[Skiboot] [PATCH] platforms/firenze: Fix clock frequency dt property

[Stewart Smith]

Vasant Hegde <hegdevasant at linux.vnet.ibm.com> writes:
> On 09/28/2016 11:53 AM, Vasant Hegde wrote:
>> Commit 5cda6f6d added 8 byte property instead of 4 byte..which resulted
>> in below calltrace.
>>
>> I think its fine to convert u64 to u32 here as we devide bus frequency
>> by 4.
>>
>> Backtrace:
>> ----------
>>    [ 1.212366090,3] DT: Unexpected property length /xscom at 3fc0000000000/i2cm at a0020/clock-frequency
>>    [ 1.212369108,3] DT: Expected len: 4 got len: 8
>>    [ 1.212370117,0] Assert fail: core/device.c:603:0
>>    [ 1.212371550,0] Aborting!
>>    CPU 0870 Backtrace:
>>    S: 0000000033dc39e0 R: 0000000030013758 .backtrace+0x24
>>    S: 0000000033dc3a60 R: 0000000030018e0c ._abort+0x4c
>>    S: 0000000033dc3ae0 R: 0000000030018e88 .assert_fail+0x34
>>    S: 0000000033dc3b60 R: 0000000030023da4 .dt_require_property+0xb4
>>    S: 0000000033dc3bf0 R: 000000003002403c .dt_prop_get_u32+0x14
>>    S: 0000000033dc3c60 R: 000000003004e884 .p8_i2c_init+0x12c
>>    S: 0000000033dc3e30 R: 0000000030014684 .main_cpu_entry+0x4a8
>>    S: 0000000033dc3f00 R: 00000000300025a0 boot_entry+0x198
>>
>> Fixes: 5cda6f6d (platforms/firenze: Fix I2C clock source frequency)
>
> Stewart,
>
> Original issue was introduced by 5cda6f6d but we never hit this issue because of 
>   5acf424a.
> I should have added "fixes 5acf424a" as well.

Added when merged.

merged to master as of 94125da004650df0133e7dbdbd8c3833c53b4902
and cherry picked into stable (as both Fixes commits were there) as of
0b8da32bb37ef806331a49c551261a58ff03b8ba


-- 
Stewart Smith
OPAL Architect, IBM.