linux-stable.git/drivers/pinctrl, branch v6.15.2

pinctrl: mediatek: eint: Fix invalid pointer dereference for v1 platforms

2025-06-10T11:17:09+00:00

commit 1c9977b263475373b31bbf86af94a5c9ae2be42c upstream.

Commit 3ef9f710efcb ("pinctrl: mediatek: Add EINT support for multiple
addresses") introduced an access to the 'soc' field of struct
mtk_pinctrl in mtk_eint_do_init() and for that an include of
pinctrl-mtk-common-v2.h.

However, pinctrl drivers relying on the v1 common driver include
pinctrl-mtk-common.h instead, which provides another definition of
struct mtk_pinctrl that does not contain an 'soc' field.

Since mtk_eint_do_init() can be called both by v1 and v2 drivers, it
will now try to dereference an invalid pointer when called on v1
platforms. This has been observed on Genio 350 EVK (MT8365), which
crashes very early in boot (the kernel trace can only be seen with
earlycon).

In order to fix this, since 'struct mtk_pinctrl' was only needed to get
a 'struct mtk_eint_pin', make 'struct mtk_eint_pin' a parameter
of mtk_eint_do_init() so that callers need to supply it, removing
mtk_eint_do_init()'s dependency on any particular 'struct mtk_pinctrl'.

Fixes: 3ef9f710efcb ("pinctrl: mediatek: Add EINT support for multiple addresses")
Suggested-by: AngeloGioacchino Del Regno 
Signed-off-by: Nícolas F. R. A. Prado 
Link: https://lore.kernel.org/20250520-genio-350-eint-null-ptr-deref-fix-v2-1-6a3ca966a7ba@collabora.com
Signed-off-by: Linus Walleij 
[ukleinek: backport to 6.15.y]
Signed-off-by: Uwe Kleine-König 
Signed-off-by: Greg Kroah-Hartman

pinctrl: armada-37xx: set GPIO output value before setting direction

2025-06-10T11:17:08+00:00

commit e6ebd4942981f8ad37189bbb36a3c8495e21ef4c upstream.

Changing the direction before updating the output value in the
OUTPUT_VAL register may result in a glitch on the output line
if the previous value in the OUTPUT_VAL register is different
from the one we want to set.

In order to avoid that, update the output value before changing
the direction.

Cc: stable@vger.kernel.org
Fixes: 6702abb3bf23 ("pinctrl: armada-37xx: Fix direction_output() callback behavior")
Signed-off-by: Imre Kaloz 
Reviewed-by: Andrew Lunn 
Signed-off-by: Gabor Juhos 
Link: https://lore.kernel.org/20250514-pinctrl-a37xx-fixes-v2-2-07e9ac1ab737@gmail.com
Signed-off-by: Linus Walleij 
Signed-off-by: Greg Kroah-Hartman

pinctrl: armada-37xx: use correct OUTPUT_VAL register for GPIOs > 31

2025-06-10T11:17:08+00:00

commit 947c93eb29c2a581c0b0b6d5f21af3c2b7ff6d25 upstream.

The controller has two consecutive OUTPUT_VAL registers and both
holds output value for 32 GPIOs. Due to a missing adjustment, the
current code always uses the first register while setting the
output value whereas it should use the second one for GPIOs > 31.

Add the missing armada_37xx_update_reg() call to adjust the register
according to the 'offset' parameter of the function to fix the issue.

Cc: stable@vger.kernel.org
Fixes: 6702abb3bf23 ("pinctrl: armada-37xx: Fix direction_output() callback behavior")
Signed-off-by: Imre Kaloz 
Reviewed-by: Andrew Lunn 
Signed-off-by: Gabor Juhos 
Link: https://lore.kernel.org/20250514-pinctrl-a37xx-fixes-v2-1-07e9ac1ab737@gmail.com
Signed-off-by: Linus Walleij 
Signed-off-by: Greg Kroah-Hartman

pinctrl: qcom: switch to devm_register_sys_off_handler()

2025-05-20T21:41:29+00:00

Error-handling paths in msm_pinctrl_probe() don't call
a function required to unroll restart handler registration,
unregister_restart_handler(). Instead of adding calls to this function,
switch the msm pinctrl code into using devm_register_sys_off_handler().

Fixes: cf1fc1876289 ("pinctrl: qcom: use restart_notifier mechanism for ps_hold")
Signed-off-by: Dmitry Baryshkov 
Link: https://lore.kernel.org/20250513-pinctrl-msm-fix-v2-2-249999af0fc1@oss.qualcomm.com
Signed-off-by: Linus Walleij

pinctrl: qcom: Fix PINGROUP definition for sm8750

2025-04-29T08:27:25+00:00

On newer SoCs intr_target_bit position is at 8 instead of 5. Fix it.

Also add missing intr_wakeup_present_bit and intr_wakeup_enable_bit which
enables forwarding of GPIO interrupts to parent PDC interrupt controller.

Fixes: afe9803e3b82 ("pinctrl: qcom: Add sm8750 pinctrl driver")
Signed-off-by: Maulik Shah 
Reviewed-by: Abel Vesa 
Reviewed-by: Dmitry Baryshkov 
Reviewed-by: Melody Olvera 
Link: https://lore.kernel.org/20250429-pinctrl_sm8750-v2-1-87d45dd3bd82@oss.qualcomm.com
Signed-off-by: Linus Walleij

pinctrl: mediatek: common-v1: Fix error checking in mtk_eint_init()

2025-04-23T10:58:57+00:00

The devm_kzalloc() function doesn't return error pointers, it returns
NULL on error.  Then on the next line it checks the same pointer again
by mistake, "->base" instead of "->base[0]".

Fixes: fe412e3a6c97 ("pinctrl: mediatek: common-v1: Fix EINT breakage on older controllers")
Signed-off-by: Dan Carpenter 
Link: https://lore.kernel.org/aAijc10fHka1WAMX@stanley.mountain
Signed-off-by: Linus Walleij

pinctrl: mediatek: Fix new design debounce issue

2025-04-23T08:45:23+00:00

Calculate the true offset of eint according to index.

Fixes: 3ef9f710efcb ("pinctrl: mediatek: Add EINT support for multiple addresses")
Signed-off-by: Hao Chang 
Signed-off-by: Qingliang Li 
Link: https://lore.kernel.org/20250422075216.14073-1-ot_chhao.chang@mediatek.com
Signed-off-by: Linus Walleij

pinctrl: mediatek: common-v1: Fix EINT breakage on older controllers

2025-04-17T07:44:52+00:00

When EINT support for multiple addresses was introduced, the driver
library for the older generations (pinctrl-mtk-common) was not fixed
together. This resulted in invalid pointer accesses.

Fix up the filled in |struct mtk_eint| in pinctrl-mtk-common to match
what is now expected by the mtk-eint library.

Reported-by: Uwe Kleine-König 
Tested-by: Uwe Kleine-König 
Closes: https://lore.kernel.org/all/43nd5jxpk7b7fv46frqlfjnqfh5jlpqsemeoakqzd4wdi3df6y@w7ycd3k5ezvn/
Fixes: 3ef9f710efcb ("pinctrl: mediatek: Add EINT support for multiple addresses")
Cc: Hao Chang 
Cc: Qingliang Li 
Signed-off-by: Chen-Yu Tsai 
Reviewed-by: AngeloGioacchino Del Regno 
Link: https://lore.kernel.org/20250415112339.2385454-1-wenst@chromium.org
Signed-off-by: Linus Walleij

pinctrl: airoha: fix wrong PHY LED mapping and PHY2 LED defines

2025-04-15T07:59:04+00:00

The current PHY2 LED define are wrong and actually set BITs outside the
related mask. Fix it and set the correct value. While at it, also use
FIELD_PREP_CONST macro to make it simple to understand what values are
actually applied for the mask.

Also fix wrong PHY LED mapping. The SoC Switch supports up to 4 port but
the register define mapping for 5 PHY port, starting from 0. The mapping
was wrongly defined starting from PHY1. Reorder the function group to
start from PHY0. PHY4 is actually never supported as we don't have a
GPIO pin to assign.

Cc: stable@vger.kernel.org
Fixes: 1c8ace2d0725 ("pinctrl: airoha: Add support for EN7581 SoC")
Reviewed-by: Benjamin Larsson 
Signed-off-by: Christian Marangi 
Acked-by: Lorenzo Bianconi 
Link: https://lore.kernel.org/20250401135026.18018-1-ansuelsmth@gmail.com
Signed-off-by: Linus Walleij

pinctrl: meson: define the pull up/down resistor value as 60 kOhm

2025-04-15T07:47:37+00:00

The public datasheets of the following Amlogic SoCs describe a typical
resistor value for the built-in pull up/down resistor:
- Meson8/8b/8m2: not documented
- GXBB (S905): 60 kOhm
- GXL (S905X): 60 kOhm
- GXM (S912): 60 kOhm
- G12B (S922X): 60 kOhm
- SM1 (S905D3): 60 kOhm

The public G12B and SM1 datasheets additionally state min and max
values:
- min value: 50 kOhm for both, pull-up and pull-down
- max value for the pull-up: 70 kOhm
- max value for the pull-down: 130 kOhm

Use 60 kOhm in the pinctrl-meson driver as well so it's shown in the
debugfs output. It may not be accurate for Meson8/8b/8m2 but in reality
60 kOhm is closer to the actual value than 1 Ohm.

Signed-off-by: Martin Blumenstingl 
Reviewed-by: Neil Armstrong 
Link: https://lore.kernel.org/20250329190132.855196-1-martin.blumenstingl@googlemail.com
Signed-off-by: Linus Walleij