Re: [PATCH] usb: musb: call pm_runtime_{get,put}_sync before reading vbus registers

[Bin Liu <b-liu@ti.com>]

On Mon, Feb 26, 2018 at 06:19:30PM +0100, Merlijn Wajer wrote:
> Hi,
> 
> On 26/02/18 16:57, Bin Liu wrote:
> > Hi,
> > 
> > On Mon, Feb 26, 2018 at 11:56:08AM +0100, Merlijn Wajer wrote:
> >> Without pm_runtime_{get,put}_sync calls in place, reading
> >> vbus status via /sys causes the following error:
> >>
> >> Unhandled fault: external abort on non-linefetch (0x1028) at 0xfa0ab060
> >> pgd = b333e822
> >> [fa0ab060] *pgd=48011452(bad)
> >>
> >> [<c05261b0>] (musb_default_readb) from [<c0525bd0>] (musb_vbus_show+0x58/0xe4)
> >> [<c0525bd0>] (musb_vbus_show) from [<c04c0148>] (dev_attr_show+0x20/0x44)
> >> [<c04c0148>] (dev_attr_show) from [<c0259f74>] (sysfs_kf_seq_show+0x80/0xdc)
> >> [<c0259f74>] (sysfs_kf_seq_show) from [<c0210bac>] (seq_read+0x250/0x448)
> >> [<c0210bac>] (seq_read) from [<c01edb40>] (__vfs_read+0x1c/0x118)
> >> [<c01edb40>] (__vfs_read) from [<c01edccc>] (vfs_read+0x90/0x144)
> >> [<c01edccc>] (vfs_read) from [<c01ee1d0>] (SyS_read+0x3c/0x74)
> >> [<c01ee1d0>] (SyS_read) from [<c0106fe0>] (ret_fast_syscall+0x0/0x54)
> >>
> >> Solution was suggested by Tony Lindgren <tony@atomide.com>.
> >>
> >> Signed-off-by: Merlijn Wajer <merlijn@wizzup.org>
> >> ---
> >>  drivers/usb/musb/musb_core.c | 4 ++++
> >>  1 file changed, 4 insertions(+)
> >>
> >> diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c
> >> index eef4ad578b31..ceaa40ef0e73 100644
> >> --- a/drivers/usb/musb/musb_core.c
> >> +++ b/drivers/usb/musb/musb_core.c
> >> @@ -1760,6 +1760,8 @@ vbus_show(struct device *dev, struct device_attribute *attr, char *buf)
> >>  	val = musb->a_wait_bcon;
> >>  	vbus = musb_platform_get_vbus_status(musb);
> >>  	if (vbus < 0) {
> >> +		pm_runtime_get_sync(dev);
> >> +
> >>  		/* Use default MUSB method by means of DEVCTL register */
> >>  		devctl = musb_readb(musb->mregs, MUSB_DEVCTL);
> >>  		if ((devctl & MUSB_DEVCTL_VBUS)
> >> @@ -1767,6 +1769,8 @@ vbus_show(struct device *dev, struct device_attribute *attr, char *buf)
> >>  			vbus = 1;
> >>  		else
> >>  			vbus = 0;
> >> +
> >> +		pm_runtime_put_sync(dev);
> >>  	}
> >>  	spin_unlock_irqrestore(&musb->lock, flags);
> > 
> > Thanks for the patch, but I got spinlock deadlock when testing it. I
> > think the function pair should be at the outside scope of the
> > spin_lock/unlock, doesn't it?
> 
> I see - sorry. Weird, it worked for me on my Nokia N900. What did you
> test this on?
> 
> I will try to submit an improved version tonight (you seem to have a
> good idea on what is wrong, so I will follow that), but it would help me
> if I can reproduce the issue.

I tested on both Beaglebone Black and AM335x GP EVM, on the USB1 host
port. I didn't test the USB0 port (yet).

Thanks,
-Bin.