in 6.18-rc1, failed to query a GPIO using a name in ACPI
Subsystems (2)
Hello,
While testing the AAEON UpBoard MFD driver with the v6.18-rc1 release, I noticed a regression. The driver does not probe anymore and I get the following error:
[ 1.977831] upboard-fpga AANT0F01:00: error -ENOENT: Failed to initialize FPGA common GPIOs [ 1.977854] upboard-fpga AANT0F01:00: probe with driver upboard-fpga
failed with error -2
The driver fails to get some GPIOs using a name [1] in the ACPI tables [2].
I did a bisect and I found the commit which introduced the regression:e121be784d35 ("ACPI: property: Refactor acpi_fwnode_get_reference_args() to support nargs_prop")
Best Regards,
Thomas
[1]
https://elixir.bootlin.com/linux/v6.17/source/drivers/mfd/upboard-fpga.c#L192
[2] https://gist.github.com/thom24/2da44ef86eacfa5d2d492ce43fa41aa4
Quoted 9 lines · Thomas Richard
Hello,
While testing the AAEON UpBoard MFD driver with the v6.18-rc1 release, I noticed a regression. The driver does not probe anymore and I get the following error:
[ 1.977831] upboard-fpga AANT0F01:00: error -ENOENT: Failed to initialize FPGA common GPIOs [ 1.977854] upboard-fpga AANT0F01:00: probe with driver upboard-fpga
Quoted 8 lines
failed with error -2
The driver fails to get some GPIOs using a name [1] in the ACPI tables [2].
I did a bisect and I found the commit which introduced the regression:
e121be784d35("ACPI: property: Refactoracpi_fwnode_get_reference_args()to supportnargs_prop")
Hi Thomas,
Could you please try with below change and see if it resolves the issue?
drivers/acpi/property.c +1 -1
@@ -1107,7 +1107,7 @@ int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode,size_t num_args, struct fwnode_reference_args *args) { - return acpi_fwnode_get_reference_args(fwnode, propname, NULL, index, num_args, args); + return acpi_fwnode_get_reference_args(fwnode, propname, NULL, num_args, index, args); }
Thanks,
Sunil
Quoted 27 lines · Sunil V L
Quoted 9 lines · Thomas Richard
Hello,
While testing the AAEON UpBoard MFD driver with the v6.18-rc1 release, I noticed a regression. The driver does not probe anymore and I get the following error:
[ 1.977831] upboard-fpga AANT0F01:00: error -ENOENT: Failed to initialize FPGA common GPIOs [ 1.977854] upboard-fpga AANT0F01:00: probe with driver upboard-fpgaQuoted 8 lines
failed with error -2
The driver fails to get some GPIOs using a name [1] in the ACPI tables [2].
I did a bisect and I found the commit which introduced the regression:
e121be784d35("ACPI: property: Refactoracpi_fwnode_get_reference_args()to supportnargs_prop")Hi Thomas,
Could you please try with below change and see if it resolves the issue?
drivers/acpi/property.c +0
-0
drivers/acpi/property.c int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode, · line 1112
Quoted 7 lines
drivers/acpi/property.c +1
-1@@ -1107,7 +1107,7 @@ int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode,size_t num_args, struct fwnode_reference_args *args) { - return acpi_fwnode_get_reference_args(fwnode, propname, NULL, index, num_args, args); + return acpi_fwnode_get_reference_args(fwnode, propname, NULL, num_args, index, args); }
Hi Sunil,
Yes your change fixes the issue !!
Thanks,
Thomas
Quoted 38 lines · Thomas Richard
Quoted 27 lines · Sunil V L
Quoted 9 lines · Thomas Richard
Hello,
While testing the AAEON UpBoard MFD driver with the v6.18-rc1 release, I noticed a regression. The driver does not probe anymore and I get the following error:
[ 1.977831] upboard-fpga AANT0F01:00: error -ENOENT: Failed to initialize FPGA common GPIOs [ 1.977854] upboard-fpga AANT0F01:00: probe with driver upboard-fpgaQuoted 8 lines
failed with error -2
The driver fails to get some GPIOs using a name [1] in the ACPI tables [2].
I did a bisect and I found the commit which introduced the regression:
e121be784d35("ACPI: property: Refactoracpi_fwnode_get_reference_args()to supportnargs_prop")Hi Thomas,
Could you please try with below change and see if it resolves the issue?
drivers/acpi/property.c +0
-0drivers/acpi/property.c int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode, · line 1112
Quoted 7 lines
drivers/acpi/property.c +1
-1@@ -1107,7 +1107,7 @@ int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode,size_t num_args, struct fwnode_reference_args *args) { - return acpi_fwnode_get_reference_args(fwnode, propname, NULL, index, num_args, args); + return acpi_fwnode_get_reference_args(fwnode, propname, NULL, num_args, index, args); }Hi Sunil,
Quoted 2 lines
Yes your change fixes the issue !!
Hi Thomas,
Thanks for confirming the fix!
I'll send out the patch shortly and include your Tested-by tag.
Apologies for the earlier issue, and thank you for your help.
Best regards,
Sunil
Quoted 35 lines · Sunil V L
Quoted 20 lines · Thomas Richard
Quoted 19 lines · Sunil V L
Quoted 8 lines · Thomas Richard
While testing the AAEON UpBoard MFD driver with the v6.18-rc1 release, I noticed a regression. The driver does not probe anymore and I get the following error:
[ 1.977831] upboard-fpga AANT0F01:00: error -ENOENT: Failed to initialize FPGA common GPIOs [ 1.977854] upboard-fpga AANT0F01:00: probe with driver upboard-fpgaQuoted 7 lines
failed with error -2
The driver fails to get some GPIOs using a name [1] in the ACPI tables [2].
I did a bisect and I found the commit which introduced the regression:
e121be784d35("ACPI: property: Refactoracpi_fwnode_get_reference_args()to supportnargs_prop")Could you please try with below change and see if it resolves the issue?
Quoted 13 lines
drivers/acpi/property.c int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode, · line 1111
Quoted 11 lines
drivers/acpi/property.c +1
-1@@ -1107,7 +1107,7 @@ int __acpi_node_get_property_reference(const struct fwnode_handle *fwnode,size_t num_args, struct fwnode_reference_args *args) { - return acpi_fwnode_get_reference_args(fwnode, propname, NULL, index, num_args, args); + return acpi_fwnode_get_reference_args(fwnode, propname, NULL, num_args, index, args); }Yes your change fixes the issue !!
Quoted 5 lines
Thanks for confirming the fix!
I'll send out the patch shortly and include your Tested-by tag.
Apologies for the earlier issue, and thank you for your help.
I haven't received anything. Have you sent anything? Am I Cc'ed?
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 2 lines · Andy Shevchenko
Quoted 1 line · Sunil V L
On Fri, Oct 17, 2025 at 10:50:05AM +0200, Thomas Richard wrote:
Seems I have had an issue with my email, and everything is now fine.
Quoted 5 lines
Quoted 3 lines
I'll send out the patch shortly and include your Tested-by tag.
Apologies for the earlier issue, and thank you for your help.
I haven't received anything. Have you sent anything? Am I Cc'ed?
Signature · 2 lines
With Best Regards, Andy Shevchenko