Re: [Qemu-devel] [PATCH 2/2] audio: add -audiodev pa, in|out.latency= to documentation

["Zoltán Kővágó" <dirty.ice.hu@gmail.com>]

On 2019-09-11 16:58, Stefan Hajnoczi wrote:
> The "latency" parameter wasn't covered by the documentation.
> 
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Zoltán Kővágó <DirtY.iCE.hu@gmail.com>
> ---
> How is this parameter related to buffer-length?

Pulseaudio being a client-server architecture is a bit different than 
the other backends, plus it also has to mix multiple streams. 
buffer-length corresponds to the buffer inside qemu, while latency 
corresponds to pulseaudio.  For playback, the latency should be "maximum 
latency that the application can deal with", if a different client 
request a lower latency, our latency will decrease too.  It's up to the 
server to figure out an optimal buffer size on the server side of the 
things.
For recording it's the size of the buffer we will read at a time from 
pulseaudio.

> ---
>   qemu-options.hx | 4 ++++
>   1 file changed, 4 insertions(+)
> 
> diff --git a/qemu-options.hx b/qemu-options.hx
> index a4f9f74f52..82154cecf8 100644
> --- a/qemu-options.hx
> +++ b/qemu-options.hx
> @@ -470,6 +470,7 @@ DEF("audiodev", HAS_ARG, QEMU_OPTION_audiodev,
>       "-audiodev pa,id=id[,prop[=value][,...]]\n"
>       "                server= PulseAudio server address\n"
>       "                in|out.name= source/sink device name\n"
> +    "                in|out.latency= desired latency in microseconds\n"
>   #endif
>   #ifdef CONFIG_AUDIO_SDL
>       "-audiodev sdl,id=id[,prop[=value][,...]]\n"
> @@ -630,6 +631,9 @@ Sets the PulseAudio @var{server} to connect to.
>   @item in|out.name=@var{sink}
>   Use the specified source/sink for recording/playback.
>   
> +@item in|out.latency=@var{usecs}
> +Desired latency in microseconds.
> +
>   @end table
>   
>   @item -audiodev sdl,id=@var{id}[,@var{prop}[=@var{value}][,...]]
>