linux-kernel - Re: [PATCH v24 16/34] ASoC: qcom: qdsp6: q6afe: Increase APR timeout

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: <bcf2c413-cb20-41eb-9993-db21c019b70d@quicinc.com>
Date: Thu, 1 Aug 2024 16:04:57 -0700
From: Wesley Cheng <quic_wcheng@...cinc.com>
To: Pierre-Louis Bossart <pierre-louis.bossart@...ux.intel.com>,
        <srinivas.kandagatla@...aro.org>, <mathias.nyman@...el.com>,
        <perex@...ex.cz>, <conor+dt@...nel.org>, <corbet@....net>,
        <broonie@...nel.org>, <lgirdwood@...il.com>, <krzk+dt@...nel.org>,
        <Thinh.Nguyen@...opsys.com>, <bgoswami@...cinc.com>, <tiwai@...e.com>,
        <gregkh@...uxfoundation.org>, <robh@...nel.org>
CC: <linux-kernel@...r.kernel.org>, <devicetree@...r.kernel.org>,
        <linux-sound@...r.kernel.org>, <linux-usb@...r.kernel.org>,
        <linux-arm-msm@...r.kernel.org>, <linux-doc@...r.kernel.org>,
        <alsa-devel@...a-project.org>
Subject: Re: [PATCH v24 16/34] ASoC: qcom: qdsp6: q6afe: Increase APR timeout

Hi Pierre,

On 8/1/2024 1:33 AM, Pierre-Louis Bossart wrote:
>
> On 8/1/24 03:17, Wesley Cheng wrote:
>> For USB offloading situations, the AFE port start command will result in a
>> QMI handshake between the Q6DSP and the main processor.  Depending on if
>> the USB bus is suspended, this routine would require more time to complete,
>> as resuming the USB bus has some overhead associated with it.  Increase the
>> timeout to 3s to allow for sufficient time for the USB QMI stream enable
>> handshake to complete.
>>
>> Reviewed-by: Srinivas Kandagatla <srinivas.kandagatla@...aro.org>
>> Signed-off-by: Wesley Cheng <quic_wcheng@...cinc.com>
>> ---
>>  sound/soc/qcom/qdsp6/q6afe.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/sound/soc/qcom/qdsp6/q6afe.c b/sound/soc/qcom/qdsp6/q6afe.c
>> index 948007955934..452d7e8b9b30 100644
>> --- a/sound/soc/qcom/qdsp6/q6afe.c
>> +++ b/sound/soc/qcom/qdsp6/q6afe.c
>> @@ -366,7 +366,7 @@
>>  #define AFE_API_VERSION_SLOT_MAPPING_CONFIG	1
>>  #define AFE_API_VERSION_CODEC_DMA_CONFIG	1
>>  
>> -#define TIMEOUT_MS 1000
>> +#define TIMEOUT_MS 3000
> Surprising that 1s is not enough to resume a bus, and conversely is 3s
> enough then if the overhead is so significant?
> It looks like either too much or too little...

To clarify, if USB bus suspend isn't enabled, then this isn't an issue.  I just picked 3s since this is a worst case timeout value.  I don't think it will take longer than that to start an audio session, otherwise there is another inherent problem.  When I initially tested this, it may have been on a platform/chipset that wasn't running with optimized CPU settings, and might have been running slower overall.

Thanks

Wesley Cheng

>>  #define AFE_CMD_RESP_AVAIL	0
>>  #define AFE_CMD_RESP_NONE	1
>>  #define AFE_CLK_TOKEN		1024