linux-kernel - Re: [PATCH 05/11] soundwire: bus: update multi-link definition with hw sync details

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

Date:   Fri, 28 Aug 2020 12:57:42 +0530
From:   Vinod Koul <vkoul@...nel.org>
To:     Pierre-Louis Bossart <pierre-louis.bossart@...ux.intel.com>
Cc:     Bard Liao <yung-chuan.liao@...ux.intel.com>,
        alsa-devel@...a-project.org, tiwai@...e.de,
        gregkh@...uxfoundation.org, linux-kernel@...r.kernel.org,
        ranjani.sridharan@...ux.intel.com, hui.wang@...onical.com,
        broonie@...nel.org, srinivas.kandagatla@...aro.org,
        jank@...ence.com, mengdong.lin@...el.com, sanyog.r.kale@...el.com,
        rander.wang@...ux.intel.com, bard.liao@...el.com
Subject: Re: [PATCH 05/11] soundwire: bus: update multi-link definition with
 hw sync details

On 26-08-20, 09:09, Pierre-Louis Bossart wrote:
> 
> 
> > > + * @hw_sync_min_links: Number of links used by a stream above which
> > > + * hardware-based synchronization is required. This value is only
> > > + * meaningful if multi_link is set. If set to 1, hardware-based
> > > + * synchronization will be used even if a stream only uses a single
> > > + * SoundWire segment.
> > 
> > Soundwire spec does not say anything about multi-link so this is left to
> > implementer. Assuming that value of 1 would mean hw based sync will
> > be used even for single stream does not make sense in generic terms.
> > Maybe yes for Intel but may not be true for everyone?
> 
> hw-based sync is required for Intel even for single stream. It's been part
> of the recommended programming flows since the beginning but ignored so far.
> 
> That said, this value is set by each master implementation, no one forces
> non-Intel users to implement an Intel-specific requirement.
> 
> > We already use m_rt_count in code for this, so the question is why is
> > that not sufficient?
> 
> Because as you rightly said above, Intel requires the hw_sync to be used
> even for single stream, but we didn't want others to be forced to use the
> hw-sync for single stream. the m_rt_count is not sufficient for Intel.
> 
> I think we are in agreement on not forcing everyone to follow what is
> required by Intel, and that's precisely why we added this setting. If you
> set it to two you would only use hw_sync when two masters are used.

Okay, it would be better if we move it to intel driver, but I see it may
not be trivial, so lets go with this approach.

-- 
~Vinod