mfd: Fix RTS5227 (and others) powermanagement
authorHans de Goede <hdegoede@redhat.com>
Wed, 6 Dec 2017 11:21:35 +0000 (12:21 +0100)
committerLee Jones <lee.jones@linaro.org>
Thu, 7 Dec 2017 10:40:57 +0000 (10:40 +0000)
Commit 8275b77a1513 ("mfd: rts5249: Add support for RTS5250S power
saving") adds powersaving support for device-ids 5249 524a and 525a.

But as a side effect it breaks ASPM support for all the other device-ids,
causing e.g. the Haswell CPU on a Lenovo T440s to not go into a higher
c-state then PC3, while previously it would go to PC7, causing the
machine to idle at 7.4W instead of 6.6W!

The problem here is the new option.dev_aspm_mode field, which only gets
explicitly initialized in the new code for the device-ids 5249 524a and
525a. Leaving the dev_aspm_mode 0 for the other device-ids.

The default dev_aspm_mode 0 is mapped to DEV_ASPM_DISABLE, but the
old behavior of calling rtsx_pci_enable_aspm() when idle and
rtsx_pci_disable_aspm() when busy happens when dev_aspm_mode ==
DEV_ASPM_DYNAMIC.

This commit changes the enum so that 0 = DEV_ASPM_DYNAMIC matching the
old default behavior, fixing the pm regression with the other device-ids.

Fixes: 8275b77a1513 ("mfd: rts5249: Add support for RTS5250S power saving")
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Rui Feng <rui_feng@realsil.com.cn>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
include/linux/mfd/rtsx_pci.h

index a2a1318a3d0c8be0a1fb3d1a08fcf671ff9d8bee..c3d3f04d8cc689eddf217c0626e71d8c16530db5 100644 (file)
@@ -915,10 +915,10 @@ enum PDEV_STAT  {PDEV_STAT_IDLE, PDEV_STAT_RUN};
 #define LTR_L1SS_PWR_GATE_CHECK_CARD_EN        BIT(6)
 
 enum dev_aspm_mode {
-       DEV_ASPM_DISABLE = 0,
        DEV_ASPM_DYNAMIC,
        DEV_ASPM_BACKDOOR,
        DEV_ASPM_STATIC,
+       DEV_ASPM_DISABLE,
 };
 
 /*