[U-Boot] Please pull u-boot-usb/topic-xhci

[Marek Vasut]

On 03/20/2018 04:34 AM, Heinrich Schuchardt wrote:
> On 03/20/2018 02:36 AM, Bin Meng wrote:
>> Hi Marek,
>>
>> On Mon, Mar 19, 2018 at 5:42 PM, Marek Vasut <marex at denx.de> wrote:
>>> On 03/19/2018 04:17 AM, Bin Meng wrote:
>>>> Hi Marek,
>>>>
>>>> The following changes since commit 958ad42b77be9d9a69f059066622ef0c15c603ee:
>>>>
>>>>   usb: dwc2: Replace printf, pr_err by dev_info, dev_err (2018-03-19
>>>> 11:03:46 +0800)
>>>>
>>>> are available in the git repository at:
>>>>
>>>>   git://git.denx.de/u-boot-usb.git topic-xhci
>>>>
>>>> for you to fetch changes up to 24b36b94c661ddbcb8d211a33ed4055178208211:
>>>>
>>>>   usb: xhci: remove superfluous assignment (2018-03-19 11:09:10 +0800)
>>>>
>>>> ----------------------------------------------------------------
>>>> Heinrich Schuchardt (4):
>>>>       usb: xhci-exynos5: correct error checking
>>>>       usb: xhci-keystone: remove superfluous assignment
>>>>       usb: xhci-ring: remove superfluous assignment
>>>>       usb: xhci: remove superfluous assignment
>>>>
>>>>  drivers/usb/host/xhci-exynos5.c  | 2 +-
>>>>  drivers/usb/host/xhci-keystone.c | 4 ++--
>>>>  drivers/usb/host/xhci-ring.c     | 2 +-
>>>>  drivers/usb/host/xhci.c          | 4 ++--
>>>
>>> I believe all of those patches were completely untested and should first
>>> be applied to Linux and then backported, so I won't take this PR.
>>
>> Why do we require U-Boot driver fixes to be firstly applied to Linux,
>> then backported? Some of them was ported from Linux before, but as
>> time goes I don't think they are a 100% match now. Besides, these 4
>> fixes are really obvious.
>>
>> Regards,
>> Bin
>>
> 
> 
> Hello Marek,
> 
> I fully understand your point that you want patches first to go
> upstream. Unfortunately for these 4 patches the patched functions do not
> exist in Linux drivers/usb. So I am not able to upstream any of these.
> 
> Out of the four patches three are just eliminating a superfluous
> assignment. If you say eliminating this is not worth the effort and
> risk, I am fine with it.
> 
> "usb: xhci-exynos5: correct error checking"
> https://lists.denx.de/pipermail/u-boot/2018-March/323104.html
> points to a real inconsistency inside the coding reading the device
> tree. I must admit I do not have the right board for testing.
> 
> drivers/usb/host/ehci-exynos.c has the same problem and should also be
> fixed.

My other concern about this flurry of automated patches is that they
were not tested at all, right ? CFR my reply to the faraday hci patch
for example.

-- 
Best regards,
Marek Vasut