Re: Keystone 2 boards boot failure

["Franklin S Cooper Jr." <fcooper@ti.com>]

On 02/02/2016 07:19 PM, Franklin S Cooper Jr. wrote:
>
> On 02/02/2016 05:26 PM, Arnd Bergmann wrote:
>> On Tuesday 02 February 2016 16:59:34 Franklin Cooper wrote:
>>> On 02/02/2016 03:26 PM, Arnd Bergmann wrote:
>>>> On Tuesday 02 February 2016 15:01:33 Franklin S Cooper Jr. wrote:
>>>>> Yes. Here is a boot log on the latest master with the below
>>>>> three patches reverted.
>>>>> http://pastebin.com/W7RWSHpE (Working)
>>>>>
>>>>> I reverted these three patches. The two latest patches seem
>>>>> to be trying to correct/expand upon the last patch on this list.
>>>>>
>>>>> commit 958d104e3d40eef5148c402887138f6594ff7e1e
>>>>> netcp: fix regression in receive processing
>>>>>
>>>>> commit 9dd2d6c5c9755b160fe0111bcdad9491676feea8
>>>>> netcp: add more __le32 annotations
>>>>>
>>>>> commit 899077791403ff7a2d8cfaa87bd1a82d729463e2
>>>>> netcp: try to reduce type confusion in descriptors
>>>>>
>>>> The middle patch should have no effect on generated code, so I'm ignoring
>>>> that for now.
>>>>
>>>> The next thing to rule out is an endianess bug. I assume you
>>>> are running this on with a little-endian kernel, correct? If
>>>> you are running big-endian, the base assumption that the driver needs
>>>> to swap the data was flawed and that portion needs to be done.
>>>>
>>>> If you are running little-endian 32-bit, please try the partial
>>>> revert below, which just undoes the attempt to make it work with
>>>> 64-bit kernels.
>>> Keystone 2 devices are little-endian 32-bit devices.
>> I meant the kernel you are running on it, not the hardware.
>> You should always be able to run both a big-endian kernel and
>> a littl-endian kernel on any ARMv7 machine, and a couple of
>> platforms use 64-bit physical addresses even on 32-bit machines
>> (with the normal 32-bit instruction set).
> I'm not sure if Keystone 2 devices support this or if we
> have support for this. I'll have to double check.
>> I wasn't completely sure if there are already keystone-derived
>> products with 64-bit CPU cores, but I guess the driver would
>> fail really badly on those (with or without the patch).
>>
>>> This partial revert fixes the boot problem for me.
>> Ok.
>>
>>
>> I tried to create a smaller version and stumbled over
>> a typo, maybe that's the whole problem. Can you try this one:
>>
>> diff --git a/drivers/net/ethernet/ti/netcp_core.c b/drivers/net/ethernet/ti/netcp_core.c
>> index c61d66d38634..8490804416dd 100644
>> --- a/drivers/net/ethernet/ti/netcp_core.c
>> +++ b/drivers/net/ethernet/ti/netcp_core.c
>> @@ -167,7 +167,7 @@ static void set_pad_info(u32 pad0, u32 pad1, u32 pad2, struct knav_dma_desc *des
>>  {
>>  	desc->pad[0] = cpu_to_le32(pad0);
>>  	desc->pad[1] = cpu_to_le32(pad1);
>> -	desc->pad[2] = cpu_to_le32(pad1);
>> +	desc->pad[2] = cpu_to_le32(pad2);
>>  }
>>  
>>  static void set_org_pkt_info(dma_addr_t buff, u32 buff_len,
>>
>>
>> 	Arnd
> So only making this change on the latest master with no
> other changes I see the boot problem again.

+Grygorii