Re: [PATCH 4.4 000/159] 4.4.203-stable review

[Jon Hunter]

On 25/11/2019 22:45, Jon Hunter wrote:
> 
> On 25/11/2019 16:03, Greg Kroah-Hartman wrote:
>> On Mon, Nov 25, 2019 at 01:22:58PM +0000, Jon Hunter wrote:
>>>
>>> On 25/11/2019 09:41, Greg Kroah-Hartman wrote:
>>>> On Sun, Nov 24, 2019 at 08:31:46PM +0000, Jon Hunter wrote:
>>>>>
>>>>> On 23/11/2019 15:46, Guenter Roeck wrote:
>>>>>> On 11/22/19 6:48 AM, Jon Hunter wrote:
>>>>>>
>>>>>> [ ... ]
>>>>>>
>>>>>>> Error: arch/arm/boot/dts/omap5-board-common.dtsi:636.1-6 Label or path
>>>>>>> dwc3 not found
>>>>>>> FATAL ERROR: Syntax error parsing input tree
>>>>>>> scripts/Makefile.lib:293: recipe for target
>>>>>>> 'arch/arm/boot/dts/omap5-igep0050.dtb' failed
>>>>>>> make[1]: *** [arch/arm/boot/dts/omap5-igep0050.dtb] Error 1
>>>>>>> arch/arm/Makefile:338: recipe for target 'dtbs' failed
>>>>>>> make: *** [dtbs] Error 2
>>>>>>>
>>>>>>>
>>>>>>> This is caused by the following commit ...
>>>>>>>
>>>>>>> commit d0abc07b3d752cbe2a8d315f662c53c772caed0f
>>>>>>> Author: H. Nikolaus Schaller <hns@xxxxxxxxxxxxx>
>>>>>>> Date:ÂÂ Fri Sep 28 17:54:00 2018 +0200
>>>>>>>
>>>>>>> ÂÂÂÂ ARM: dts: omap5: enable OTG role for DWC3 controller
>>>>>>>
>>>>>>
>>>>>> On top of the breakage caused by this patch, I would also argue
>>>>>> that it is not a bug fix and should not have been included
>>>>>> in the first place.
>>>>>>
>>>>>> The dwc3 label was added with commit 4c387984618fe ("ARM: dts: omap5:
>>>>>> Add l4 interconnect hierarchy and ti-sysc data"). Given the size of
>>>>>> that patch, I highly doubt that a backport to 4.4 would work.
>>>>
>>>> Good catch, I have now dropped both of these patches and pushed out a
>>>> -rc3
>>>>
>>>>> FYI ... I am still seeing a build failure because of this with -rc2 ...
>>>>
>>>> Can you see if -rc3 is also giving you problems?
>>>
>>> Better, but I appear to be seeing some random suspend failures with this
>>> now on one board. I will try to bisect this.
>>>
>>> Test results for stable-v4.4:
>>>     6 builds:	6 pass, 0 fail
>>>     12 boots:	12 pass, 0 fail
>>>     19 tests:	18 pass, 1 fail
>>>
>>> Linux version:	4.4.203-rc3-g2576206c30b5
>>> Boards tested:	tegra124-jetson-tk1, tegra20-ventana,
>>>                 tegra30-cardhu-a04
>>
>> Odd.  If you find anything interesting, please let me know.
> 
> Yes will do. Bisect has not found anything yet, so will keeping looking
> to see if this is a false-positive or not.

I have been doing some more testing and it appears that I see
intermittent suspend failures on Tegra124 with linux-4.4.y. Apparently
this issue has always been there and appears to be fixed in v4.5. I am
still trying to find a way to fix this. The changes in v4.5 that appear
to resolve this are too complex for stable (several patches in Tegra clk
drivers). Anyway, ignore this for now, I will see how we can fix or
workaround for linux-4.4.y.

Jon

-- 
nvpublic