* strange probe order issue after platform_add_devices()
@ 2007-08-18 14:29 hufey
2007-08-22 16:27 ` hufey
0 siblings, 1 reply; 2+ messages in thread
From: hufey @ 2007-08-18 14:29 UTC (permalink / raw)
To: linux-kernel
Hi list,
I am facing a very strange issue about device probe order.
I have a piece of code to support network interface card for an embedded box.
It likes following:
...
static struct platform_device *custom_devices[] __initdata = {
&a_device,
&b_device,
&c_device,
};
static int __init device_init()
{
return platform_add_devices(custom_devices,
ARRAY_SIZE(custom_devices);
}
subsys_initcall(device_init);
/* end */
This code works good with kernel 2.6.11. After platform_add_devices()
is executed, a_device's probe function will be called first then
b_device's probe function be called.
I merged it to 2.6.17, but strange thing is b_device's probe be called
first always now.
Since b_device probe depends on a_device, so it always fails now.
Does anyone know how to fix it? Or give me the hint about how kernel
decide which device be called to probe in order by what mechanism?
Thanks in advance and best regards,
hufeyy
^ permalink raw reply [flat|nested] 2+ messages in thread* Re: strange probe order issue after platform_add_devices()
2007-08-18 14:29 strange probe order issue after platform_add_devices() hufey
@ 2007-08-22 16:27 ` hufey
0 siblings, 0 replies; 2+ messages in thread
From: hufey @ 2007-08-22 16:27 UTC (permalink / raw)
To: linux-kernel
Seems never reach to the list...
On 8/18/07, hufey <hufeyy@gmail.com> wrote:
> Hi list,
>
> I am facing a very strange issue about device probe order.
> I have a piece of code to support network interface card for an embedded box.
>
> It likes following:
>
> ...
> static struct platform_device *custom_devices[] __initdata = {
> &a_device,
> &b_device,
> &c_device,
> };
>
> static int __init device_init()
> {
> return platform_add_devices(custom_devices,
> ARRAY_SIZE(custom_devices);
> }
>
> subsys_initcall(device_init);
>
> /* end */
>
> This code works good with kernel 2.6.11. After platform_add_devices()
> is executed, a_device's probe function will be called first then
> b_device's probe function be called.
> I merged it to 2.6.17, but strange thing is b_device's probe be called
> first always now.
> Since b_device probe depends on a_device, so it always fails now.
>
> Does anyone know how to fix it? Or give me the hint about how kernel
> decide which device be called to probe in order by what mechanism?
>
> Thanks in advance and best regards,
> hufeyy
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2007-08-22 16:27 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-08-18 14:29 strange probe order issue after platform_add_devices() hufey
2007-08-22 16:27 ` hufey
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®