Re: Unbalanced IRQ wake disable during resume from static suspend

[Kevin Hilman <khilman@deeprootsystems.com>]

Paul Walmsley <paul@pwsan.com> writes:

> Hi Kevin
>
> On Thu, 2 Dec 2010, Kevin Hilman wrote:
>
>> I guess this hasn't been seen before since we haven't tested the sysfs
>> wakeup interface for the omap-serial driver.  For on-chip OMAP UARTs,
>> using the sysfs interface isn't needed as the serial core is already
>> doing device_init_wakeup(dev, true);
>
> Is this the code you're referring to, in serial_core.c?
>
> 	tty_dev = tty_register_device(drv->tty_driver, uport->line, uport->dev);
> 	if (likely(!IS_ERR(tty_dev))) {
> 		device_init_wakeup(tty_dev, 1);
> 		device_set_wakeup_enable(tty_dev, 0);
> 	} else
>
> I may be misreading it, but it appears that the code leaves wakeups 
> disabled for the serial port, by default.

No, I was referring to the code in mach-omap2/serial.c, in omap_serial_init_port():

	if ((cpu_is_omap34xx() && uart->padconf) ||
	    (uart->wk_en && uart->wk_mask)) {
		device_init_wakeup(&od->pdev.dev, true);
		DEV_CREATE_FILE(&od->pdev.dev, &dev_attr_sleep_timeout);
	}

Kevin

> As an aside, this code is somewhat perplexing: it doesn't seem accurate to 
> assume that every serial device really is capable of waking up the system.