Re: [PATCH] spi: mark device nodes only in case of successful instantiation

[Pantelis Antoniou]

Hi Ralf,

> On Oct 16, 2016, at 12:55 , Ralf Ramsauer <ralf@xxxxxxxxxxxxxxxxxxxxxx> wrote:
> 
> Hi Geert,
> 
> On 10/16/2016 10:49 AM, Geert Uytterhoeven wrote:
>> Hi Ralf,
>> 
>> (Cc i2c)
>> 
>> On Fri, Oct 14, 2016 at 9:31 PM, Ralf Ramsauer
>> <ralf@xxxxxxxxxxxxxxxxxxxxxx> wrote:
>>> Instantiated SPI device nodes are marked with OF_POPULATE. This was
>>> introduced in bd6c164. On unloading, loaded device nodes will of course
>>> be unmarked. The problem are nodes the fail during initialisation: If a
>>> node failed during registration, it won't be unloaded and hence never be
>>> unmarked again.
>>> 
>>> So if a SPI driver module is unloaded and reloaded, it will skip nodes
>>> that failed before.
>>> 
>>> Skip device nodes that are already populated and mark them only in case
>>> of success.
>>> 
>>> Fixes: bd6c164 ("spi: Mark instantiated device nodes with OF_POPULATE")
>>> Signed-off-by: Ralf Ramsauer <ralf@xxxxxxxxxxxxxxxxxxxxxx>
>>> Cc: Geert Uytterhoeven <geert+renesas@xxxxxxxxx>
>>> ---
>>> Hi,
>>> 
>>> imagine the following situation: you loaded a spi driver as module, but
>>> it fails to instantiate, because of some reasons (e.g. some resources,
>>> like gpios, might be in use in userspace).
>>> 
>>> When reloading the driver, _all_ nodes, including previously failed
>>> ones, should be probed again. This is not the case at the moment.
>>> Current behaviour only re-registers nodes that were previously
>>> successfully loaded.
>>> 
>>> This small patches fixes this behaviour. I stumbled over this while
>>> working on a spi driver.
>> 
>> Thanks for your patch!
>> 
>>> drivers/spi/spi.c | 7 +++++--
>>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>> 
>>> diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
>>> index 200ca22..f96a04e 100644
>>> --- a/drivers/spi/spi.c
>>> +++ b/drivers/spi/spi.c
>>> @@ -1604,12 +1604,15 @@ static void of_register_spi_devices(struct spi_master *master)
>>>                return;
>>> 
>>>        for_each_available_child_of_node(master->dev.of_node, nc) {
>>> -               if (of_node_test_and_set_flag(nc, OF_POPULATED))
>>> +               if (of_node_check_flag(nc, OF_POPULATED))
>>>                        continue;
>>>                spi = of_register_spi_device(master, nc);
>>> -               if (IS_ERR(spi))
>>> +               if (IS_ERR(spi)) {
>>>                        dev_warn(&master->dev, "Failed to create SPI device for %s\n",
>>>                                nc->full_name);
>>> +                       continue;
>>> +               }
>>> +               of_node_set_flag(nc, OF_POPULATED);
>> 
>> I think it's safer to keep the atomic test-and-set, but clear the flag on
>> failure, cfr. of_platform_device_create_pdata() and of_amba_device_create().
> Ack, no prob. Let me change this in the next version.
>> 
>> Shouldn't of_spi_notify() be fixed, too?
> Right, that's almost the same path.
>> 
>> The same issue exists for i2c in of_i2c_register_devices() and of_i2c_notify(),
>> which is what I had used as an example.
> Good old c&p ;-)
> I'll fix and test that tomorrow and come back with two patches, as it
> touches different subsystems.
> 

Thanks for this. This is a very rare case thatâs easy to slip through.
It is good to be consistent :)

> Best
>  Ralf
>> 

Regards

â Pantelis

>> Gr{oetje,eeting}s,
>> 
>>                        Geert
>> 
>> --
>> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
>> 
>> In personal conversations with technical people, I call myself a hacker. But
>> when I'm talking to journalists I just say "programmer" or something like that.
>>                                -- Linus Torvalds
>> 
> 
> 
> -- 
> Ralf Ramsauer
> GPG: 0x8F10049B