On 6/10/2016 2:46 PM, Roger Quadros wrote:
Implementations might use different IRQs for
host, gadget so use named interrupt resources
to allow device tree to specify the interrupts.
Following are the interrupt names
Peripheral Interrupt - peripheral
HOST Interrupt - host
Maintain backward compatibility for a single named
interrupt ("dwc3_usb3") for all interrupts as well as
unnamed interrupt at index 0 for all interrupts.
As platform_get_irq_() variants are used, tackle
platform_get_irq().
OK.
the -EPROBE_DEFER case as well.
Signed-off-by: Roger Quadros <rogerq@xxxxxx>
---
v10:
- don't mention otg irq since we are not using it yet
- use platform_get_irq() and friends and check -EPROBE_DEFER case.
drivers/usb/dwc3/core.c | 22 ++++++++--------------
drivers/usb/dwc3/gadget.c | 29 ++++++++++++++++++++++++++---
drivers/usb/dwc3/host.c | 41 ++++++++++++++++++++++++++++++++++++++++-
3 files changed, 74 insertions(+), 18 deletions(-)
diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c
index 8fceeb1..131e7eb 100644
--- a/drivers/usb/dwc3/core.c
+++ b/drivers/usb/dwc3/core.c
[...]
diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
index 0f6fb8e..774a0d8 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
[...]
@@ -2866,7 +2865,31 @@ static irqreturn_t dwc3_interrupt(int irq, void *_evt)
*/
int dwc3_gadget_init(struct dwc3 *dwc)
{
- int ret;
+ int ret, irq;
+ struct platform_device *dwc3_pdev = to_platform_device(dwc->dev);
+
+ irq = platform_get_irq_byname(dwc3_pdev, "peripheral");
+ if (irq == -EPROBE_DEFER)
+ return irq;
+
+ if (irq <= 0) {
+ irq = platform_get_irq_byname(dwc3_pdev, "dwc_usb3");
+ if (irq == -EPROBE_DEFER)
+ return irq;
+
+ if (irq <= 0) {
+ irq = platform_get_irq(dwc3_pdev, 0);
+ if (irq <= 0) {
+ if (irq != -EPROBE_DEFER) {
+ dev_err(dwc->dev,
+ "missing peripheral IRQ\n");
+ }
+ return irq;
Iff irq == 0, you'll return success despite IRQ was "invalid". Was that intended?
good catch. It wasn't intended. I guess i'll return -EINVAL then?
[...]
I'd just consider 0 a valid IRQ, that's simpler. FYI, I've submitted to Greg KH a patch fixing platform_get_irq[_byname]() to not return 0 on failure. No reaction so far...
Maybe till your patch is in we can't really differentiate if it is error or not
so it is safer to consider it as error IMO.
Returning -EINVAL is fine then.
cheers,
-roger
MBR, Sergei
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
