mbox series

[GIT,PULL] i2c-host-fixes for v6.11-rc8

Message ID 2itivkfhp7s2dzmihtlo67w53wpdiukmmrcf625isex5rrtu23@4naen2rzimxw
State New
Headers show
Series [GIT,PULL] i2c-host-fixes for v6.11-rc8 | expand

Pull-request

git://git.kernel.org/pub/scm/linux/kernel/git/andi.shyti/linux.git tags/i2c-host-fixes-6.11-rc8

Message

Andi Shyti Sept. 13, 2024, 9:53 a.m. UTC
Hi Wolfram,

Here’s the fixes pull request. It’s funny, after a few weeks with
no fixes, right at the merge window, I’m sending in five!

I’ve also brought in from the past two fixes from Robert Hancock
for the Xilinx controller.

See you later today for the next pull request.

Thanks,
Andi

The following changes since commit da3ea35007d0af457a0afc87e84fddaebc4e0b63:

  Linux 6.11-rc7 (2024-09-08 14:50:28 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/andi.shyti/linux.git tags/i2c-host-fixes-6.11-rc8

for you to fetch changes up to e2c85d85a05f16af2223fcc0195ff50a7938b372:

  i2c: qcom-geni: Use IRQF_NO_AUTOEN flag in request_irq() (2024-09-12 09:08:22 +0200)

----------------------------------------------------------------
The Aspeed driver tracks the controller's state (stop, pending,
start, etc.). Previously, when the stop command was sent, the
state was not updated. The fix in this pull request ensures the
driver's state is aligned with the device status.

The Intel SCH driver receives a new look, and among the cleanups,
there is a fix where, due to an oversight, an if/else statement
was missing the else, causing it to move forward instead of
exiting the function in case of an error.

The Qualcomm GENI I2C driver adds the IRQF_NO_AUTOEN flag to the
IRQ setup to prevent unwanted interrupts during probe.

The Xilinx XPS controller fixes TX FIFO handling to avoid missed
NAKs. Another fix ensures the controller is reinitialized when
the bus appears busy.

----------------------------------------------------------------
Andy Shevchenko (1):
      i2c: isch: Add missed 'else'

Jinjie Ruan (1):
      i2c: qcom-geni: Use IRQF_NO_AUTOEN flag in request_irq()

Robert Hancock (2):
      i2c: xiic: Wait for TX empty to avoid missed TX NAKs
      i2c: xiic: Try re-initialization on bus busy timeout

Tommy Huang (1):
      i2c: aspeed: Update the stop sw state when the bus recovery occurs

 drivers/i2c/busses/i2c-aspeed.c    | 16 ++++++++--------
 drivers/i2c/busses/i2c-isch.c      |  3 +--
 drivers/i2c/busses/i2c-qcom-geni.c |  4 +---
 drivers/i2c/busses/i2c-xiic.c      | 60 +++++++++++++++++++++++++++++++++++-------------------------
 4 files changed, 45 insertions(+), 38 deletions(-)

Comments

Wolfram Sang Sept. 16, 2024, 12:13 p.m. UTC | #1
Hi Andi,

> Here’s the fixes pull request. It’s funny, after a few weeks with
> no fixes, right at the merge window, I’m sending in five!

And sorry for not getting it into 6.11-final. I was on my way to Vienna
to the conferences but the weather conditions made the travel quite
chaotic, so I didn't have enough bandwidth to handle the PR in time.

I will send it to Linus as the first I2C pull request right away. Before
our "main one". Sorry for the troubles!

All the best,

   Wolfram
Wolfram Sang Sept. 16, 2024, 1:23 p.m. UTC | #2
> I can take the opportunity to include a few more fixes and send
> a new pull request.

Let's make a new PR out of these. I don't want to send Linus a PR which
we modified until the last second. This one has been stable for a few
days, let's keep it in this state. The new PR will have its own
timeline.

> Also, there’s a patch from you (a fix of a fix) that I’ve already
> commented on, which can be integrated as well.

Further comment coming up ;)