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  PHC 
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Thu, 21 May 2020 12:11:56 -0700
From:   Jakub Kicinski <>
To:     Mark Starovoytov <>
Cc:     "" <>,
        "David S . Miller" <>,
        Dmitry Bezrukov <>,
        Igor Russkikh <>
Subject: Re: [EXT] Re: [PATCH net-next 03/12] net: atlantic: changes for
 multi-TC support

On Thu, 21 May 2020 09:18:44 +0000 Mark Starovoytov wrote:
> Hi Jakub,
> > > In the first generation of our hardware (A1), a whole traffic class is
> > > consumed for PTP handling in FW (FW uses it to send the ptp data and
> > > to send back timestamps).
> > > Since this conflicts with QoS (user is unable to use the reserved
> > > TC2), we suggest using module param to give the user a choice:
> > > disabling PTP allows using all available TCs.  
> > 
> > Is there really no way to get the config automatically chosen when user sets
> > up TCs or does SIOCSHWTSTAMP? It's fine to return -EOPNOTSUPP when too
> > many things are enabled, but user having to set module parameters upfront
> > is quite painful.  
> Module param is not a must have for usage, default config allows the
> user to use TCs and PTP features simultaneously with one major
> limitation: TC2 is reserved for PTP, so if the user tries to
> send/receive anything on TC2, if won't quite work unfortunately. If
> the user wants to get "everything" from QoS/TC (e.g. use all the TCs)
> - he can explicitly disable the PTP via module param.
> Right now we really aren't sure we can dynamically rearrange
> resources between QoS and PTP, since disabling/enabling PTP requires
> a complete HW reconfiguration unfortunately. Even more unfortunate is
> the fact that we can't change the TC, which is reserved for PTP,
> because TC2 is hardcoded in firmware.
> We would prefer to keep things as is for now, if possible. We'll
> discuss this with HW/FW team(s) and submit a follow-up patch, if we
> find a way to automatically choose the config.

Module parameters are very strongly discouraged for networking drivers.
They also constitute uAPI, and can't be changed, short-term solution
like that is really far from ideal.

Powered by blists - more mailing lists