[PATCH] ASoC: wm_adsp: Fix compilation when debugfs is disabled
Guard the use of component->debufs_root against CONFIG_DEBUG_FS
Reported-by: kernel test robot lkp@intel.com Signed-off-by: Simon Trimmer simont@opensource.cirrus.com --- sound/soc/codecs/wm_adsp.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index f17c749c24c3..03cfaa82b031 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -1028,7 +1028,9 @@ int wm_adsp2_component_probe(struct wm_adsp *dsp, struct snd_soc_component *comp snprintf(preload, ARRAY_SIZE(preload), "%s Preload", dsp->cs_dsp.name); snd_soc_component_disable_pin(component, preload);
+#if IS_ENABLED(CONFIG_DEBUG_FS) cs_dsp_init_debugfs(&dsp->cs_dsp, component->debugfs_root); +#endif
dsp->component = component;
On Wed, Sep 29, 2021 at 03:50:40PM +0100, Simon Trimmer wrote:
+#if IS_ENABLED(CONFIG_DEBUG_FS) cs_dsp_init_debugfs(&dsp->cs_dsp, component->debugfs_root); +#endif
The more common pattern is to declare a static inline stub function instead of the prototype if debugfs is disabled.
On Wednesday, September 29, 2021 3:52 PM, Mark Brown wrote: The more common pattern is to declare a static inline stub function instead of the prototype if debugfs is disabled.
Hi Mark, Yes there is one for the cs_dsp_init_debugfs function though the compilation failure the bot detected is with the parameter dereferencing the disappearing snd_soc_component member.
I could make a utility inline for use in this function that would do the dereference or return NULL based on the state of the config option if you think that would be neater? (For the expense of a new function it would separate the call of the stub from the conditional access of the member based on IS_ENABLED)
Thanks, Simon
On Wed, Sep 29, 2021 at 04:27:19PM +0100, Simon Trimmer wrote:
Yes there is one for the cs_dsp_init_debugfs function though the compilation failure the bot detected is with the parameter dereferencing the disappearing snd_soc_component member.
At that point just remove the conditional definition in the struct, it's not like a single pointer in a struct we are likely to have well under 10 of in the system is going to make an appreciable difference to memory consumption.
participants (2)
-
Mark Brown -
Simon Trimmer