Re: [PATCH] gpio: shared: call gpio_chip::of_xlate() if set

[Jon Hunter]

On 17/03/2026 11:44, Bartosz Golaszewski wrote:
> On Tue, Mar 17, 2026 at 11:12 AM Jon Hunter <jonathanh@xxxxxxxxxx> wrote:
> > Thanks for sending this. However, I am seeing a different issue now ...
> >
> >    ------------[ cut here ]------------
> >    WARNING: drivers/gpio/gpiolib-shared.c:499 at gpio_shared_add_proxy_lookup+0x118/0x1d8, CPU#8: swapper/0/1
> >    Modules linked in:
> >    CPU: 8 UID: 0 PID: 1 Comm: swapper/0 Not tainted 7.0.0-rc3-next-20260309-00005-g02826fefa46f #14 PREEMPT
> >    Hardware name: NVIDIA NVIDIA Jetson AGX Orin Developer Kit/Jetson, BIOS buildbrain-gcid-42974706 11/20/2025
> >    pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> >    pc : gpio_shared_add_proxy_lookup+0x118/0x1d8
> >    lr : gpio_shared_add_proxy_lookup+0xfc/0x1d8
> >    sp : ffff8000832bba30
> >    x29: ffff8000832bba30 x28: ffff000080d01010 x27: ffffffffffffefff
> >    x26: 0000000000000001 x25: ffff800082df0538 x24: ffff800082df0528
> >    x23: 0000000000000000 x22: ffff00008012c158 x21: ffff000081455010
> >    usb 1-3: new full-speed USB device number 2 using tegra-xusb
> >    x20: ffff000080d5d430 x19: ffff00008012c158 x18: 00000000ffffffff
> >    x17: ffff8000830786a8 x16: ffff800083078718 x15: ffff8000832bb880
> >    x14: ffffffffffffffff x13: 0000000000000008 x12: 0101010101010101
> >    x11: 7f7f7f7f7f7f7f7f x10: ffff8000827f20d0 x9 : 0000000000000003
> >    x8 : 0101010101010101 x7 : 0080808080808000 x6 : 15151a0a59460209
> >    x5 : 000000000000003c x4 : ffff8000832bb990 x3 : ffff0000800fe800
> >    x2 : ffff0000801c2f40 x1 : ffff0000801c2f40 x0 : ffff800082df0538
> >    Call trace:
> >     gpio_shared_add_proxy_lookup+0x118/0x1d8 (P)
> >     gpiod_find_and_request+0x1bc/0x548
> >     devm_fwnode_gpiod_get_index+0x1c/0x6c
> >     gpio_keys_probe+0x494/0x9fc
> >     platform_probe+0x5c/0x98
> >     really_probe+0xbc/0x2a8
> >     __driver_probe_device+0x78/0x12c
> >     driver_probe_device+0x3c/0x15c
> >     __driver_attach+0x90/0x19c
> >     bus_for_each_dev+0x78/0xd4
> >     driver_attach+0x24/0x30
> >     bus_add_driver+0xe4/0x208
> >     driver_register+0x5c/0x124
> >     __platform_driver_register+0x24/0x30
> >     gpio_keys_init+0x1c/0x28
> >     do_one_initcall+0x7c/0x1c0
> >     kernel_init_freeable+0x204/0x2ec
> >     kernel_init+0x24/0x1e0
> >     ret_from_fork+0x10/0x20
> >    ---[ end trace 0000000000000000 ]---
> >
> > I see the comment says ...
> >
> >    /* We warn here because this can only happen if the programmer borked. */
> >    WARN_ON(1);
> >
> > I will take a closer look, but let me know if you have any thoughts?
>
> I suppose this is not a reset-gpio-like use-case? Could you point me
> to the DTS you're using? I've reproduced the bug with a dummy dts and
> xlate function in gpio-sim and this patch fixes it but maybe I'm
> missing something.

So the board I originally observed this on is a farm board with camera
hardware I don't have. So like you I started off reproducing locally
with a dummy dts by making the following change ...

diff --git a/arch/arm64/boot/dts/nvidia/tegra234-p3737-0000+p3701.dtsi b/arch/arm64/boot/dts/nvidia/tegra234-p3737-0000+p3701.dtsi
index f6cad29355e6..5e62ffb425f4 100644
--- a/arch/arm64/boot/dts/nvidia/tegra234-p3737-0000+p3701.dtsi
+++ b/arch/arm64/boot/dts/nvidia/tegra234-p3737-0000+p3701.dtsi
@@ -389,14 +389,14 @@ gpio-keys {
 
                key-force-recovery {
                        label = "Force Recovery";
-                       gpios = <&gpio TEGRA234_MAIN_GPIO(G, 0) GPIO_ACTIVE_LOW>;
+                       gpios = <&gpio TEGRA234_MAIN_GPIO(AF, 2) GPIO_ACTIVE_LOW>;
                        linux,input-type = <EV_KEY>;
                        linux,code = <BTN_1>;
                };
 
                key-power {
                        label = "Power";
-                       gpios = <&gpio_aon TEGRA234_AON_GPIO(EE, 4) GPIO_ACTIVE_LOW>;
+                       gpios = <&gpio TEGRA234_MAIN_GPIO(AF, 2) GPIO_ACTIVE_LOW>;
                        linux,input-type = <EV_KEY>;
                        linux,code = <KEY_POWER>;
                        wakeup-event-action = <EV_ACT_ASSERTED>;


With this I see ...

 gpiolib_shared: GPIO 154 owned by tegra234-gpio is shared by multiple consumers
 gpiolib_shared: Setting up a shared GPIO entry for key-force-recovery (con_id: '(none)')
 gpiolib_shared: Created an auxiliary GPIO proxy gpiolib_shared.proxy.3 for GPIO device tegra234-gpio
 gpiolib_shared: Setting up a shared GPIO entry for key-power (con_id: '(none)')
 gpiolib_shared: Created an auxiliary GPIO proxy gpiolib_shared.proxy.4 for GPIO device tegra234-gpio
 ...
 gpio-keys gpio-keys: error -ENXIO: Unable to determine IRQ# for button #0

I am not sure if this is because these are child nodes of gpio-keys?
Obviously this is not a proper example, but something quick and dirty
for local testing :-)

Jon

--
nvpublic