Somehow it went completely under my radar that v4.8 added another struct pci_platform_pm_ops for Intel Mobile Internet Devices. There's a commit queued for v4.9 on Bjorn's pci/pm branch which adds a mandatory ->get_power hook to pci_platform_pm_ops and implements it for acpi_pci_platform_pm, but not mid_pci_platform_pm. The following patch fixes that. It's not a build issue, the missing hook in mid_pci_platform_pm only becomes a problem at runtime. @Andy: Could you look over this and provide an ack? I do not have an Intel MID so I was only able to compile-test the patch. The ->get_power hook is currently only called to update a device's current_state after resume (both at runtime and after system sleep) and after changing its power state using the platform in pci_platform_power_transition(). The result of the ->get_power hook is ignored unless it's D3cold. Since intel-mid PCI devices can only be suspended to D3hot, the return value is irrelevant on this platform, the hook just needs to be present to make pci_set_platform_pm() happy. If you want to test the patch, you need to apply it either on Bjorn's pci/pm branch or linux-next: http://git.kernel.org/cgit/linux/kernel/git/helgaas/pci.git/log/?h=pci/pm Thanks, Lukas Lukas Wunner (1): x86/platform/intel-mid: Retrofit pci_platform_pm_ops ->get_power hook arch/x86/include/asm/intel-mid.h | 1 + arch/x86/platform/intel-mid/pwr.c | 19 +++++++++++++++++++ drivers/pci/pci-mid.c | 6 ++++++ 3 files changed, 26 insertions(+) -- 2.9.3 -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
