linux-kernel - [GIT PULL] pwm: Two fixes for 6.17-rc1

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: <ljulzcaas5usyyqjzcnhjtgrv7loxz5hgabdjso237kwzndwln@5zfghmq3kqjt>
Date: Thu, 31 Jul 2025 11:38:29 +0200
From: Uwe Kleine-König <ukleinek@...nel.org>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: linux-pwm@...r.kernel.org, linux-kernel@...r.kernel.org, 
	Laurentiu Mihalcea <laurentiu.mihalcea@....com>, 
	AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>
Subject: [GIT PULL] pwm: Two fixes for 6.17-rc1

Hello Linus,

the following changes since commit 68b9272ca7ac948b71aba482ef8244dee8032f46:

  pwm: raspberrypi-poe: Fix spelling mistake "Firwmware" -> "Firmware" (2025-07-24 23:04:16 +0200)

are available in the Git repository at:

  https://git.kernel.org/pub/scm/linux/kernel/git/ukleinek/linux.git tags/pwm/for-6.17-rc1-fixes

for you to fetch changes up to 65c6f742ab14ab1a2679fba72b82dcc0289d96f1:

  pwm: imx-tpm: Reset counter if CMOD is 0 (2025-07-29 17:51:27 +0200)

.

Note that there are three commits for these two fixes; that's because
the mediatek PWM fix is split into a preparing commit and the actual
fix. The first introduces no semantic change and only makes the second
commit smaller and easier to understand. The commits are tagged for
backporting to stable and contained in yesterday's and today's next
without reported issues.

Please pull these commits before tagging 6.17, preferably considerably
earlier :-)

Thanks
Uwe

----------------------------------------------------------------
pwm: Two fixes for 6.17-rc1

Here contained are two fixes for the mediatek and the imx-tpm driver.
Both are already old but (v4.12-rc1 and v5.2-rc1 respectively) given
it's still very early in the development cycle I consider them fine to
go in.

The mediatek issue is that both period and duty_cycle were configured to
higher values than requested. For most applications the period part is
no tragedy, but a PWM that is configured for duty_cycle = 0 should
really emit a constant inactive signal. That was noticed by an LED not
being completely off in this case.

For the imx-tpm PWM driver the fixed issue is that the first period is
quite a bit too long under some circumstances. So it might take up to
UINT32_MAX << 7 clock ticks until the PWM starts toggling. With an
assumed input clock rate of 166 MHz (completely made up) that's 55
minutes.
----------------------------------------------------------------

Laurentiu Mihalcea (1):
      pwm: imx-tpm: Reset counter if CMOD is 0

Uwe Kleine-König (2):
      pwm: mediatek: Handle hardware enable and clock enable separately
      pwm: mediatek: Fix duty and period setting

 drivers/pwm/pwm-imx-tpm.c  |  9 ++++++
 drivers/pwm/pwm-mediatek.c | 71 ++++++++++++++++++++++++----------------------
 2 files changed, 46 insertions(+), 34 deletions(-)

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)