linux-kernel - Re: [PATCH 3/4] usb: dwc3: core: add a core init flag for device mode resume

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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20240813235642.zaxom3jbuomfwghf@synopsys.com>
Date: Tue, 13 Aug 2024 23:56:48 +0000
From: Thinh Nguyen <Thinh.Nguyen@...opsys.com>
To: Frank Li <Frank.li@....com>
CC: Thinh Nguyen <Thinh.Nguyen@...opsys.com>,
        Mathias Nyman <mathias.nyman@...el.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Shawn Guo <shawnguo@...nel.org>, Sascha Hauer <s.hauer@...gutronix.de>,
        Pengutronix Kernel Team <kernel@...gutronix.de>,
        Fabio Estevam <festevam@...il.com>,
        "linux-usb@...r.kernel.org" <linux-usb@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "imx@...ts.linux.dev" <imx@...ts.linux.dev>,
        "linux-arm-kernel@...ts.infradead.org" <linux-arm-kernel@...ts.infradead.org>,
        "jun.li@....com" <jun.li@....com>, Peter Chen <peter.chen@....com>
Subject: Re: [PATCH 3/4] usb: dwc3: core: add a core init flag for device mode
 resume

On Tue, Aug 13, 2024, Frank Li wrote:
> On Fri, Aug 09, 2024 at 11:18:53PM +0000, Thinh Nguyen wrote:
> > On Fri, Aug 09, 2024, Frank Li wrote:
> > > On Wed, Aug 07, 2024 at 01:13:52AM +0000, Thinh Nguyen wrote:
> > > > On Fri, Jul 12, 2024, Frank Li wrote:
> > > > > From: Li Jun <jun.li@....com>
> > > > >
> > > > > The runtime resume will happen before system resume if system wakeup by
> > > > > device mode wakeup event(e.g, VBUS). Add a flag to avoid init twice.
> > > >
> > > > What's the consequence of going through the resuming sequence twice?
> > > > Will this cause any functional issue?
> > >
> > > static int dwc3_core_init_for_resume(struct dwc3 *dwc)
> > > {
> > >
> > > 	...
> > >         ret = dwc3_clk_enable(dwc);
> > >         if (ret)
> > >                 goto assert_reset;
> > >
> > > 	...
> > > }
> > >
> > > clk will be enabled twice, the reference count in clk will be wrong because
> > > clk_disable() only once at suspend.
> >
> > Please capture these info in the commit message.
> >
> > >
> > > So clk will be always ON at next suspend.
> > >
> > > Frank Li
> > >
> >
> > dwc3_clk_enable() happens in probe() and dwc3_core_init_for_resume(),
> > but you're checking dwc->core_inited in dwc3_core_init(). Why?
> 
> I have not check dwc->core_inited in dwc3_core_init(). Just set init value
> dwc->core_inited = true
> 
> Do you means dwc3_core_init() as dwc3_resume_common()?
> 

You set the dwc->core_inited in dwc3_core_init(). However, you use this
flag to also check if dwc3_clk_enable() is done. dwc3_clk_enable() is
not done in dwc3_core_init().

> >
> > If the issue is only related to clk_enable(), perhaps just check that?
> 
> This should be major one. Any paired operator between suspend/resume should
> have issue.
> 

I don't know what else could be a problem in dwc3_core_init(). As far as
I know, it's only dwc3_clk_enable().

I questioned because of the mismatch use of the check, where the issue
you mentioned is outside of dwc3_core_init(). Perhaps the flag should
not be set there. If that's the case, rename the flag to match the check
condition (if necessary).

> 
> >
> > (minor nit: rename core_inited to core_initialized if you plan to do
> > that)
> >
> > Thanks,
> > Thinh
> >
> > > > >
> > > > > Reviewed-by: Peter Chen <peter.chen@....com>
> > > > > Signed-off-by: Li Jun <jun.li@....com>
> > > > > Signed-off-by: Frank Li <Frank.Li@....com>
> > > > > ---
> > > > >  drivers/usb/dwc3/core.c | 13 +++++++++++++
> > > > >  drivers/usb/dwc3/core.h |  1 +
> > > > >  2 files changed, 14 insertions(+)
> > > > >
> > > > > diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c
> > > > > index 734de2a8bd212..d60917fad8c52 100644
> > > > > --- a/drivers/usb/dwc3/core.c
> > > > > +++ b/drivers/usb/dwc3/core.c
> > > > > @@ -950,6 +950,8 @@ static void dwc3_core_exit(struct dwc3 *dwc)
> > > > >  	dwc3_phy_exit(dwc);
> > > > >  	dwc3_clk_disable(dwc);
> > > > >  	reset_control_assert(dwc->reset);
> > > > > +
> > > > > +	dwc->core_inited = false;
> > > > >  }
> > > > >
> > > > >  static bool dwc3_core_is_valid(struct dwc3 *dwc)
> > > > > @@ -1446,6 +1448,8 @@ static int dwc3_core_init(struct dwc3 *dwc)
> > > > >  		dwc3_writel(dwc->regs, DWC3_LLUCTL, reg);
> > > > >  	}
> > > > >
> > > > > +	dwc->core_inited = true;
> > > > > +
> > > > >  	return 0;
> > > > >
> > > > >  err_power_off_phy:
> > > > > @@ -2375,6 +2379,15 @@ static int dwc3_resume_common(struct dwc3 *dwc, pm_message_t msg)
> > > > >
> > > > >  	switch (dwc->current_dr_role) {
> > > > >  	case DWC3_GCTL_PRTCAP_DEVICE:
> > > > > +		/*
> > > > > +		 * system resume may come after runtime resume
> > > > > +		 * e.g. rpm suspend -> pm suspend -> wakeup
> > > > > +		 * -> rpm resume -> system resume, so if already
> > > > > +		 *  runtime resumed, system resume should skip it.
> > > > > +		 */
> > > > > +		if (dwc->core_inited)
> > > > > +			break;
> > > > > +
> > > > >  		ret = dwc3_core_init_for_resume(dwc);
> > > > >  		if (ret)
> > > > >  			return ret;
> > > > > diff --git a/drivers/usb/dwc3/core.h b/drivers/usb/dwc3/core.h
> > > > > index 1e561fd8b86e2..8a4bfd9a25b19 100644
> > > > > --- a/drivers/usb/dwc3/core.h
> > > > > +++ b/drivers/usb/dwc3/core.h
> > > > > @@ -1195,6 +1195,7 @@ struct dwc3 {
> > > > >  	struct clk		*utmi_clk;
> > > > >  	struct clk		*pipe_clk;
> > > > >
> > > > > +	bool			core_inited;

Document new member of the struct.

> > > > >  	struct reset_control	*reset;
> > > > >
> > > > >  	struct usb_phy		*usb2_phy;
> > > > >
> > > > > --
> > > > > 2.34.1
> > > > >

Thanks,
Thinh