[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20191104001819.2300-1-chris.packham@alliedtelesis.co.nz>
Date: Mon, 4 Nov 2019 13:18:17 +1300
From: Chris Packham <chris.packham@...iedtelesis.co.nz>
To: linus.walleij@...aro.org, rjui@...adcom.com, sbranden@...adcom.com,
bcm-kernel-feedback-list@...adcom.com
Cc: linux-arm-kernel@...ts.infradead.org, linux-gpio@...r.kernel.org,
linux-kernel@...r.kernel.org,
Chris Packham <chris.packham@...iedtelesis.co.nz>
Subject: [PATCH v3 0/2] pinctrl: bcm: nsp: gpio improvements
I'm working on a platform using the BCM 58525 SoC. I noticed that some of
the peripherals were being deferred (not that that's a problem) and debugfs
was complaining "File ':axi@...00000:gpio@20' in directory 'domains' already
present!" which is more of a sign that things were not right.
The debugfs error was because the manually created irq domain was not
cleaned up on failure (or deferral).
I've dropped the patch from this series that changes the order in the
device tree. I can probably live with the deferrals.
While I was debugging another issue I noticed my gpio-hogs weren't
showing up correctly in /sys/kernel/debug/gpio. At first I thought I was
missing commit d95da993383c ("gpiolib: Preserve desc->flags when setting
state") but as it turns out pinctrl-nsp-gpio.c didn't provide a
get_direction function so the generic code assumed they all started as
inputs. I've added a new patch to address that.
Chris Packham (2):
pinctrl: bcm: nsp: use gpiolib infrastructure for interrupts
pinctrl: bcm: nsp: implement get_direction
drivers/pinctrl/bcm/pinctrl-nsp-gpio.c | 119 ++++++++++++-------------
1 file changed, 56 insertions(+), 63 deletions(-)
--
2.23.0
Powered by blists - more mailing lists