2021-12-15 13:49:24

by Mark Brown

Subject: linux-next: build failure after merge of the phy-next tree

Hi all,

After merging the phy-next tree, today's linux-next build (x86_64
allmodconfig) failed like this:

/tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c: In function 'imx8_pcie_phy_init':
/tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c:41:37: error: implicit declaration of function 'FIELD_PREP' [-Werror=implicit-function-declaration]
41 | #define IMX8MM_GPR_PCIE_REF_CLK_EXT FIELD_PREP(IMX8MM_GPR_PCIE_REF_CLK_SEL, 0x2)
| ^~~~~~~~~~
/tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c:85:7: note: in expansion of macro 'IMX8MM_GPR_PCIE_REF_CLK_EXT'
85 | IMX8MM_GPR_PCIE_REF_CLK_EXT :
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors

Caused by commit

1aa97b002258a190d77 ("phy: freescale: pcie: Initialize the imx8 pcie standalone phy driver")

I have used the phy-next tree from yesterday.

2021-12-15 14:14:35

by Marcel Ziswiler

Subject: Re: linux-next: build failure after merge of the phy-next tree

On Wed, 2021-12-15 at 13:49 +0000, [email protected] wrote:
> Hi all,
>
> After merging the phy-next tree, today's linux-next build (x86_64
> allmodconfig) failed like this:

I admit I never tried building any of this as a module. I assume neither did Richard (now on CC) or anybody
else.

> /tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c: In function 'imx8_pcie_phy_init':
> /tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c:41:37: error: implicit declaration of function
> 'FIELD_PREP' [-Werror=implicit-function-declaration]

Hm, that should come from include/linux/bitfield.h. However, that seems not explicitly included. Likely, in the
module case it no longer finds it. That's a problem.

>    41 | #define IMX8MM_GPR_PCIE_REF_CLK_EXT FIELD_PREP(IMX8MM_GPR_PCIE_REF_CLK_SEL, 0x2)
>       |                                     ^~~~~~~~~~
> /tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c:85:7: note: in expansion of macro
> 'IMX8MM_GPR_PCIE_REF_CLK_EXT'
>    85 |       IMX8MM_GPR_PCIE_REF_CLK_EXT :
>       |       ^~~~~~~~~~~~~~~~~~~~~~~~~~~
> cc1: all warnings being treated as errors
>
> Caused by commit
>
>   1aa97b002258a190d77 ("phy: freescale: pcie: Initialize the imx8 pcie standalone phy driver")
>
> I have used the phy-next tree from yesterday.

Yes, I also noticed it having been applied there now.

Anyway, let me cook up a fix for this. Thanks for reporting.

2021-12-15 15:22:08

by Vinod Koul

Subject: Re: linux-next: build failure after merge of the phy-next tree

On 15-12-21, 14:14, Marcel Ziswiler wrote:
> On Wed, 2021-12-15 at 13:49 +0000, [email protected] wrote:
> > Hi all,
> >
> > After merging the phy-next tree, today's linux-next build (x86_64
> > allmodconfig) failed like this:
>
> I admit I never tried building any of this as a module. I assume neither did Richard (now on CC) or anybody
> else.
>
> > /tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c: In function 'imx8_pcie_phy_init':
> > /tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c:41:37: error: implicit declaration of function
> > 'FIELD_PREP' [-Werror=implicit-function-declaration]
>
> Hm, that should come from include/linux/bitfield.h. However, that seems not explicitly included. Likely, in the
> module case it no longer finds it. That's a problem.
>
> > ?? 41 | #define IMX8MM_GPR_PCIE_REF_CLK_EXT FIELD_PREP(IMX8MM_GPR_PCIE_REF_CLK_SEL, 0x2)
> > ????? |???????????????????????????????????? ^~~~~~~~~~
> > /tmp/next/build/drivers/phy/freescale/phy-fsl-imx8m-pcie.c:85:7: note: in expansion of macro
> > 'IMX8MM_GPR_PCIE_REF_CLK_EXT'
> > ?? 85 |?????? IMX8MM_GPR_PCIE_REF_CLK_EXT :
> > ????? |?????? ^~~~~~~~~~~~~~~~~~~~~~~~~~~
> > cc1: all warnings being treated as errors
> >
> > Caused by commit
> >
> > ? 1aa97b002258a190d77 ("phy: freescale: pcie: Initialize the imx8 pcie standalone phy driver")
> >
> > I have used the phy-next tree from yesterday.
>
> Yes, I also noticed it having been applied there now.
>
> Anyway, let me cook up a fix for this. Thanks for reporting.

I have fixed it up: https://lore.kernel.org/all/[email protected]/

will push now

--
~Vinod