mbox series

[stable-5.4,0/2] net: dsa: lantiq_gswip: backports for Linux 5.4

Message ID 20210411102344.2834328-1-martin.blumenstingl@googlemail.com
Headers show
Series net: dsa: lantiq_gswip: backports for Linux 5.4 | expand

Message

Martin Blumenstingl April 11, 2021, 10:23 a.m. UTC
Hello,

This backports two patches (which could not be backported automatically
because the gswip_phylink_mac_link_up function is different in Linux 5.4
compared to 5.7 and newer) for the lantiq_gswip driver:
- commit 3e9005be87777afc902b9f5497495898202d335d upstream.
- commit 4b5923249b8fa427943b50b8f35265176472be38 upstream.

This is the first time that I am doing such a backport so I am not
sure how to mention the required modifications. I added them at the
bottom of each patch with another Signed-off-by. If this is not correct
then please suggest how I can do it rights.


Thank you!
Martin


Martin Blumenstingl (2):
  net: dsa: lantiq_gswip: Don't use PHY auto polling
  net: dsa: lantiq_gswip: Configure all remaining GSWIP_MII_CFG bits

 drivers/net/dsa/lantiq_gswip.c | 203 ++++++++++++++++++++++++++++-----
 1 file changed, 175 insertions(+), 28 deletions(-)

Comments

Martin Blumenstingl April 11, 2021, 5:17 p.m. UTC | #1
Hi Sasha,

On Sun, Apr 11, 2021 at 6:48 PM Sasha Levin <sashal@kernel.org> wrote:
>
> On Sun, Apr 11, 2021 at 12:23:42PM +0200, Martin Blumenstingl wrote:
> >Hello,
> >
> >This backports two patches (which could not be backported automatically
> >because the gswip_phylink_mac_link_up function is different in Linux 5.4
> >compared to 5.7 and newer) for the lantiq_gswip driver:
> >- commit 3e9005be87777afc902b9f5497495898202d335d upstream.
> >- commit 4b5923249b8fa427943b50b8f35265176472be38 upstream.
> >
> >This is the first time that I am doing such a backport so I am not
> >sure how to mention the required modifications. I added them at the
> >bottom of each patch with another Signed-off-by. If this is not correct
> >then please suggest how I can do it rights.
>
> Hey Martin,
>
> Your backport works, but I'd rather take 5b502a7b2992 ("net: dsa:
> propagate resolved link config via mac_link_up()") along with the
> backport instead. This means that we don't diverge from upstream too
> much and will make future backports easier.
>
> I've queued up these 3 commits to 5.4, thanks!
in general I am fine with your suggested approach. however, I think at
least one more backport is required then:
91a208f2185ad4855ff03c342d0b7e4f5fc6f5df ("net: phylink: propagate
resolved link config via mac_link_up()")
Patches should be backported in a specific order also so we don't
break git bisect:
- phylink patch
- dsa patch
- the two lantiq GSWIP patches


Best regards,
Martin
Sasha Levin April 11, 2021, 6:17 p.m. UTC | #2
On Sun, Apr 11, 2021 at 07:17:46PM +0200, Martin Blumenstingl wrote:
>Hi Sasha,
>
>On Sun, Apr 11, 2021 at 6:48 PM Sasha Levin <sashal@kernel.org> wrote:
>>
>> On Sun, Apr 11, 2021 at 12:23:42PM +0200, Martin Blumenstingl wrote:
>> >Hello,
>> >
>> >This backports two patches (which could not be backported automatically
>> >because the gswip_phylink_mac_link_up function is different in Linux 5.4
>> >compared to 5.7 and newer) for the lantiq_gswip driver:
>> >- commit 3e9005be87777afc902b9f5497495898202d335d upstream.
>> >- commit 4b5923249b8fa427943b50b8f35265176472be38 upstream.
>> >
>> >This is the first time that I am doing such a backport so I am not
>> >sure how to mention the required modifications. I added them at the
>> >bottom of each patch with another Signed-off-by. If this is not correct
>> >then please suggest how I can do it rights.
>>
>> Hey Martin,
>>
>> Your backport works, but I'd rather take 5b502a7b2992 ("net: dsa:
>> propagate resolved link config via mac_link_up()") along with the
>> backport instead. This means that we don't diverge from upstream too
>> much and will make future backports easier.
>>
>> I've queued up these 3 commits to 5.4, thanks!
>in general I am fine with your suggested approach. however, I think at
>least one more backport is required then:
>91a208f2185ad4855ff03c342d0b7e4f5fc6f5df ("net: phylink: propagate
>resolved link config via mac_link_up()")
>Patches should be backported in a specific order also so we don't
>break git bisect:
>- phylink patch
>- dsa patch
>- the two lantiq GSWIP patches

Good point, I haven't realized there's an additional phylink patch
(which on it's own requires more dependencies and follow-ups). In this
case I'll just grab your backport, thanks!