Re: [PATCH] [v2] ALSA: hdsp: fix some memleaks in snd_hdsp_hwdep_ioctl

[Takashi Iwai]

On Sun, 07 Jan 2024 09:25:12 +0100,
alexious@xxxxxxxxxx wrote:
> 
>  
> > On Wed, 27 Dec 2023 07:03:20 +0100,
> > Zhipeng Lu wrote:
> > > 
> > > When snd_hdsp_load_firmware_from_cache and snd_hdsp_enable_io fails,
> > > the hdsp->fw_uploaded needs to be free.Or there could be memleaks in
> > > snd_hdsp_hwdep_ioctl.
> > > 
> > > Fixes: 90caaef6a1ce ("ALSA: hdsp: improve firmware caching")
> > > Signed-off-by: Zhipeng Lu <alexious@xxxxxxxxxx>
> > 
> > Thanks for the patch, but the patch description is misleading.
> > The allocated object isn't really "leaked"; it's released at the
> > removal of the driver.
> 
> It is a little confused to me that I can't find a clear "removal" operation for 
> hdsp->fw_uploaded. Could you please give a more detailed explaination, 
> such as call chain, why hdsp->fw_uploaded is not leaked?

It's released in snd_hdsp_card_free().


Takashi