linux-kernel - [PATCH 0/2] iio: adc: ad4695: add oversampling support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20241217-ad4695-oversampling-v1-0-0b045d835dac@baylibre.com>
Date: Tue, 17 Dec 2024 16:47:27 -0500
From: Trevor Gamblin <tgamblin@...libre.com>
To: Michael Hennerich <michael.hennerich@...log.com>, 
 Nuno Sá <nuno.sa@...log.com>, 
 David Lechner <dlechner@...libre.com>, Lars-Peter Clausen <lars@...afoo.de>, 
 Jonathan Cameron <jic23@...nel.org>, Jonathan Corbet <corbet@....net>
Cc: linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org, 
 linux-doc@...r.kernel.org, Trevor Gamblin <tgamblin@...libre.com>
Subject: [PATCH 0/2] iio: adc: ad4695: add oversampling support

Add driver logic and documentation for the oversampling feature of the
AD469x parts from Analog Devices. For now, this only works with offload
support, and takes advantage of that mode's higher performance to make
oversampling possible on multiple channels with varying sampling
frequencies. Some significant rework of the driver had to be done in
order to conditionally support this feature, including use of
iio_scan_types to help determine the appropriate spi message
configurations depending on oversampling ratio.

A dilemma that came up during development of this feature was whether or
not the implementation of oversampling ratios against sampling frequency
was actually correct. More specifically, it's unclear if the sampling
frequency attribute is supposed to be the conversion rate or the data
read rate (according to the IIO subsystem). If it's the former, then
this implementation is probably incorrect. David Lechner pointed out
during review that it would be easier if it were defined as the
conversion rate and that it was userspace's responsibility to handle
oversampling ratio, but that might also require more work in the IIO
subsystem. Two other ADC drivers that were referenced for inspiration
when working through this were the ad7380 and the rtq6056. The ad7380
has a global oversampling setting rather than per-channel, and the
rtq6056 seems at least partially broken because it only takes
oversampling ratio into account when getting the sampling frequency (but
not when setting it). Instead of per-driver implementation, these three
drivers might serve as inspiration for changes to how oversampling is
handled in IIO?

This series depends on David's recent SPI engine changes for adding
offload support:

https://lore.kernel.org/all/20241211-dlech-mainline-spi-engine-offload-2-v6-0-88ee574d5d03@baylibre.com/

Signed-off-by: Trevor Gamblin <tgamblin@...libre.com>
---
Trevor Gamblin (2):
      iio: adc: ad4695: add offload-based oversampling support
      doc: iio: ad4695: describe oversampling support

 Documentation/iio/ad4695.rst |  36 ++++-
 drivers/iio/adc/ad4695.c     | 378 +++++++++++++++++++++++++++++++++++++++----
 2 files changed, 383 insertions(+), 31 deletions(-)
---
base-commit: 0c6c3bf84f541fb4ec7097baf9eac10136f98c62
change-id: 20241217-ad4695-oversampling-2946fbe3aff3

Best regards,
-- 
Trevor Gamblin <tgamblin@...libre.com>