mmc: sdhci-acpi: Only powered up enabled acpi child devices
authorHans de Goede <hdegoede@redhat.com>
Tue, 20 Dec 2016 23:19:19 +0000 (00:19 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 26 Jan 2017 07:24:38 +0000 (08:24 +0100)
commitc1274eeb2fe3524ed075d3343213923262c845c8
tree5b5eab569201adeddcdcd0b64e87b00c9b6243f8
parent35028599bc8884f33e36a9353be6f1401940aa29
mmc: sdhci-acpi: Only powered up enabled acpi child devices

commit e1d070c3793a2766122865a7c2142853b48808c5 upstream.

Commit e5bbf30733f9 ("mmc: sdhci-acpi: Ensure connected devices are
powered when probing") introduced code to powerup any acpi child
nodes listed in the dstd. But some dstd-s list all possible devices
used on some board variants, while reporting if the device is actually
present and enabled in the status field of the device.

So we end up calling the acpi _PS0 (power-on) method for devices which
are not actually present. This does not always end well, e.g. on my
cube iwork8 air tablet, this results in freezing the entire tablet as
soon as the r8723bs module is loaded.

This commit fixes this by checking the child device's status.present
and status.enabled bits and only call acpi_device_fix_up_power()
if both are set.

Fixes: e5bbf30733f9 ("mmc: sdhci-acpi: Ensure connected devices are powered when probing")
BugLink: https://github.com/hadess/rtl8723bs/issues/80
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/mmc/host/sdhci-acpi.c