From: Ulf Hansson ulf.hansson@linaro.org
In case the PM domain fails to be powered on in genpd_dev_pm_attach(), it returns -EPROBE_DEFER, but keeping the device attached to its PM domain. This leads to problems when the next attempt to attach is re-tried. More precisely, in that situation an -EEXIST error code is returned, because the device already has its PM domain pointer assigned, from the first attempt.
Now, because of the sloppy error handling by the existing callers of dev_pm_domain_attach(), probing is allowed to continue when -EEXIST is returned. However, in such case there are no guarantees that the PM domain is powered on by genpd, which may lead to hangs when buses/drivers tried to access their devices.
Let's fix this behaviour, simply by detaching the device when powering on fails in genpd_dev_pm_attach().
Cc: v4.11+ stable@vger.kernel.org # v4.11+ Signed-off-by: Ulf Hansson ulf.hansson@linaro.org Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com
(cherry picked from commit 72038df3c580c4c326b83c86149d7ac34007532a)
Cherry-picked to imx_4.9.y with minimal conflicts so that we can properly handle errors from SCFW pm calls Signed-off-by: Leonard Crestez leonard.crestez@nxp.com --- drivers/base/power/domain.c | 3 +++ 1 file changed, 3 insertions(+)
diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 9c3e535795a0..d9681d372997 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1936,10 +1936,13 @@ int genpd_dev_pm_attach(struct device *dev) dev->pm_domain->sync = genpd_dev_pm_sync;
mutex_lock(&pd->lock); ret = genpd_poweron(pd, 0); mutex_unlock(&pd->lock); + + if (ret) + genpd_remove_device(pd, dev); out: return ret ? -EPROBE_DEFER : 0; } EXPORT_SYMBOL_GPL(genpd_dev_pm_attach); #endif /* CONFIG_PM_GENERIC_DOMAINS_OF */
-----Original Message----- From: Leonard Crestez Sent: Monday, September 3, 2018 6:16 PM To: Anson Huang anson.huang@nxp.com; A.s. Dong aisheng.dong@nxp.com Cc: Ranjani Vaidyanathan ranjani.vaidyanathan@nxp.com; Nitin Garg nitin.garg@nxp.com; Jason Liu jason.hui.liu@nxp.com; Pete Zhang pete.zhang@nxp.com; Eagle Zhou eagle.zhou@nxp.com; LnxRevLi LnxRevLi@nxp.com; Ulf Hansson ulf.hansson@linaro.org; v4 . 11+ stable@vger.kernel.org; Rafael J . Wysocki rafael.j.wysocki@intel.com Subject: [PATCH imx_4.9.y 1/2] PM / Domains: Fix error path during attach in genpd
From: Ulf Hansson ulf.hansson@linaro.org
In case the PM domain fails to be powered on in genpd_dev_pm_attach(), it returns -EPROBE_DEFER, but keeping the device attached to its PM domain. This leads to problems when the next attempt to attach is re-tried. More precisely, in that situation an -EEXIST error code is returned, because the device already has its PM domain pointer assigned, from the first attempt.
Now, because of the sloppy error handling by the existing callers of dev_pm_domain_attach(), probing is allowed to continue when -EEXIST is returned. However, in such case there are no guarantees that the PM domain is powered on by genpd, which may lead to hangs when buses/drivers tried to access their devices.
Let's fix this behaviour, simply by detaching the device when powering on fails in genpd_dev_pm_attach().
Cc: v4.11+ stable@vger.kernel.org # v4.11+
Although this patch is stated only needed for v4.11+, but it seems to me 4.9 has the same issue and also needs it.
Reviewed-by: Dong Aisheng Aisheng.dong@nxp.com
Hi Leonard, I guess you could also try to send it to 4.9 LTS kernel maillist to test maintainer later.
Regards Dong Aisheng
Signed-off-by: Ulf Hansson ulf.hansson@linaro.org Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com
(cherry picked from commit 72038df3c580c4c326b83c86149d7ac34007532a)
Cherry-picked to imx_4.9.y with minimal conflicts so that we can properly handle errors from SCFW pm calls Signed-off-by: Leonard Crestez leonard.crestez@nxp.com
drivers/base/power/domain.c | 3 +++ 1 file changed, 3 insertions(+)
diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 9c3e535795a0..d9681d372997 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1936,10 +1936,13 @@ int genpd_dev_pm_attach(struct device *dev) dev->pm_domain->sync = genpd_dev_pm_sync;
mutex_lock(&pd->lock); ret = genpd_poweron(pd, 0); mutex_unlock(&pd->lock);
- if (ret)
genpd_remove_device(pd, dev);out: return ret ? -EPROBE_DEFER : 0; } EXPORT_SYMBOL_GPL(genpd_dev_pm_attach);
#endif /* CONFIG_PM_GENERIC_DOMAINS_OF */
2.17.1
-----Original Message----- From: Leonard Crestez Sent: Monday, September 3, 2018 6:16 PM To: Anson Huang anson.huang@nxp.com; A.s. Dong aisheng.dong@nxp.com Cc: Ranjani Vaidyanathan ranjani.vaidyanathan@nxp.com; Nitin Garg nitin.garg@nxp.com; Jason Liu jason.hui.liu@nxp.com; Pete Zhang pete.zhang@nxp.com; Eagle Zhou eagle.zhou@nxp.com; LnxRevLi LnxRevLi@nxp.com; Ulf Hansson ulf.hansson@linaro.org; v4 . 11+ stable@vger.kernel.org; Rafael J . Wysocki rafael.j.wysocki@intel.com Subject: [PATCH imx_4.9.y 1/2] PM / Domains: Fix error path during attach in genpd
From: Ulf Hansson ulf.hansson@linaro.org
In case the PM domain fails to be powered on in genpd_dev_pm_attach(), it returns -EPROBE_DEFER, but keeping the device attached to its PM domain. This leads to problems when the next attempt to attach is re-tried. More precisely, in that situation an -EEXIST error code is returned, because the device already has its PM domain pointer assigned, from the first attempt.
Now, because of the sloppy error handling by the existing callers of dev_pm_domain_attach(), probing is allowed to continue when -EEXIST is returned. However, in such case there are no guarantees that the PM domain is powered on by genpd, which may lead to hangs when buses/drivers tried to access their devices.
Let's fix this behaviour, simply by detaching the device when powering on fails in genpd_dev_pm_attach().
Cc: v4.11+ stable@vger.kernel.org # v4.11+ Signed-off-by: Ulf Hansson ulf.hansson@linaro.org Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com
(cherry picked from commit 72038df3c580c4c326b83c86149d7ac34007532a)
Cherry-picked to imx_4.9.y with minimal conflicts so that we can properly handle errors from SCFW pm calls Signed-off-by: Leonard Crestez leonard.crestez@nxp.com
drivers/base/power/domain.c | 3 +++ 1 file changed, 3 insertions(+)
diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 9c3e535795a0..d9681d372997 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1936,10 +1936,13 @@ int genpd_dev_pm_attach(struct device *dev) dev->pm_domain->sync = genpd_dev_pm_sync;
mutex_lock(&pd->lock); ret = genpd_poweron(pd, 0); mutex_unlock(&pd->lock);
- if (ret)
genpd_remove_device(pd, dev);out: return ret ? -EPROBE_DEFER : 0; } EXPORT_SYMBOL_GPL(genpd_dev_pm_attach); #endif /* CONFIG_PM_GENERIC_DOMAINS_OF */
BTW, it looks to me this is not a complete fix.
static int platform_drv_probe(struct device *_dev) {
ret = dev_pm_domain_attach(_dev, true); if (ret != -EPROBE_DEFER) { <---- here is risk as it's only hanlle EPROBE_DEFFER. Latest upstream is already changed. However, it may not cause real issue for our case. if (drv->probe) { ret = drv->probe(dev); if (ret) dev_pm_domain_detach(_dev, true); } else { /* don't fail if just dev_pm_domain_attach failed */ ret = 0; } } ...
return ret; }
-- 2.17.1
linux-stable-mirror@lists.linaro.org
-
A.s. Dong -
Leonard Crestez