summaryrefslogtreecommitdiff
path: root/tools/lib/python
diff options
context:
space:
mode:
authorTakashi Iwai <tiwai@suse.de>2026-10-02 17:04:19 +0200
committerTakashi Iwai <tiwai@suse.de>2026-10-03 11:09:54 +0200
commit9f6f4e97fd53908aecb0979507d9ff10a22288eb (patch)
tree564575c23543fe33e744bec37dded24961c5be0c /tools/lib/python
parentcb8e306fd99c70cc6f57d1a3ec39659dd4c5a6e6 (diff)
ALSA: pcm: Optimize cpu_latency_qos_*() calls
We have a few cpu_latency_qos_*() calls in PCM callbacks, but they seem suboptimal in a few aspects: - We call them at each time when the hw_params changed, but explicitly do remove and (re-)add; this could be cpu_latency_qos_update_request(), instead - The setup of cpu_latency_qos_*() doesn't have to be in hw_params at all; it's not the place where the actual parameters are set up, but rather better at the prepare callback. - The hw_free calls cpu_latency_qos_remove_request() unconditionally, which may be poking a WARNING when it wasn't queued. For addressing the issues above, this patch cleans up and reorganize the cpu_latency_qos_*() calls in the PCM core by a couple helper functions. Now, they are set up at prepare callback, and cleared at hw_free and release callbacks via the new helpers. Link: https://patch.msgid.link/20261002150420.879056-1-tiwai@suse.de Signed-off-by: Takashi Iwai <tiwai@suse.de>
Diffstat (limited to 'tools/lib/python')
0 files changed, 0 insertions, 0 deletions