Re: [PATCH 12/12] USB: ehci-omap: Fix detection in HSIC mode

[Roger Quadros <rogerq-l0cyMroinI0@public.gmane.org>]

On 03/12/2013 12:51 PM, kishon wrote:
> Hi,
> 
> On Tuesday 12 March 2013 04:14 PM, Roger Quadros wrote:
>> The HSIC devices need to be kept in reset while the EHCI controller
>> is being initialized and only brought out of reset after the
>> initialization is complete, else HSIC devices will not be detected.
>>
>> Signed-off-by: Roger Quadros <rogerq-l0cyMroinI0@public.gmane.org>
>> CC: Alan Stern <stern-nwvwT67g6+6dFdvTe/nMLpVzexx5G7lz@public.gmane.org>
>> ---
>>   drivers/usb/host/ehci-omap.c |   28 +++++++++++-----------------
>>   1 files changed, 11 insertions(+), 17 deletions(-)
>>
>> diff --git a/drivers/usb/host/ehci-omap.c b/drivers/usb/host/ehci-omap.c
>> index 1ba1df8..52a2d46 100644
>> --- a/drivers/usb/host/ehci-omap.c
>> +++ b/drivers/usb/host/ehci-omap.c
>> @@ -50,6 +50,7 @@
>>   #include <linux/usb/hcd.h>
>>   #include <linux/of.h>
>>   #include <linux/dma-mapping.h>
>> +#include <linux/delay.h>
> 
> not needed here I guess.

Right. will remove.

>>
>>   #include "ehci.h"
>>
>> @@ -90,26 +91,13 @@ static inline u32 ehci_read(void __iomem *base, u32 reg)
>>   static int omap_ehci_init(struct usb_hcd *hcd)
>>   {
>>       struct ehci_hcd    *ehci = hcd_to_ehci(hcd);
>> -    struct omap_hcd    *omap = (struct omap_hcd *)ehci->priv;
>> -    int rc, i;
>> -
>> -    /* Hold PHYs in reset while initializing EHCI controller */
>> -    for (i = 0; i < omap->nports; i++) {
>> -        if (omap->phy[i])
>> -            usb_phy_shutdown(omap->phy[i]);
>> -    }
>> +    int rc;
>>
>>       /* we know this is the memory we want, no need to ioremap again */
>>       ehci->caps = hcd->regs;
>>
>>       rc = ehci_setup(hcd);
>>
>> -    /* Bring PHYs out of reset */
>> -    for (i = 0; i < omap->nports; i++) {
>> -        if (omap->phy[i])
>> -            usb_phy_init(omap->phy[i]);
>> -    }
>> -
>>       return rc;
>>   }
>>
>> @@ -219,9 +207,6 @@ static int ehci_hcd_omap_probe(struct platform_device *pdev)
>>           }
>>
>>           omap->phy[i] = phy;
>> -        usb_phy_init(omap->phy[i]);
>> -        /* bring PHY out of suspend */
>> -        usb_phy_set_suspend(omap->phy[i], 0);
>>       }
>>
>>       pm_runtime_enable(dev);
>> @@ -245,6 +230,15 @@ static int ehci_hcd_omap_probe(struct platform_device *pdev)
>>           goto err_pm_runtime;
>>       }
>>
>> +    /* Bring PHYs out of reset */
>> +    for (i = 0; i < omap->nports; i++) {
>> +        if (!omap->phy[i])
>> +            continue;
>> +
>> +        usb_phy_init(omap->phy[i]);
>> +        /* bring PHY out of suspend */
>> +        usb_phy_set_suspend(omap->phy[i], 0);
>> +    }
> 
> But isn't HSIC supposed to be PHYless devices. Maybe a comment explaining it might be helpful.
> 

Yes HSIC is PHYless but they do have reset line between the chips.
I'll put a comment.

cheers,
-roger