diff options
author | Stefan Wahren <stefan.wahren@i2se.com> | 2023-04-27 13:21:52 +0200 |
---|---|---|
committer | Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> | 2023-05-08 08:13:35 +0200 |
commit | dca5480ab7b77a889088ab7cac81934604510ac7 (patch) | |
tree | 2fd18ad52ee9970bde0dffede433377daeb78610 /lib/syscall.c | |
parent | ac9a78681b921877518763ba0e89202254349d1b (diff) | |
download | linux-dca5480ab7b77a889088ab7cac81934604510ac7.tar.gz linux-dca5480ab7b77a889088ab7cac81934604510ac7.tar.bz2 linux-dca5480ab7b77a889088ab7cac81934604510ac7.zip |
w1: w1_therm: fix locking behavior in convert_t
The commit 67b392f7b8ed ("w1_therm: optimizing temperature read timings")
accidentially inverted the logic for lock handling of the bus mutex.
Before:
pullup -> release lock before sleep
no pullup -> release lock after sleep
After:
pullup -> release lock after sleep
no pullup -> release lock before sleep
This cause spurious measurements of 85 degree (powerup value) on the
Tarragon board with connected 1-w temperature sensor
(w1_therm.w1_strong_pull=0).
In the meantime a new feature for polling the conversion
completion has been integrated in these branches with
commit 021da53e65fd ("w1: w1_therm: Add sysfs entries to control
conversion time and driver features"). But this feature isn't
available for parasite power mode, so handle this separately.
Link: https://lore.kernel.org/regressions/2023042645-attentive-amends-7b0b@gregkh/T/
Fixes: 67b392f7b8ed ("w1_therm: optimizing temperature read timings")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Link: https://lore.kernel.org/r/20230427112152.12313-1-stefan.wahren@i2se.com
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Diffstat (limited to 'lib/syscall.c')
0 files changed, 0 insertions, 0 deletions