RE: [PATCH] dma: dw: Add suspend and resume handling for PCI modeDW_DMAC.

[Chew, Chiau Ee]

Vinod,

As mentioned by Andy, we are using *_noirq  verion of suspend/resume PM callback whereby the callbacks would be executed after IRQ handlers have been disabled. If using SET_SYSTEM_SLEEP_PM_OPS, it would be the normal suspend/resume PM callback. Looking at the Desginware DMAC platform code (drivers/dma/dw/platform.c), it is using the *_noirq suspend/resume PM callback. Is it advisable to use the normal suspend/resume PM callback instead of *_noirq suspend/PM callback? 

Thanks,
Chiau Ee

-----Original Message-----
From: Shevchenko, Andriy 
Sent: Tuesday, December 10, 2013 7:57 PM
To: Koul, Vinod
Cc: Chew, Chiau Ee; Viresh Kumar; Andy Shevchenko; Williams, Dan J; dmaengine@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
Subject: Re: [PATCH] dma: dw: Add suspend and resume handling for PCI mode DW_DMAC.

On Tue, 2013-12-10 at 15:40 +0530, Vinod Koul wrote:
> On Tue, Dec 10, 2013 at 10:06:13PM +0800, Chew Chiau Ee wrote:
> > From: Chew, Chiau Ee <chiau.ee.chew@xxxxxxxxx>
> > 
> > This is to disable/enable DW_DMAC hw during suspend/resume.
> > 
> > Signed-off-by: Chew, Chiau Ee <chiau.ee.chew@xxxxxxxxx>
> > Acked-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> > ---
> >  drivers/dma/dw/pci.c |   33 +++++++++++++++++++++++++++++++++
> >  1 files changed, 33 insertions(+), 0 deletions(-)
> > 
> > diff --git a/drivers/dma/dw/pci.c b/drivers/dma/dw/pci.c index 
> > e89fc24..97bc3a2 100644
> > --- a/drivers/dma/dw/pci.c
> > +++ b/drivers/dma/dw/pci.c
> > @@ -75,6 +75,36 @@ static void dw_pci_remove(struct pci_dev *pdev)
> >  		dev_warn(&pdev->dev, "can't remove device properly: %d\n", ret);  
> > }
> >  
> > +#ifdef CONFIG_PM_SLEEP
> > +
> > +static int dw_pci_suspend_noirq(struct device *dev) {
> > +	struct pci_dev *pci = to_pci_dev(dev);
> > +	struct dw_dma_chip *chip = pci_get_drvdata(pci);
> > +
> > +	return dw_dma_suspend(chip);
> > +};
> > +
> > +static int dw_pci_resume_noirq(struct device *dev) {
> > +	struct pci_dev *pci = to_pci_dev(dev);
> > +	struct dw_dma_chip *chip = pci_get_drvdata(pci);
> > +
> > +	return dw_dma_resume(chip);
> > +};
> > +
> > +#else /* !CONFIG_PM_SLEEP */
> > +
> > +#define dw_pci_suspend_noirq	NULL
> > +#define dw_pci_resume_noirq	NULL
> > +
> > +#endif /* !CONFIG_PM_SLEEP */
> How about SET_SYSTEM_SLEEP_PM_OPS instead?

So, we are using *_noirq versions of the functions here. What happened when we switch to normal ones? Any side effects?

> 
> --
> ~Vinod
> > +
> > +static const struct dev_pm_ops dw_pci_dev_pm_ops = {
> > +	.suspend_noirq = dw_pci_suspend_noirq,
> > +	.resume_noirq = dw_pci_resume_noirq, };
> > +
> >  static DEFINE_PCI_DEVICE_TABLE(dw_pci_id_table) = {
> >  	/* Medfield */
> >  	{ PCI_VDEVICE(INTEL, 0x0827), (kernel_ulong_t)&dw_pci_pdata }, @@ 
> > -92,6 +122,9 @@ static struct pci_driver dw_pci_driver = {
> >  	.id_table	= dw_pci_id_table,
> >  	.probe		= dw_pci_probe,
> >  	.remove		= dw_pci_remove,
> > +	.driver	= {
> > +		.pm	= &dw_pci_dev_pm_ops,
> > +	},
> >  };
> >  
> >  module_pci_driver(dw_pci_driver);
> > --
> > 1.7.4.4
> > 
> 

--
Andy Shevchenko <andriy.shevchenko@xxxxxxxxx> Intel Finland Oy
N‹§²æìr¸›yúèšØb²X¬¶ÇvØ^–)Þ{.nÇ+‰·¥Š{±‘êçzX§¶›¡Ü}©ž²ÆzÚ&j:+v‰¨¾«‘êçzZ+€Ê+zf£¢·hšˆ§~†­†Ûiÿûàz¹®w¥¢¸?™¨è­Ú&¢)ßf”ù^jÇy§m…á@A«a¶Úÿ0¶ìh®å’i