linux-kernel - Re: [PATCH v4 1/2] Partially revert ASoC: qcom: Fix enabling BCLK and LRCLK in LPAIF invalid state

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]

Message-ID: <966993b7-4720-bdd2-cf4d-cf5a7c11a0c1@codeaurora.org>
Date:   Tue, 1 Dec 2020 23:01:21 +0530
From:   Srinivasa Rao Mandadapu <srivasam@...eaurora.org>
To:     Mark Brown <broonie@...nel.org>
Cc:     agross@...nel.org, bjorn.andersson@...aro.org, lgirdwood@...il.com,
        robh+dt@...nel.org, plai@...eaurora.org, bgoswami@...eaurora.org,
        perex@...ex.cz, tiwai@...e.com, srinivas.kandagatla@...aro.org,
        rohitkr@...eaurora.org, linux-arm-msm@...r.kernel.org,
        alsa-devel@...a-project.org, devicetree@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        V Sujith Kumar Reddy <vsujithk@...eaurora.org>
Subject: Re: [PATCH v4 1/2] Partially revert ASoC: qcom: Fix enabling BCLK and
 LRCLK in LPAIF invalid state

Thanks Mark for your time!!!

On 11/30/2020 6:16 PM, Mark Brown wrote:
> On Sat, Nov 28, 2020 at 10:29:18AM +0530, Srinivasa Rao Mandadapu wrote:
>> This reverts part of commit b1824968221c
>> ("ASoC: qcom: Fix enabling BCLK and LRCLK in LPAIF invalid state")
>>
>> To identify LPAIF invalid state after device suspend and resume,
>> made I2S and DMA control registers not volatile, which is not necessary.
>> Instead invalid reg state can be handled with regcache APIs.
>> The BCLK ref count is necessary to enable clock only it's in disable state.
> Part of this commit message says that the problem was making the registers
> non-volatile but both the change and the rest of the commit message say
> that the issue was that the registers were made volatile.  I'm also
> still unclear as to what the issue is either way - how does reading the
> state of the registers from the hardware instead of the cache affect
> things?

Initial problem was, during playback if device suspended, I2S and DMA 
control registers

are getting reset and unable to recover playback after resume.

As these registers were non volatile registers, driver is not getting 
actual register value

and unable to report error state to application. Due to this application

keeps on polling for HW current pointer state and not exited from PCM 
running state.

To handle this scenario I made registers volatile and if they are in 
reset state, reported error

to application(commit b1824968221c).

Later from review comments by Srinivas kandagatla, I got to know

about regcache sync APIs, which can be used  to sync cache after resume and

HW registers can be updated with  original values. With that playback 
can be continued.

So is the reason, I am reverting partial changes in the commit b1824968221c.


> Please submit patches using subject lines reflecting the style for the
> subsystem, this makes it easier for people to identify relevant patches.
> Look at what existing commits in the area you're changing are doing and
> make sure your subject lines visually resemble what they're doing.
> There's no need to resubmit to fix this alone.

-- 
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc.,
is a member of Code Aurora Forum, a Linux Foundation Collaborative Project.