linux-kernel - Re: Possible regression caused by commit a192aa923b66a

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 12 Jun 2018 11:17:08 +0200
From:   "Rafael J. Wysocki" <rjw@...ysocki.net>
To:     Kai-Heng Feng <kai.heng.feng@...onical.com>
Cc:     Ulf Hansson <ulf.hansson@...aro.org>,
        P HeLiOn <perryhelionsemail@...il.com>,
        ACPI Devel Maling List <linux-acpi@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Mika Westerberg <mika.westerberg@...ux.intel.com>
Subject: Re: Possible regression caused by commit a192aa923b66a

On Monday, June 11, 2018 11:52:34 PM CEST Rafael J. Wysocki wrote:
> 
> --000000000000703623056e64c488
> Content-Type: text/plain; charset="UTF-8"
> 
> On Mon, Jun 11, 2018 at 10:09 AM, Rafael J. Wysocki <rafael@...nel.org> wrote:
> > On Mon, Jun 11, 2018 at 8:26 AM, Kai-Heng Feng
> > <kai.heng.feng@...onical.com> wrote:
> >> Hi Rafael,
> >>
> >> There's a regression report [1] that says commit a192aa923b66a ("ACPI /
> >> LPSS: Consolidate runtime PM and system sleep handling") is the first bad
> >> commit.
> >>
> >> From the looks of it, it didn't introduce any behavioral change. So your
> >> help is appreciated.
> >>
> >> [1] https://bugs.launchpad.net/bugs/1774950
> >
> > Well, the only difference is the iosf quirk AFAICS, but that should be
> > easy enough to check.  I'll try to cut a patch for that later today.
> 
> If the iosf quirk is the source of the problem, the attached patch should help.

The one below should be slightly better, please test this one.

---
 drivers/acpi/acpi_lpss.c |   18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

Index: linux-pm/drivers/acpi/acpi_lpss.c
===================================================================
--- linux-pm.orig/drivers/acpi/acpi_lpss.c
+++ linux-pm/drivers/acpi/acpi_lpss.c
@@ -22,6 +22,7 @@
 #include <linux/pm_domain.h>
 #include <linux/pm_runtime.h>
 #include <linux/pwm.h>
+#include <linux/suspend.h>
 #include <linux/delay.h>
 
 #include "internal.h"
@@ -940,9 +941,10 @@ static void lpss_iosf_exit_d3_state(void
 	mutex_unlock(&lpss_iosf_mutex);
 }
 
-static int acpi_lpss_suspend(struct device *dev, bool wakeup)
+static int acpi_lpss_suspend(struct device *dev, bool runtime)
 {
 	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+	bool wakeup = runtime || device_may_wakeup(dev);
 	int ret;
 
 	if (pdata->dev_desc->flags & LPSS_SAVE_CTX)
@@ -955,13 +957,14 @@ static int acpi_lpss_suspend(struct devi
 	 * wrong status for devices being about to be powered off. See
 	 * lpss_iosf_enter_d3_state() for further information.
 	 */
-	if (lpss_quirks & LPSS_QUIRK_ALWAYS_POWER_ON && iosf_mbi_available())
+	if ((runtime || !pm_suspend_via_firmware()) &&
+	    lpss_quirks & LPSS_QUIRK_ALWAYS_POWER_ON && iosf_mbi_available())
 		lpss_iosf_enter_d3_state();
 
 	return ret;
 }
 
-static int acpi_lpss_resume(struct device *dev)
+static int acpi_lpss_resume(struct device *dev, bool runtime)
 {
 	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
 	int ret;
@@ -970,7 +973,8 @@ static int acpi_lpss_resume(struct devic
 	 * This call is kept first to be in symmetry with
 	 * acpi_lpss_runtime_suspend() one.
 	 */
-	if (lpss_quirks & LPSS_QUIRK_ALWAYS_POWER_ON && iosf_mbi_available())
+	if ((runtime || !pm_resume_via_firmware()) &&
+	    lpss_quirks & LPSS_QUIRK_ALWAYS_POWER_ON && iosf_mbi_available())
 		lpss_iosf_exit_d3_state();
 
 	ret = acpi_dev_resume(dev);
@@ -994,12 +998,12 @@ static int acpi_lpss_suspend_late(struct
 		return 0;
 
 	ret = pm_generic_suspend_late(dev);
-	return ret ? ret : acpi_lpss_suspend(dev, device_may_wakeup(dev));
+	return ret ? ret : acpi_lpss_suspend(dev, false);
 }
 
 static int acpi_lpss_resume_early(struct device *dev)
 {
-	int ret = acpi_lpss_resume(dev);
+	int ret = acpi_lpss_resume(dev, false);
 
 	return ret ? ret : pm_generic_resume_early(dev);
 }
@@ -1014,7 +1018,7 @@ static int acpi_lpss_runtime_suspend(str
 
 static int acpi_lpss_runtime_resume(struct device *dev)
 {
-	int ret = acpi_lpss_resume(dev);
+	int ret = acpi_lpss_resume(dev, true);
 
 	return ret ? ret : pm_generic_runtime_resume(dev);
 }