diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2017-07-14 22:24:25 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2017-07-14 22:24:25 -0700 |
commit | e37720e25df1cb70958a59716135385683e728b3 (patch) | |
tree | db852b7c4a35164ea54fd5c3bcc56c787bbde958 /drivers/pci/pci.c | |
parent | 867eacd7fb975273703766f52f485f08471a1ae9 (diff) | |
parent | 6df609b24b23845d2ad85b01814e7ad27f607e92 (diff) | |
download | talos-op-linux-e37720e25df1cb70958a59716135385683e728b3.tar.gz talos-op-linux-e37720e25df1cb70958a59716135385683e728b3.zip |
Merge tag 'pm-fixes-4.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management fixes from Rafael Wysocki:
"These fix a recently exposed issue in the PCI device wakeup code and
one older problem related to PCI device wakeup that has been reported
recently, modify one more piece of computations in intel_pstate to get
rid of a rounding error, fix a possible race in the schedutil cpufreq
governor, fix the device PM QoS sysfs interface to correctly handle
invalid user input, fix return values of two probe routines in devfreq
drivers and constify an attribute_group structure in devfreq.
Specifics:
- Avoid clearing the PCI PME Enable bit for devices as a result of
config space restoration which confuses AML executed afterward and
causes wakeup events to be lost on some systems (Rafael Wysocki).
- Fix the native PCIe PME interrupts handling in the cases when the
PME IRQ is set up as a system wakeup one so that runtime PM remote
wakeup works as expected after system resume on systems where that
happens (Rafael Wysocki).
- Fix the device PM QoS sysfs interface to handle invalid user input
correctly instead of using an unititialized variable value as the
latency tolerance for the device at hand (Dan Carpenter).
- Get rid of one more rounding error from intel_pstate computations
(Srinivas Pandruvada).
- Fix the schedutil cpufreq governor to prevent it from possibly
accessing unititialized data structures from governor callbacks in
some cases on systems when multiple CPUs share a single cpufreq
policy object (Vikram Mulukutla).
- Fix the return values of probe routines in two devfreq drivers
(Gustavo Silva).
- Constify an attribute_group structure in devfreq (Arvind Yadav)"
* tag 'pm-fixes-4.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
PCI / PM: Fix native PME handling during system suspend/resume
PCI / PM: Restore PME Enable after config space restoration
cpufreq: schedutil: Fix sugov_start() versus sugov_update_shared() race
PM / QoS: return -EINVAL for bogus strings
cpufreq: intel_pstate: Fix ratio setting for min_perf_pct
PM / devfreq: constify attribute_group structures.
PM / devfreq: tegra: fix error return code in tegra_devfreq_probe()
PM / devfreq: rk3399_dmc: fix error return code in rk3399_dmcfreq_probe()
Diffstat (limited to 'drivers/pci/pci.c')
-rw-r--r-- | drivers/pci/pci.c | 16 |
1 files changed, 8 insertions, 8 deletions
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index d88edf5c563b..af0cc3456dc1 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -1801,7 +1801,11 @@ static void __pci_pme_active(struct pci_dev *dev, bool enable) pci_write_config_word(dev, dev->pm_cap + PCI_PM_CTRL, pmcsr); } -static void pci_pme_restore(struct pci_dev *dev) +/** + * pci_pme_restore - Restore PME configuration after config space restore. + * @dev: PCI device to update. + */ +void pci_pme_restore(struct pci_dev *dev) { u16 pmcsr; @@ -1811,6 +1815,7 @@ static void pci_pme_restore(struct pci_dev *dev) pci_read_config_word(dev, dev->pm_cap + PCI_PM_CTRL, &pmcsr); if (dev->wakeup_prepared) { pmcsr |= PCI_PM_CTRL_PME_ENABLE; + pmcsr &= ~PCI_PM_CTRL_PME_STATUS; } else { pmcsr &= ~PCI_PM_CTRL_PME_ENABLE; pmcsr |= PCI_PM_CTRL_PME_STATUS; @@ -1907,14 +1912,9 @@ int pci_enable_wake(struct pci_dev *dev, pci_power_t state, bool enable) { int ret = 0; - /* - * Don't do the same thing twice in a row for one device, but restore - * PME Enable in case it has been updated by config space restoration. - */ - if (!!enable == !!dev->wakeup_prepared) { - pci_pme_restore(dev); + /* Don't do the same thing twice in a row for one device. */ + if (!!enable == !!dev->wakeup_prepared) return 0; - } /* * According to "PCI System Architecture" 4th ed. by Tom Shanley & Don |