linux-kernel - Re: [PATCH] USB: DWC2: Add VBUS overcurrent detection control.

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:   Tue, 15 Jun 2021 11:06:17 +0100
From:   Paul Cercueil <paul@...ndingux.net>
To:     Greg KH <gregkh@...uxfoundation.org>
Cc:     周琰杰 <zhouyanjie@...yeetech.com>,
        hminas@...opsys.com, paul@...pouillou.net,
        linux-mips@...r.kernel.org, linux-usb@...r.kernel.org,
        linux-kernel@...r.kernel.org, dongsheng.qiu@...enic.com,
        aric.pzqi@...enic.com, sernia.zhou@...mail.com,
        Dragan Čečavac <dragancecavac@...oo.com>
Subject: Re: [PATCH] USB: DWC2: Add VBUS overcurrent detection control.

Hi Greg,

Le mar., juin 15 2021 at 11:48:30 +0200, Greg KH 
<gregkh@...uxfoundation.org> a écrit :
> On Tue, Jun 15, 2021 at 09:52:20AM +0100, Paul Cercueil wrote:
>>  Hi Zhou,
>> 
>>  Le mar., juin 15 2021 at 16:16:39 +0800, 周琰杰 
>> <zhouyanjie@...yeetech.com>
>>  a écrit :
>>  > Hi Greg,
>>  >
>>  > Sorry for taking so long to reply.
>>  >
>>  > 于 Tue, 23 Mar 2021 16:31:29 +0100
>>  > Greg KH <gregkh@...uxfoundation.org> 写道:
>>  >
>>  > >  On Tue, Mar 23, 2021 at 11:24:26PM +0800, 周琰杰 (Zhou 
>> Yanjie)
>>  > > wrote:
>>  > >  > Introduce configurable option for enabling GOTGCTL register
>>  > >  > bits VbvalidOvEn and VbvalidOvVal. Once selected it disables
>>  > >  > VBUS overcurrent detection.
>>  > >  >
>>  > >  > This patch is derived from Dragan Čečavac (in the kernel 
>> 3.18
>>  > >  > tree of CI20). It is very useful for the MIPS Creator 
>> CI20(r1).
>>  > >  > Without this patch, CI20's OTG port has a great probability 
>> to
>>  > >  > face overcurrent warning, which breaks the OTG functionality.
>>  > >  >
>>  > >  > Signed-off-by: 周琰杰 (Zhou Yanjie) 
>> <zhouyanjie@...yeetech.com>
>>  > >  > Signed-off-by: Dragan Čečavac <dragancecavac@...oo.com>
>>  > >  > ---
>>  > >  >  drivers/usb/dwc2/Kconfig | 6 ++++++
>>  > >  >  drivers/usb/dwc2/core.c  | 9 +++++++++
>>  > >  >  2 files changed, 15 insertions(+)
>>  > >  >
>>  > >  > diff --git a/drivers/usb/dwc2/Kconfig 
>> b/drivers/usb/dwc2/Kconfig
>>  > >  > index c131719..e40d187 100644
>>  > >  > --- a/drivers/usb/dwc2/Kconfig
>>  > >  > +++ b/drivers/usb/dwc2/Kconfig
>>  > >  > @@ -94,4 +94,10 @@ config USB_DWC2_DEBUG_PERIODIC
>>  > >  >  	  non-periodic transfers, but of course the debug logs
>>  > >  > will be incomplete. Note that this also disables some debug
>>  > > messages
>>  > >  >  	  for which the transfer type cannot be deduced.
>>  > >  > +
>>  > >  > +config USB_DWC2_DISABLE_VOD
>>  > >  > +	bool "Disable VBUS overcurrent detection"
>>  > >  > +	help
>>  > >  > +	  Say Y here to switch off VBUS overcurrent detection. It
>>  > >  > enables USB
>>  > >  > +	  functionality blocked by overcurrent detection.
>>  > >
>>  > >  Why would this be a configuration option?  Shouldn't this be 
>> dynamic
>>  > >  and just work properly automatically?
>>  > >
>>  > >  You should not have to do this on a build-time basis, it 
>> should be
>>  > >  able to be detected and handled properly at run-time for all
>>  > > devices.
>>  > >
>>  >
>>  > I consulted the original author Dragan Čečavac, he think since 
>> this is
>>  > a feature which disables overcurrent detection, so we are not 
>> sure if
>>  > it could be harmful for some devices. Therefore he advise against
>>  > enabling it in runtime, and in favor that user explicitely has to
>>  > enable it.
>> 
>>  This could still be enabled at runtime, though, via a module 
>> parameter.
>>  Leave it enabled by default, and those who want to disable it can 
>> do it.
> 
> This is not the 1990's, please NEVER add new module parameters,

First time I hear this.

> especially ones that are somehow supposed to be device-specific.
> 
> Remember, module options are code-wide, not device-specific.

Right. I thought "just make the option available on devices that 
support it" but that's not how it works.

-Paul

> Just do this based on the device type, or something else dynamic, do 
> NOT
> make this be forced to be selected by a kernel configuration option 
> or a
> random module option at runtime.
> 
> thanks,
> 
> greg k-h