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

From: Merlijn Wajer
Date: Mon Feb 26 2018 - 12:20:11 EST


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@xxxxxxxxxxx>.
>>
>> Signed-off-by: Merlijn Wajer <merlijn@xxxxxxxxxx>
>> ---
>> 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.

Cheers,
Merlijn

Attachment: signature.asc
Description: OpenPGP digital signature