diff options
| author | Frieder Schrempf <frieder.schrempf@kontron.de> | 2026-09-21 12:16:46 +0200 |
|---|---|---|
| committer | Mark Brown <broonie@kernel.org> | 2026-09-21 14:15:05 +0200 |
| commit | 38d4b42e4a2af5066b602867042c10c699bd08ad (patch) | |
| tree | 07e75de516c2d3ab93093eee409f7db4fdadc255 /tools/perf/scripts/python/task-analyzer.py | |
| parent | 631a5633638136add14406d5bc18ca113492cd77 (diff) | |
spi: dw: Use the rx-sample-delay-ns value parsed by the core
The core now parses the "rx-sample-delay-ns" peripheral property into
spi_device.rx_sample_delay_ns, so drop the private copy of that parsing.
The controller-wide default is deliberately left alone. It is read from
the *controller* node into dws->def_rx_sample_dly_ns, which is a
different node and none of the core's business when it parses properties
of a peripheral.
One corner case changes: the peripheral value is now treated as unset
when it is zero, whereas before an absent property could be told apart
from an explicit "rx-sample-delay-ns = <0>", the latter overriding a
non-zero controller default with no delay at all. There are no in-tree
users affected by this and it is very unlikely for out-of-tree users
to set rx-sample-delay-ns on the controller level and then revert it
to 0 at the chip level. We should abandon this configuration that was
probably never meant to work like this and doesn't make sense.
Assisted-by: Claude:claude-opus-5
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Link: https://patch.msgid.link/20260921-spi-sample-delay-cleanup-v2-3-b460d572fd32@kontron.de
Signed-off-by: Mark Brown <broonie@kernel.org>
Diffstat (limited to 'tools/perf/scripts/python/task-analyzer.py')
0 files changed, 0 insertions, 0 deletions
