| Message ID | 20141110223104.6419229d@endymion.delvare |
|---|---|
| State | Superseded |
| Headers | show |
On Mon, Nov 10, 2014 at 10:31:04PM +0100, Jean Delvare wrote: > The i2c-i801 driver can work without interrupts, so there is no reason > to make a request_irq failure fatal. Instead we can simply fallback > to polling. > > Signed-off-by: Jean Delvare <jdelvare@suse.de> > Cc: Wolfram Sang <wsa@the-dreams.de> > --- > drivers/i2c/busses/i2c-i801.c | 7 ++++--- > 1 file changed, 4 insertions(+), 3 deletions(-) > > --- linux-3.18-rc4.orig/drivers/i2c/busses/i2c-i801.c 2014-11-10 22:29:42.788955868 +0100 > +++ linux-3.18-rc4/drivers/i2c/busses/i2c-i801.c 2014-11-10 22:29:44.416991298 +0100 > @@ -1242,9 +1242,11 @@ static int i801_probe(struct pci_dev *de > if (err) { > dev_err(&dev->dev, "Failed to allocate irq %d: %d\n", > dev->irq, err); > - goto exit_release; > + priv->features &= ~FEATURE_IRQ; > + dev_info(&dev->dev, "SMBus using polling\n"); Shouldn't this message also be printed for !FEATURE_IRQ? I'd think it should be moved to another place. We can also deduce from the dev_err above that polling will be used, no? > + } else { > + dev_info(&dev->dev, "SMBus using PCI interrupt\n"); > } > - dev_info(&dev->dev, "SMBus using PCI Interrupt\n"); > } > > /* set up the sysfs linkage to our parent device */ > @@ -1272,7 +1274,6 @@ static int i801_probe(struct pci_dev *de > exit_free_irq: > if (priv->features & FEATURE_IRQ) > free_irq(dev->irq, priv); > -exit_release: > pci_release_region(dev, SMBBAR); > exit: > kfree(priv); > > -- > Jean Delvare > SUSE L3 Support
Hi Wolfram, Thanks for the quick review, very appreciated. On Tue, 11 Nov 2014 11:57:34 +0100, Wolfram Sang wrote: > On Mon, Nov 10, 2014 at 10:31:04PM +0100, Jean Delvare wrote: > > The i2c-i801 driver can work without interrupts, so there is no reason > > to make a request_irq failure fatal. Instead we can simply fallback > > to polling. > > > > Signed-off-by: Jean Delvare <jdelvare@suse.de> > > Cc: Wolfram Sang <wsa@the-dreams.de> > > --- > > drivers/i2c/busses/i2c-i801.c | 7 ++++--- > > 1 file changed, 4 insertions(+), 3 deletions(-) > > > > --- linux-3.18-rc4.orig/drivers/i2c/busses/i2c-i801.c 2014-11-10 22:29:42.788955868 +0100 > > +++ linux-3.18-rc4/drivers/i2c/busses/i2c-i801.c 2014-11-10 22:29:44.416991298 +0100 > > @@ -1242,9 +1242,11 @@ static int i801_probe(struct pci_dev *de > > if (err) { > > dev_err(&dev->dev, "Failed to allocate irq %d: %d\n", > > dev->irq, err); > > - goto exit_release; > > + priv->features &= ~FEATURE_IRQ; > > + dev_info(&dev->dev, "SMBus using polling\n"); > > Shouldn't this message also be printed for !FEATURE_IRQ? I'd think it > should be moved to another place. Good point, I'll rework that part of the code and resubmit. > We can also deduce from the dev_err > above that polling will be used, no? Having to deduce things doesn't make my supporter's life easy. Given the problems that have been reported by the i2c-i801 driver users lately, I'd rather be verbose during initialization to make bug reports easier to analyze. > > + } else { > > + dev_info(&dev->dev, "SMBus using PCI interrupt\n"); > > } > > - dev_info(&dev->dev, "SMBus using PCI Interrupt\n"); > > } > > > > /* set up the sysfs linkage to our parent device */ > > @@ -1272,7 +1274,6 @@ static int i801_probe(struct pci_dev *de > > exit_free_irq: > > if (priv->features & FEATURE_IRQ) > > free_irq(dev->irq, priv); > > -exit_release: > > pci_release_region(dev, SMBBAR); > > exit: > > kfree(priv); > >
--- linux-3.18-rc4.orig/drivers/i2c/busses/i2c-i801.c 2014-11-10 22:29:42.788955868 +0100 +++ linux-3.18-rc4/drivers/i2c/busses/i2c-i801.c 2014-11-10 22:29:44.416991298 +0100 @@ -1242,9 +1242,11 @@ static int i801_probe(struct pci_dev *de if (err) { dev_err(&dev->dev, "Failed to allocate irq %d: %d\n", dev->irq, err); - goto exit_release; + priv->features &= ~FEATURE_IRQ; + dev_info(&dev->dev, "SMBus using polling\n"); + } else { + dev_info(&dev->dev, "SMBus using PCI interrupt\n"); } - dev_info(&dev->dev, "SMBus using PCI Interrupt\n"); } /* set up the sysfs linkage to our parent device */ @@ -1272,7 +1274,6 @@ static int i801_probe(struct pci_dev *de exit_free_irq: if (priv->features & FEATURE_IRQ) free_irq(dev->irq, priv); -exit_release: pci_release_region(dev, SMBBAR); exit: kfree(priv);
The i2c-i801 driver can work without interrupts, so there is no reason to make a request_irq failure fatal. Instead we can simply fallback to polling. Signed-off-by: Jean Delvare <jdelvare@suse.de> Cc: Wolfram Sang <wsa@the-dreams.de> --- drivers/i2c/busses/i2c-i801.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)