diff options
| author | Marc Zyngier <maz@misterjones.org> | 2008-08-18 13:08:42 +0200 |
|---|---|---|
| committer | Greg Kroah-Hartman <gregkh@suse.de> | 2008-10-17 23:40:52 +0200 |
| commit | 27140219373327f2291da5d74a78db0105b15060 (patch) | |
| tree | 1cbb0eab32d9b847b0f030982429045ff0e8f041 /drivers/usb/host/sl811-hcd.c | |
| parent | ub: remove sg_stat (diff) | |
| download | linux-27140219373327f2291da5d74a78db0105b15060.tar.xz linux-27140219373327f2291da5d74a78db0105b15060.zip | |
USB: Let some USB host controllers get IRQ flags from resource
[This version fixes a thinko in the r8a66597 driver]
This patch let a few discrete USB host controllers drivers (isp116x-hcd,
r8a66597-hcd and sl811-hcd) obtain IRQ flags from their IORESOURCE_IRQ
resource if configured as such, much like it's been done for the smc91x
driver.
It spares people writing support for specific boards the burden to
configure the interrupt controller independantly, and keeps all IRQ
related information in a single resource.
HCD that are integrally part of a SoC have been left aside, as there
is probably no "wiring" options...
Tested on an Xscale PXA-255 based platform with isp116x-hcd.
Signed-off-by: Marc Zyngier <marc.zyngier@altran.com>
Cc: David Brownell <david-b@pacbell.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Diffstat (limited to 'drivers/usb/host/sl811-hcd.c')
| -rw-r--r-- | drivers/usb/host/sl811-hcd.c | 15 |
1 files changed, 11 insertions, 4 deletions
diff --git a/drivers/usb/host/sl811-hcd.c b/drivers/usb/host/sl811-hcd.c index 8a74bbb57d08..e106e9d48d4a 100644 --- a/drivers/usb/host/sl811-hcd.c +++ b/drivers/usb/host/sl811-hcd.c @@ -1620,22 +1620,26 @@ sl811h_probe(struct platform_device *dev) { struct usb_hcd *hcd; struct sl811 *sl811; - struct resource *addr, *data; + struct resource *addr, *data, *ires; int irq; void __iomem *addr_reg; void __iomem *data_reg; int retval; u8 tmp, ioaddr = 0; + unsigned long irqflags; /* basic sanity checks first. board-specific init logic should * have initialized these three resources and probably board * specific platform_data. we don't probe for IRQs, and do only * minimal sanity checking. */ - irq = platform_get_irq(dev, 0); - if (dev->num_resources < 3 || irq < 0) + ires = platform_get_resource(dev, IORESOURCE_IRQ, 0); + if (dev->num_resources < 3 || !ires) return -ENODEV; + irq = ires->start; + irqflags = ires->flags & IRQF_TRIGGER_MASK; + /* refuse to confuse usbcore */ if (dev->dev.dma_mask) { DBG("no we won't dma\n"); @@ -1717,8 +1721,11 @@ sl811h_probe(struct platform_device *dev) * triggers (e.g. most ARM CPUs). Initial driver stress testing * was on a system with single edge triggering, so most sorts of * triggering arrangement should work. + * + * Use resource IRQ flags if set by platform device setup. */ - retval = usb_add_hcd(hcd, irq, IRQF_DISABLED | IRQF_SHARED); + irqflags |= IRQF_SHARED; + retval = usb_add_hcd(hcd, irq, IRQF_DISABLED | irqflags); if (retval != 0) goto err6; |