Re: [PATCH 1/1] ARM: STi: Restore secondary CPU's bringup

[Patrice CHOTARD]

Hi Russell

On 12/18/18 6:27 PM, Russell King - ARM Linux wrote:
> On Tue, Dec 18, 2018 at 05:05:18PM +0000, Patrice CHOTARD wrote:
>> Hi Russell
>>
>> On 12/18/18 4:52 PM, Russell King - ARM Linux wrote:
>>> On Tue, Dec 18, 2018 at 03:48:13PM +0100, patrice.chotard@xxxxxx wrote:
>>>> From: Patrice Chotard <patrice.chotard@xxxxxx>
>>>>
>>>> Due to pen_release and boot_lock removal, secondary CPU's bringup
>>>> was broken. Restore CPU's bringup by reworking properly
>>>> .smp_prepare_cpus and .smp_boot_secondary STi callbacks.
>>>
>>> Sorry, maybe I don't understand your commit message, but you seem to be
>>> saying that removal of the pen_release and boot_lock broke STi's secondary
>>> CPU bring up?  Please clarify, and explain how that happened.
>>
>> Correct, CPU1 failed to come online.
>>
>> It seems that writing secondary_startup address at cpu-release-addr in
>> .smp_prepare_cpus callback was too early.
>>
>> Doing it in .smp_boot_secondary callback, insures that secondary_data
>> struct is populated in __cpu_up() (stack, pgdir and swapper_pg_dir fields).
> 
> Ah, you're saying that it causes the CPU to jump to secondary_startup
> while the boot CPU is in smp_prepare_cpus()?  What triggers the CPU

Yes

> to jump to the address written to cpu_strt_ptr?  What you're saying
> seems to suggest that it's the write to that address, rather than the
> IPI that's sent in sti_boot_secondary().

At platform startup, an U-Bootrom firmware initialize secondary CPU and
make it spinning waiting for a jump address to be written at cpu_strt_ptr.

I didn't pay attention to the IPI, you are right IPI is useless, i will
remove it.

> 
> If the IPI in sti_boot_secondary() isn't doing anything, it ought to
> be removed.  It'd also be a good idea to document what's going on as
> comments in the code for future maintanence.

Agree, i will add a comment.

Thanks

Patrice

> 
>> If you are ok, i will pick up your patch [1] and this one to prepare a
>> STi pull-request.
> 
> Yes, although I'll have to delay patch 9 as a result.
>