linux-kernel - Re: [PATCH v3 2/4] platform/chrome: cros_ec_typec: Use Thunderbolt 3 cable discover mode VDO in USB4 mode

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

Message-ID: <20201120120700.GD4160865@google.com>
Date:   Fri, 20 Nov 2020 04:07:00 -0800
From:   Prashant Malani <pmalani@...omium.org>
To:     Heikki Krogerus <heikki.krogerus@...ux.intel.com>
Cc:     Utkarsh Patel <utkarsh.h.patel@...el.com>,
        linux-kernel@...r.kernel.org, linux-usb@...r.kernel.org,
        enric.balletbo@...labora.com, rajmohan.mani@...el.com,
        azhar.shaikh@...el.com
Subject: Re: [PATCH v3 2/4] platform/chrome: cros_ec_typec: Use Thunderbolt 3
 cable discover mode VDO in USB4 mode

On Fri, Nov 20, 2020 at 01:22:18PM +0200, Heikki Krogerus wrote:
> On Thu, Nov 19, 2020 at 12:09:06AM -0800, Prashant Malani wrote:
> > Hi Utkarsh,
> > 
> > On Wed, Nov 18, 2020 at 10:32:09PM -0800, Utkarsh Patel wrote:
> > > Configure Thunderbolt 3 cable generation value by filling Thunderbolt 3
> > > cable discover mode VDO to support rounded Thunderbolt 3 cables.
> > > While we are here use Thunderbolt 3 cable discover mode VDO to fill active
> > > cable plug link training value.
> > > 
> > > Suggested-by: Heikki Krogerus <heikki.krogerus@...ux.intel.com>
> > > Signed-off-by: Utkarsh Patel <utkarsh.h.patel@...el.com>
> > > 
> > > --
> > > Changes in v3:
> > > - Added a check for cable's TBT support before filling TBT3 discover mode
> > >   VDO.
> > > 
> > > Changes in v2:
> > > - No change.
> > > --
> > > ---
> > >  drivers/platform/chrome/cros_ec_typec.c | 14 ++++++++++++--
> > >  1 file changed, 12 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/drivers/platform/chrome/cros_ec_typec.c b/drivers/platform/chrome/cros_ec_typec.c
> > > index 8111ed1fc574..68b17ee1d1ae 100644
> > > --- a/drivers/platform/chrome/cros_ec_typec.c
> > > +++ b/drivers/platform/chrome/cros_ec_typec.c
> > > @@ -514,8 +514,18 @@ static int cros_typec_enable_usb4(struct cros_typec_data *typec,
> > >  	else if (pd_ctrl->control_flags & USB_PD_CTRL_ACTIVE_CABLE)
> > >  		data.eudo |= EUDO_CABLE_TYPE_RE_TIMER << EUDO_CABLE_TYPE_SHIFT;
> > >  
> > > -	data.active_link_training = !!(pd_ctrl->control_flags &
> > > -				       USB_PD_CTRL_ACTIVE_LINK_UNIDIR);
> > > +	/*
> > > +	 * Filling TBT3 Cable VDO when TBT3 cable is being used to establish
> > > +	 * USB4 connection.
> > > +	 */
> > > +	if (pd_ctrl->cable_gen) {
> > > +		data.tbt_cable_vdo = TBT_MODE;
> > > +
> > > +		if (pd_ctrl->control_flags & USB_PD_CTRL_ACTIVE_LINK_UNIDIR)
> > > +			data.tbt_cable_vdo |= TBT_CABLE_LINK_TRAINING;
> > > +
> > > +		data.tbt_cable_vdo |= TBT_SET_CABLE_ROUNDED(pd_ctrl->cable_gen);
> > > +	}
> > 
> > I think the following would decouple Rounded Support and Active Cable Link
> > Training?:
> > 
> > 	struct typec_thunderbolt_data data = {};
> > ...
> > 	if (pd_ctrl->control_flags & USB_PD_CTRL_ACTIVE_LINK_UNIDIR)
> > 		data.tbt_cable_vdo |= TBT_CABLE_LINK_TRAINING;
> > 
> > 	data.tbt_cable_vdo |= TBT_SET_CABLE_ROUNDED(pd_ctrl->cable_gen);
> 
> I agree with this. We should not modify the data that we actually
> have access to.
> 
> > 	if (data.tbt_cable_vdo)
> > 		data.tbt_cable_vdo |= TBT_MODE;
> 
> That is wrong. If the LSRX communication is bi-directional, and/or the
> data rates are non-rounded, the cable has to be TBT3 cable. So I think
> what you would want is:

Thanks for pointing this out, I didn't consider this case.

> 
> 	if (!data.tbt_cable_vdo)
>  		data.tbt_cable_vdo = TBT_MODE;
> 
> But of course we can not do that either, because we have to set the
> TBT_MODE bit if there is any other data in tbt_cable_vdo (USB Type-C
> spec does not define any other valid values except 0x0001 = TBT_MODE
> for that field). Otherwise the mux driver should consider the data
> corrupted. So we would have to do this:
> 
>         if (pd_ctrl->cable_gen &&
>             pd_ctrl->control_flags & USB_PD_CTRL_ACTIVE_LINK_UNIDIR)
>                 data.tbt_cable_vdo = 0; /* We assume USB4 cable */
>         else
>  		data.tbt_cable_vdo |= TBT_MODE; /* It is for sure TBT3 cable */
> 
> But I would not do that. TBT3 cable can also support unidirectional
> SBU communication and rounded data rates.
> 
> IMO safer bet for now would be to just claim that the cable is always
> TBT3 cable until we have access to information that can really tell us
> is the cable USB4 or TBT3.

Which brings us back to v1 of the patch :S

That still leaves my underlying concern that we'll be telling the Mux
implementation that a TBT3 cable is connected when in fact it's a USB4
active cable.

How about we don't set the TBT_MODE bit at all ? IMO it's equally bad as setting it
always, but with the additional advantage:

- USB4 active cable case : you are covered (since if we unilaterally set
TBT_MODE then the Active USB4 cable case never gets executed in
pmc_usb_mux_usb4() in drivers/usb/typec/mux/intel_pmc_mux.c Patch 3/4)

- Bidirectional LSRX non-rounded TBT: Still supported since
  the code path in the Intel Mux agent is the same.

I understand neither of the options are ideal, but WDYT?

BR,

-Prashant