Re: [PATCH 2/4] usb: dwc3: dwc3-omap: return -EPROBE_DEFER if probehas not yet executed

From: Felipe Balbi
Date: Tue Mar 05 2013 - 10:22:40 EST


On Tue, Mar 05, 2013 at 08:48:24PM +0530, kishon wrote:
> Hi,
>
> On Tuesday 05 March 2013 08:36 PM, Felipe Balbi wrote:
> >On Tue, Mar 05, 2013 at 08:31:58PM +0530, kishon wrote:
> >>Hi,
> >>
> >>On Tuesday 05 March 2013 08:26 PM, Felipe Balbi wrote:
> >>>On Tue, Mar 05, 2013 at 07:51:58PM +0530, Kishon Vijay Abraham I wrote:
> >>>>return -EPROBE_DEFER from dwc3_omap_mailbox in dwc3-omap.c, if the probe of
> >>>>dwc3-omap has not yet been executed or failed.
> >>>>
> >>>>Signed-off-by: Kishon Vijay Abraham I <kishon@xxxxxx>
> >>>>---
> >>>> drivers/usb/dwc3/dwc3-omap.c | 7 +++++--
> >>>> include/linux/usb/dwc3-omap.h | 6 +++---
> >>>> 2 files changed, 8 insertions(+), 5 deletions(-)
> >>>>
> >>>>diff --git a/drivers/usb/dwc3/dwc3-omap.c b/drivers/usb/dwc3/dwc3-omap.c
> >>>>index 19c6e72..9428f4e 100644
> >>>>--- a/drivers/usb/dwc3/dwc3-omap.c
> >>>>+++ b/drivers/usb/dwc3/dwc3-omap.c
> >>>>@@ -138,11 +138,14 @@ static inline void dwc3_omap_writel(void __iomem *base, u32 offset, u32 value)
> >>>> writel(value, base + offset);
> >>>> }
> >>>>
> >>>>-void dwc3_omap_mailbox(enum omap_dwc3_vbus_id_status status)
> >>>>+int dwc3_omap_mailbox(enum omap_dwc3_vbus_id_status status)
> >>>> {
> >>>> u32 val;
> >>>> struct dwc3_omap *omap = _omap;
> >>>>
> >>>>+ if (!omap)
> >>>>+ return -EPROBE_DEFER;
> >>>>+
> >>>> switch (status) {
> >>>> case OMAP_DWC3_ID_GROUND:
> >>>> dev_dbg(omap->dev, "ID GND\n");
> >>>>@@ -185,7 +188,7 @@ void dwc3_omap_mailbox(enum omap_dwc3_vbus_id_status status)
> >>>> dev_dbg(omap->dev, "ID float\n");
> >>>> }
> >>>>
> >>>>- return;
> >>>>+ return IRQ_HANDLED;
> >>>
> >>>IRQ_HANDLED ????
> >>
> >>Actually I did it that way since palmas_vbus_wakeup_irq can directly
> >>return the return value from dwc3_omap_mailbox. If this seems hacky
> >>to you, I'll change it.
> >
> >it does seem hacky :-) Try something like:
> >
> >ret = dwc3_omap_mailbox();
> >if (ret)
> > print_error();
> >
> >return IRQ_HANDLED;
>
> hmm.. But there is one case where palmas_vbus_wakeup_irq should
> return EPROBE_DEFER. In the cold plug case, if palmas gets loaded
> before dwc3-omap, omap in dwc3_omap_mailbox will be NULL. So ideally
> palmas should be probed after dwc3-omap for this case. Returning
> IRQ_HANDLED or IRQ_NONE wont help here.

you can return IRQ_HANDLED but don't clear the IRQ status bits, which
will make the IRQ retrigger, right ?

If you return -EPROBE_DEFER from within the IRQ handler, what will
happen ? IRQ subsystem only understands IRQ_NONE, IRQ_WAKE_THREAD, and
IRQ_HANDLED. How do you suppose it will treat -EPROBE_DEFER ?

--
balbi

Attachment: signature.asc
Description: Digital signature