On Friday 28 April 2023 21:35:19 Randy Dunlap wrote:
> Neither FSL_SOC_BOOKE nor PPC_86xx enables CONFIG_PCI by
> default, so it may be unset in some randconfigs.
> When that happens, FSL_ULI1575 may be set when it should not be
> since it is a PCI driver. When it is set, there are 3 kconfig
> warnings and a slew of build errors
>
> WARNING: unmet direct dependencies detected for PCI_QUIRKS
> Depends on [n]: PCI [=n]
> Selected by [y]:
> - FSL_PCI [=y]
>
> WARNING: unmet direct dependencies detected for GENERIC_ISA_DMA
> Depends on [n]: ISA_DMA_API [=n]
> Selected by [y]:
> - FSL_ULI1575 [=y] && (FSL_SOC_BOOKE [=n] || PPC_86xx [=y])
>
> WARNING: unmet direct dependencies detected for PPC_INDIRECT_PCI
> Depends on [n]: PCI [=n]
> Selected by [y]:
> - FSL_PCI [=y]
>
> and 30+ build errors.
>
> Fixes: 22fdf79171e8 ("powerpc/fsl_uli1575: Allow to disable FSL_ULI1575
> support")
> Signed-off-by: Randy Dunlap <rdun...@infradead.org>
> Cc: Pali Rohár <p...@kernel.org>
> Cc: Michael Ellerman <m...@ellerman.id.au>
> Cc: linuxppc-dev@lists.ozlabs.org
> ---
> arch/powerpc/platforms/Kconfig | 1 +
> 1 file changed, 1 insertion(+)
>
> diff -- a/arch/powerpc/platforms/Kconfig b/arch/powerpc/platforms/Kconfig
> --- a/arch/powerpc/platforms/Kconfig
> +++ b/arch/powerpc/platforms/Kconfig
> @@ -265,6 +265,7 @@ config CPM2
> config FSL_ULI1575
> bool "ULI1575 PCIe south bridge support"
> depends on FSL_SOC_BOOKE || PPC_86xx
> + depends on PCI
> select FSL_PCI
> select GENERIC_ISA_DMA
> help
Ok, looks like a reasonable solution for this issue.
These kind of cleanups (as done in uli1575 patch series) are always
problematic as they can break some unusual configuration...
Reviewed-by: Pali Rohár <p...@kernel.org>
