patches onto broonie/sound.git for-next
> $ git log --oneline
> b4471777f5d8 (HEAD -> draft) ASoC: mediatek: mt8183: support WoV
> b6bb558fa59d ASoC: dt-bindings: mt8183: add ec-codec
> a08bede115d4 ASoC: mediatek: mt6358: support WoV
> f67068fd0c91 ASoC: cros_ec_codec: support WoV
> 7e11271c070e ASoC: dt-bindings: cros_ec_codec: add SHM bindings
> fd04f20e77d3 ASoC: cros_ec_codec: read max DMIC gain from EC codec
> c008f01d5bc3 platform/chrome: cros_ec: add common commands for EC codec
> 50d2c1f9b1f4 ASoC: cros_ec_codec: extract DMIC EC command from I2S RX
> 00e5a1c121eb ASoC: cros_ec_codec: refactor I2S RX
> 3f0c475d6ec8 platform/chrome: cros_ec: remove unused EC feature
> 3877dcd0194c (mark/for-next, asoc-next) Merge branch 'asoc-5.5' into
asoc-next
>
> My reproduce steps:
> $ make ARCH=x86_64 mrproper
> $ make ARCH=x86_64 allmodconfig
> $ make ARCH=x86_64 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' -j40
Find a typo from my previous message. It should be "make ARCH=... C=1 ...".
> 2>&1 | grep -v sched | tee log.txt
> (Note: filter sched out to not get flood sparse errors)
> $ grep cros_ec_codec log.txt
> CHECK sound/soc/codecs/cros_ec_codec.c
> CC [M] sound/soc/codecs/cros_ec_codec.o
>
> It did not generate the same message as 0day reported.
>
> One difference would be the sparse version (it is from "apt install"
> in my environment):
> $ sparse --version
> 0.6.0 (Debian: 0.6.0-3)
> On the other hand, 0day used "v0.6.1-rc1-42-g38eda53-dirty".
Hi,
The sparse warnings could be generated by the latest sparse
(
https://github.com/lucvoo/sparse.git). Could you try again?
Thanks. By using the version from github, it can generate the same
sparse errors.
$ sparse --version
v0.6.1-rc1-43-g0ccb3b4
It seems current debian's version (i.e. 0.6.0 (Debian: 0.6.0-3))
cannot reproduce the errors even without the typo mentioned above.
>
> Best Regards,
> Rong Chen
>
> >
> > Guenter, what we could do in the case? Do you have any idea?
> > _______________________________________________
> > kbuild-all mailing list -- kbuild-all(a)lists.01.org
> > To unsubscribe send an email to kbuild-all-leave(a)lists.01.org
>