summaryrefslogtreecommitdiff
path: root/tools/perf/scripts/python
diff options
context:
space:
mode:
authorUday Khare <udaykhare77@gmail.com>2026-07-20 16:09:50 +0530
committerMark Brown <broonie@kernel.org>2026-07-20 15:49:34 +0100
commit317e21532e6ffa1de026bdbce5ba98e1b70ca5c6 (patch)
tree679a34cea5919ba53d134d8f9743aeb987daf8fb /tools/perf/scripts/python
parenta9058504f11150308c44f6c0683eb04c822b821d (diff)
ASoC: max98095: fix missing IS_ERR() before PTR_ERR() on mclk lookup
In max98095_probe(), the -EPROBE_DEFER check after devm_clk_get() is broken due to a missing IS_ERR() guard. The code intends to return -EPROBE_DEFER only when the clock lookup fails with that specific error. However, without IS_ERR() the check: if (PTR_ERR(max98095->mclk) == -EPROBE_DEFER) is called unconditionally, including when devm_clk_get() succeeds and returns a valid pointer. Calling PTR_ERR() on a valid pointer reinterprets its address as a signed long; the result is arbitrary and is almost never equal to -EPROBE_DEFER, so the check silently does nothing in the success case. When devm_clk_get() fails with any error other than -EPROBE_DEFER the check is also skipped, leaving max98095->mclk holding an error pointer with no indication to the caller. This means a deferred probe will never actually be triggered for this device, and any non-EPROBE_DEFER clock error is silently swallowed with the error pointer left in the mclk field. Fix this by adding the missing IS_ERR() guard around the PTR_ERR() call, matching the pattern already used in the sibling max98088 and wm8960 drivers. Fixes: e3048c3d2be5 ("ASoC: max98095: Add master clock handling") Signed-off-by: Uday Khare <udaykhare77@gmail.com> Link: https://patch.msgid.link/20260720103950.14474-1-udaykhare77@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
Diffstat (limited to 'tools/perf/scripts/python')
0 files changed, 0 insertions, 0 deletions