linux-kernel - Re: [PATCH] usb: musb: dsps: implement vbus

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Fri, 9 Oct 2015 16:19:03 -0500
From:	Felipe Balbi <balbi@...com>
To:	Roman Alyautdin <ralyautdin@....rtsoft.ru>,
	Sergei Shtylyov <sergei.shtylyov@...entembedded.com>
CC:	<linux-usb@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] usb: musb: dsps: implement vbus_status method


Hi,

Roman Alyautdin <ralyautdin@....rtsoft.ru> writes:
> On 08/10/15 17:07, Sergei Shtylyov wrote:
>> On 10/8/2015 4:50 PM, Sergei Shtylyov wrote:
>>
>>>> Implement vbus_status  method of musb_platform_ops that allows
>>>> musb_core to properly represent the VBUS status of musb_dsps devices in
>>>> corresponding sysfs entry
>>>>
>>>> Signed-off-by: Roman Alyautdin <ralyautdin@....rtsoft.ru>
>>>> ---
>>>>   drivers/usb/musb/musb_dsps.c |   13 +++++++++++++
>>>>   1 file changed, 13 insertions(+)
>>>>
>>>> diff --git a/drivers/usb/musb/musb_dsps.c 
>>>> b/drivers/usb/musb/musb_dsps.c
>>>> index 84512d1..9c00edf 100644
>>>> --- a/drivers/usb/musb/musb_dsps.c
>>>> +++ b/drivers/usb/musb/musb_dsps.c
>>>> @@ -632,6 +632,18 @@ static void dsps_read_fifo32(struct musb_hw_ep 
>>>> *hw_ep,
>>>> u16 len, u8 *dst)
>>>>       }
>>>>   }
>>>>
>>>> +static int dsps_musb_vbus_status(struct musb *musb)
>>>> +{
>>>> +    void __iomem *mregs = musb->mregs;
>>>> +    u8 devctl;
>>>> +
>>>> +    devctl = dsps_readb(mregs, MUSB_DEVCTL);
>>>> +    if ((devctl & MUSB_DEVCTL_VBUS) == (3 << MUSB_DEVCTL_VBUS_SHIFT))
>>>
>>>     I don't see why this has to be is implemented in the DSPS glue 
>>> layer. The
>>> DevCtl register is a standard MUSB one.
>>
>>    Looking at the musb_platform_get_vbus_status(), it's unclear why it 
>> returns 0 if the vbus_status() method id undefined. I think we should 
>> read the DevCtl registre here, removing the FIXME at the call site.
>
> Maybe it is worth to return -EINVAL from musb_platform_get_vbus_status 
> in case of method is undefined and keep current logic of interface 
> implementation in platform-specific driver.
>
> Hence print "Vbus unknown" in musb_vbus_show() in case of -EINVAL from 
> musb_platform_get_vbus_status

I actually like the idea of reading DevCtl. That's a nice, safe default
afaict. That's one of the information DevCtl is supposed to tell MUSB anyway.

-- 
balbi

Download attachment "signature.asc" of type "application/pgp-signature" (819 bytes)