From d42d28176e1740c03d36d2953272e70918083b3a Mon Sep 17 00:00:00 2001 From: Marc Herbert Date: Wed, 2 Dec 2020 10:40:43 -0800 Subject: [PATCH] multiple-pipeline: remove initial sleep and clarify second sleep Having the same -w parameter for both this initial sleep and the test duration made no sense: fixed by removing the first sleep. It's not clear what the purpose of this first sleep was, it has been in multiple-pipeline-capture/playback.sh since the dawn of time (= commit 39cd964ddc776c) For sure this first sleep was helping hide "uninterruptible state" bug #472 by giving the DSP more time to wake up! Signed-off-by: Marc Herbert --- test-case/multiple-pipeline.sh | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/test-case/multiple-pipeline.sh b/test-case/multiple-pipeline.sh index 445264eb..6ed7b7b0 100755 --- a/test-case/multiple-pipeline.sh +++ b/test-case/multiple-pipeline.sh @@ -47,7 +47,7 @@ OPT_NAME['f']='first' OPT_DESC['f']='Fill either playback (p) or capture (c) first or any (a) for all pipelines' OPT_HAS_ARG['f']=1 OPT_VAL['f']='p' -OPT_NAME['w']='wait' OPT_DESC['w']='perpare wait time by sleep' +OPT_NAME['w']='wait' OPT_DESC['w']='duration of one (sub)test iteration' OPT_HAS_ARG['w']=1 OPT_VAL['w']=5 OPT_NAME['r']='random' OPT_DESC['r']='random load pipeline' @@ -163,13 +163,10 @@ do die "Wrong -f argument $f_arg, see -h" esac - dlogi "sleep ${OPT_VAL['w']}s for sound device wakeup" - sleep ${OPT_VAL['w']} - dlogi "checking pipeline status" ps_checks - dlogi "preparing sleep ${OPT_VAL['w']}" + dlogi "Letting playback/capture run for ${OPT_VAL['w']}s" sleep ${OPT_VAL['w']} # check processes again