Hi all,
Today's linux-next merge of the arm-soc tree got a conflict in
arch/arm/mach-shmobile/clock-r8a7740.c between commit e5c9b4cd6651
("sh_eth: get R8A7740 support out of #ifdef") from the net-next tree and
commit 9e0b428f079d ("ARM: shmobile: r8a7740: Add interim sh-eth device
name to clocks list") from the arm-soc tree.
I fixed it up (see below) and can carry the fix as necessary (no action
is required).
--
Cheers,
Stephen Rothwell [email protected]
diff --cc arch/arm/mach-shmobile/clock-r8a7740.c
index ae93f94,7fd32d6..0000000
--- a/arch/arm/mach-shmobile/clock-r8a7740.c
+++ b/arch/arm/mach-shmobile/clock-r8a7740.c
@@@ -591,7 -594,9 +594,9 @@@ static struct clk_lookup lookups[] =
CLKDEV_DEV_ID("e6860000.sdhi", &mstp_clks[MSTP313]),
CLKDEV_DEV_ID("sh_mmcif", &mstp_clks[MSTP312]),
CLKDEV_DEV_ID("e6bd0000.mmcif", &mstp_clks[MSTP312]),
- CLKDEV_DEV_ID("sh-eth", &mstp_clks[MSTP309]),
+ CLKDEV_DEV_ID("r8a7740-gether", &mstp_clks[MSTP309]),
+ CLKDEV_DEV_ID("e9a00000.sh-eth", &mstp_clks[MSTP309]),
+ CLKDEV_DEV_ID("renesas_tpu_pwm", &mstp_clks[MSTP304]),
CLKDEV_DEV_ID("sh_mobile_sdhi.2", &mstp_clks[MSTP415]),
CLKDEV_DEV_ID("e6870000.sdhi", &mstp_clks[MSTP415]),
Hello.
On 17-06-2013 10:39, Stephen Rothwell wrote:
> Today's linux-next merge of the arm-soc tree got a conflict in
> arch/arm/mach-shmobile/clock-r8a7740.c between commit e5c9b4cd6651
> ("sh_eth: get R8A7740 support out of #ifdef") from the net-next tree and
> commit 9e0b428f079d ("ARM: shmobile: r8a7740: Add interim sh-eth device
> name to clocks list") from the arm-soc tree.
To me that latter commit looked utterly pointless, as there's no
device tree support for the 'sh-eth' driver yet (and at this stage it
isn't even going to happen due to procedural platform data). I'm instead
going to use OF_DEV_AUXDATA() in the platform code which again would
render that commit pointless.
> I fixed it up (see below) and can carry the fix as necessary (no action
> is required).
Thanks.
WBR, Sergei
Hello Sergei, hello Stephen,
Sergei is completely right I think. He was sceptic about the commit in
a mail some days ago already. Unfortunately I recently haven't had
access to my development environment to check what's up with this
patch. I'm sorry that it ended up to you both to deal with it. I have
written some simple bindings for the sh-eth and wanted to use it for
our DT reference version for the Armadillo board but discarded it as I
saw Nobuhiro Iwamatsu, the author of the sh-eth driver, cooked up some
more mature version. I don't see them in the next branch of Simon, so
until they appear there we can delete commit 9e0b428f079d ("ARM:
shmobile: r8a7740: Add interim sh-eth device name to clocks list")
that is only needed for sh-eth DT usage.
Thanks,
Bastian
2013/6/17 Sergei Shtylyov <[email protected]>:
> Hello.
>
>
> On 17-06-2013 10:39, Stephen Rothwell wrote:
>
>> Today's linux-next merge of the arm-soc tree got a conflict in
>> arch/arm/mach-shmobile/clock-r8a7740.c between commit e5c9b4cd6651
>> ("sh_eth: get R8A7740 support out of #ifdef") from the net-next tree and
>> commit 9e0b428f079d ("ARM: shmobile: r8a7740: Add interim sh-eth device
>> name to clocks list") from the arm-soc tree.
>
>
> To me that latter commit looked utterly pointless, as there's no device
> tree support for the 'sh-eth' driver yet (and at this stage it isn't even
> going to happen due to procedural platform data). I'm instead going to use
> OF_DEV_AUXDATA() in the platform code which again would render that commit
> pointless.
>
>
>> I fixed it up (see below) and can carry the fix as necessary (no action
>> is required).
>
>
> Thanks.
>
> WBR, Sergei
>
Hello.
On 18-06-2013 12:33, Bastian Hecht wrote:
> Sergei is completely right I think. He was sceptic about the commit in
> a mail some days ago already. Unfortunately I recently haven't had
> access to my development environment to check what's up with this
> patch. I'm sorry that it ended up to you both to deal with it. I have
> written some simple bindings for the sh-eth and wanted to use it for
> our DT reference version for the Armadillo board but discarded it as I
> saw Nobuhiro Iwamatsu, the author of the sh-eth driver, cooked up some
> more mature version. I don't see them in the next branch of Simon, so
> until they appear there we can delete commit 9e0b428f079d ("ARM:
It won't appear there. I'll just NAK it, I think I've alredy told
Iwamatsu-san that the full blown device tree binding for this driver is
no-go. The best we can do is use OF_DEV_AUXDATA() to rename the DT
sourced platform device to a traditional one and assign to it the
platform data.
> shmobile: r8a7740: Add interim sh-eth device name to clocks list")
Unfortunately, it's now too late to just delete it at this point. I
missed the patch unfortunately when it was just submitted.
> that is only needed for sh-eth DT usage.
> Thanks,
> Bastian
WBR, Sergei