mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH v2 0/6] iio: proximity: sx93xx: fix raw reads, thresholds, hysteresis and gain
@ 2026-11-22  7:46 Arnav Kapoor
  2026-11-22  7:46 ` [PATCH v2 1/6] iio: proximity: sx_common: fix raw reads while the trigger is in use Arnav Kapoor
                   ` (5 more replies)
  0 siblings, 6 replies; 7+ messages in thread
From: Arnav Kapoor @ 2026-11-22  7:46 UTC (permalink / raw)
  To: Jonathan Cameron
  Cc: David Lechner, Nuno Sá,
	Andy Shevchenko, Andy Shevchenko, Gwendal Grignou, Stephen Boyd,
	linux-iio, linux-kernel, Arnav Kapoor

The remaining sx93xx fixes, grouped by driver.

1: sx_common: a raw read can return on a stale completion (left by an
   interrupted read, or by conversions while another device uses the
   sensor's trigger), and it then turns off the CONVDONE interrupt the
   trigger depends on.
2, 5: sx9324/sx9360 write_thresh() lowers the threshold by one register
   value when the value just read from an odd register value is
   written back, and lets some negative values through.
3, 6: sx9324/sx9360 write_hysteresis() turns negative values into the
   largest hysteresis.
4: sx9360 gain encoding, as in Semtech's own SX9360 driver and as
   sx9324 got in 108e4d4de2b5.

Changes in v2:
- Rebase on the current iio testing, which now has all the patches v1
  depended on
- 2, 5: check _val before it is used, and before reg is assigned;
  val > 1 instead of val >= 1 (Andy)
- 2, 5: commit message rewritten as agreed with Andy: drop the
  (r * r - 1) / 2 formula, explain why int_sqrt(2 * val + 1) is right
  in integer arithmetic, and say that the negative value wrap happens
  on both 32-bit and 64-bit
- 3, 6: check _val before it is used (Andy); say in the commit message
  that it happens on both 32-bit and 64-bit (Andy)

v1: https://lore.kernel.org/all/20261116205320.26070-1-kapoorarnav43@gmail.com/

The threshold change was checked against a userspace copy of both
conversions for every val from 0 to 65535: all 256 register values
round-trip, and every write picks the largest register value whose
reported threshold doesn't exceed the request.
Compile-tested only (W=1, checkpatch --strict clean); I don't have
SX93xx hardware.

Arnav Kapoor (6):
  iio: proximity: sx_common: fix raw reads while the trigger is in use
  iio: proximity: sx9324: fix proximity threshold writes
  iio: proximity: sx9324: reject negative hysteresis values
  iio: proximity: sx9360: fix hardware gain encoding
  iio: proximity: sx9360: fix proximity threshold writes
  iio: proximity: sx9360: reject negative hysteresis values

 drivers/iio/proximity/sx9324.c    | 18 ++++++++++----
 drivers/iio/proximity/sx9360.c    | 39 +++++++++++++++++++++++--------
 drivers/iio/proximity/sx_common.c | 25 +++++++++++++++-----
 3 files changed, 62 insertions(+), 20 deletions(-)


base-commit: 42d25ad999bd60b96b227c2e54391c82cb03d00d
-- 
2.53.0


^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2026-11-22  7:47 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-11-22  7:46 [PATCH v2 0/6] iio: proximity: sx93xx: fix raw reads, thresholds, hysteresis and gain Arnav Kapoor
2026-11-22  7:46 ` [PATCH v2 1/6] iio: proximity: sx_common: fix raw reads while the trigger is in use Arnav Kapoor
2026-11-22  7:46 ` [PATCH v2 2/6] iio: proximity: sx9324: fix proximity threshold writes Arnav Kapoor
2026-11-22  7:46 ` [PATCH v2 3/6] iio: proximity: sx9324: reject negative hysteresis values Arnav Kapoor
2026-11-22  7:46 ` [PATCH v2 4/6] iio: proximity: sx9360: fix hardware gain encoding Arnav Kapoor
2026-11-22  7:46 ` [PATCH v2 5/6] iio: proximity: sx9360: fix proximity threshold writes Arnav Kapoor
2026-11-22  7:46 ` [PATCH v2 6/6] iio: proximity: sx9360: reject negative hysteresis values Arnav Kapoor

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®