| Message ID | 20140911222402.16809.94145.stgit@bhelgaas-glaptop2.roam.corp.google.com |
|---|---|
| State | Accepted |
| Headers | show |
Hi, On Thu, Sep 11, 2014 at 3:24 PM, Bjorn Helgaas <bhelgaas@google.com> wrote: > Powering off a hot-pluggable device, e.g., with pci_set_power_state(D3cold), > normally generates a hot-remove event that unbinds the driver. > > Some drivers expect to remain bound to a device even while they power it > off and back on again. This can be dangerous, because if the device is > removed or replaced while it is powered off, the driver doesn't know that > anything changed. But some drivers accept that risk. > > Add pci_ignore_hotplug() for use by drivers that know their device cannot > be removed. Using pci_ignore_hotplug() tells the PCI core that hot-plug > events for the device should be ignored. > > The radeon and nouveau drivers use this to switch between a low-power, > integrated GPU and a higher-power, higher-performance discrete GPU. They > power off the unused GPU, but they want to remain bound to it. > > This is a reimplementation of f244d8b623da ("ACPIPHP / radeon / nouveau: > Fix VGA switcheroo problem related to hotplug") but extends it to work with > both acpiphp and pciehp. > > This fixes a problem where systems with dual GPUs using the radeon drivers > become unusable, freezing every few seconds (see bugzillas below). The > resume of the radeon device may also fail, e.g., > > This fixes problems on dual GPU systems where the radeon driver becomes > unusable because of problems while suspending the device, as in bug 79701: > > [drm] radeon: finishing device. > radeon 0000:01:00.0: Userspace still has active objects ! > radeon 0000:01:00.0: ffff8800cb4ec288 ffff8800cb4ec000 16384 4294967297 force free > ... > WARNING: CPU: 0 PID: 67 at /home/apw/COD/linux/drivers/gpu/drm/radeon/radeon_gart.c:234 radeon_gart_unbind+0xd2/0xe0 [radeon]() > trying to unbind memory from uninitialized GART ! > > or while resuming it, as in bug 77261: > > radeon 0000:01:00.0: ring 0 stalled for more than 10158msec > radeon 0000:01:00.0: GPU lockup ... > radeon 0000:01:00.0: GPU pci config reset > pciehp 0000:00:01.0:pcie04: Card not present on Slot(1-1) > radeon 0000:01:00.0: GPU reset succeeded, trying to resume > *ERROR* radeon: dpm resume failed > radeon 0000:01:00.0: Wait for MC idle timedout ! > > Link: https://bugzilla.kernel.org/show_bug.cgi?id=77261 > Link: https://bugzilla.kernel.org/show_bug.cgi?id=79701 > Reported-by: Shawn Starr <shawn.starr@rogers.com> > Reported-by: Jose P. <lbdkmjdf@sharklasers.com> > Tested-by: SpacemanSpiff <a818958@trbvm.com> > Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> > CC: stable@vger.kernel.org # v3.15+ Acked-by: Rajat Jain <rajatxjain@gmail.com> Just some additional notes (not related to the current problem). I think there are other places where we could make use of the same newly introduced pci_ignore_hotplug(): https://lkml.org/lkml/2013/12/15/151 I think that at least there are 2 places where this can potentially be utilized: 1) While handling AER, we may reset the link while trying to do a recovery. This will result in a hot-remove followed by a hot-add. We may utilize this call in that situation (will also need a pci_resume_hotplug() call to cleat the ignore_hotplug flag) 2) Some drivers [mpt3sas_base.c: _base_diag_reset()] may reset the device during a firmware upgrade or diagnostics. Thanks, Rajat > --- > drivers/gpu/drm/nouveau/nouveau_drm.c | 1 + > drivers/gpu/drm/radeon/radeon_drv.c | 1 + > drivers/pci/hotplug/acpiphp_glue.c | 16 ++++++---------- > drivers/pci/hotplug/pciehp_hpc.c | 12 ++++++++++++ > include/linux/pci.h | 6 ++++++ > 5 files changed, 26 insertions(+), 10 deletions(-) > > diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c > index 250a5e88c751..9c3af96a7153 100644 > --- a/drivers/gpu/drm/nouveau/nouveau_drm.c > +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c > @@ -627,6 +627,7 @@ int nouveau_pmops_suspend(struct device *dev) > > pci_save_state(pdev); > pci_disable_device(pdev); > + pci_ignore_hotplug(pdev); > pci_set_power_state(pdev, PCI_D3hot); > return 0; > } > diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c > index 8df888908833..abbd87adfd75 100644 > --- a/drivers/gpu/drm/radeon/radeon_drv.c > +++ b/drivers/gpu/drm/radeon/radeon_drv.c > @@ -440,6 +440,7 @@ static int radeon_pmops_runtime_suspend(struct device *dev) > ret = radeon_suspend_kms(drm_dev, false, false); > pci_save_state(pdev); > pci_disable_device(pdev); > + pci_ignore_hotplug(pdev); > pci_set_power_state(pdev, PCI_D3cold); > drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF; > > diff --git a/drivers/pci/hotplug/acpiphp_glue.c b/drivers/pci/hotplug/acpiphp_glue.c > index 70741c8c46a0..6cd5160fc057 100644 > --- a/drivers/pci/hotplug/acpiphp_glue.c > +++ b/drivers/pci/hotplug/acpiphp_glue.c > @@ -560,19 +560,15 @@ static void disable_slot(struct acpiphp_slot *slot) > slot->flags &= (~SLOT_ENABLED); > } > > -static bool acpiphp_no_hotplug(struct acpi_device *adev) > -{ > - return adev && adev->flags.no_hotplug; > -} > - > static bool slot_no_hotplug(struct acpiphp_slot *slot) > { > - struct acpiphp_func *func; > + struct pci_bus *bus = slot->bus; > + struct pci_dev *dev; > > - list_for_each_entry(func, &slot->funcs, sibling) > - if (acpiphp_no_hotplug(func_to_acpi_device(func))) > + list_for_each_entry(dev, &bus->devices, bus_list) { > + if (PCI_SLOT(dev->devfn) == slot->device && dev->ignore_hotplug) > return true; > - > + } > return false; > } > > @@ -645,7 +641,7 @@ static void trim_stale_devices(struct pci_dev *dev) > > status = acpi_evaluate_integer(adev->handle, "_STA", NULL, &sta); > alive = (ACPI_SUCCESS(status) && device_status_valid(sta)) > - || acpiphp_no_hotplug(adev); > + || dev->ignore_hotplug; > } > if (!alive) > alive = pci_device_is_present(dev); > diff --git a/drivers/pci/hotplug/pciehp_hpc.c b/drivers/pci/hotplug/pciehp_hpc.c > index 9da84b8b27d8..5e01ae39ec46 100644 > --- a/drivers/pci/hotplug/pciehp_hpc.c > +++ b/drivers/pci/hotplug/pciehp_hpc.c > @@ -506,6 +506,8 @@ static irqreturn_t pcie_isr(int irq, void *dev_id) > { > struct controller *ctrl = (struct controller *)dev_id; > struct pci_dev *pdev = ctrl_dev(ctrl); > + struct pci_bus *subordinate = pdev->subordinate; > + struct pci_dev *dev; > struct slot *slot = ctrl->slot; > u16 detected, intr_loc; > > @@ -539,6 +541,16 @@ static irqreturn_t pcie_isr(int irq, void *dev_id) > wake_up(&ctrl->queue); > } > > + if (subordinate) { > + list_for_each_entry(dev, &subordinate->devices, bus_list) { > + if (dev->ignore_hotplug) { > + ctrl_dbg(ctrl, "ignoring hotplug event %#06x (%s requested no hotplug)\n", > + intr_loc, pci_name(dev)); > + return IRQ_HANDLED; > + } > + } > + } > + > if (!(intr_loc & ~PCI_EXP_SLTSTA_CC)) > return IRQ_HANDLED; > > diff --git a/include/linux/pci.h b/include/linux/pci.h > index 61978a460841..96453f9bc8ba 100644 > --- a/include/linux/pci.h > +++ b/include/linux/pci.h > @@ -303,6 +303,7 @@ struct pci_dev { > D3cold, not set for devices > powered on/off by the > corresponding bridge */ > + unsigned int ignore_hotplug:1; /* Ignore hotplug events */ > unsigned int d3_delay; /* D3->D0 transition time in ms */ > unsigned int d3cold_delay; /* D3cold->D0 transition time in ms */ > > @@ -1021,6 +1022,11 @@ bool pci_dev_run_wake(struct pci_dev *dev); > bool pci_check_pme_status(struct pci_dev *dev); > void pci_pme_wakeup_bus(struct pci_bus *bus); > > +static inline void pci_ignore_hotplug(struct pci_dev *dev) > +{ > + dev->ignore_hotplug = 1; > +} > + > static inline int pci_enable_wake(struct pci_dev *dev, pci_power_t state, > bool enable) > { > -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index 250a5e88c751..9c3af96a7153 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -627,6 +627,7 @@ int nouveau_pmops_suspend(struct device *dev) pci_save_state(pdev); pci_disable_device(pdev); + pci_ignore_hotplug(pdev); pci_set_power_state(pdev, PCI_D3hot); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index 8df888908833..abbd87adfd75 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -440,6 +440,7 @@ static int radeon_pmops_runtime_suspend(struct device *dev) ret = radeon_suspend_kms(drm_dev, false, false); pci_save_state(pdev); pci_disable_device(pdev); + pci_ignore_hotplug(pdev); pci_set_power_state(pdev, PCI_D3cold); drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF; diff --git a/drivers/pci/hotplug/acpiphp_glue.c b/drivers/pci/hotplug/acpiphp_glue.c index 70741c8c46a0..6cd5160fc057 100644 --- a/drivers/pci/hotplug/acpiphp_glue.c +++ b/drivers/pci/hotplug/acpiphp_glue.c @@ -560,19 +560,15 @@ static void disable_slot(struct acpiphp_slot *slot) slot->flags &= (~SLOT_ENABLED); } -static bool acpiphp_no_hotplug(struct acpi_device *adev) -{ - return adev && adev->flags.no_hotplug; -} - static bool slot_no_hotplug(struct acpiphp_slot *slot) { - struct acpiphp_func *func; + struct pci_bus *bus = slot->bus; + struct pci_dev *dev; - list_for_each_entry(func, &slot->funcs, sibling) - if (acpiphp_no_hotplug(func_to_acpi_device(func))) + list_for_each_entry(dev, &bus->devices, bus_list) { + if (PCI_SLOT(dev->devfn) == slot->device && dev->ignore_hotplug) return true; - + } return false; } @@ -645,7 +641,7 @@ static void trim_stale_devices(struct pci_dev *dev) status = acpi_evaluate_integer(adev->handle, "_STA", NULL, &sta); alive = (ACPI_SUCCESS(status) && device_status_valid(sta)) - || acpiphp_no_hotplug(adev); + || dev->ignore_hotplug; } if (!alive) alive = pci_device_is_present(dev); diff --git a/drivers/pci/hotplug/pciehp_hpc.c b/drivers/pci/hotplug/pciehp_hpc.c index 9da84b8b27d8..5e01ae39ec46 100644 --- a/drivers/pci/hotplug/pciehp_hpc.c +++ b/drivers/pci/hotplug/pciehp_hpc.c @@ -506,6 +506,8 @@ static irqreturn_t pcie_isr(int irq, void *dev_id) { struct controller *ctrl = (struct controller *)dev_id; struct pci_dev *pdev = ctrl_dev(ctrl); + struct pci_bus *subordinate = pdev->subordinate; + struct pci_dev *dev; struct slot *slot = ctrl->slot; u16 detected, intr_loc; @@ -539,6 +541,16 @@ static irqreturn_t pcie_isr(int irq, void *dev_id) wake_up(&ctrl->queue); } + if (subordinate) { + list_for_each_entry(dev, &subordinate->devices, bus_list) { + if (dev->ignore_hotplug) { + ctrl_dbg(ctrl, "ignoring hotplug event %#06x (%s requested no hotplug)\n", + intr_loc, pci_name(dev)); + return IRQ_HANDLED; + } + } + } + if (!(intr_loc & ~PCI_EXP_SLTSTA_CC)) return IRQ_HANDLED; diff --git a/include/linux/pci.h b/include/linux/pci.h index 61978a460841..96453f9bc8ba 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -303,6 +303,7 @@ struct pci_dev { D3cold, not set for devices powered on/off by the corresponding bridge */ + unsigned int ignore_hotplug:1; /* Ignore hotplug events */ unsigned int d3_delay; /* D3->D0 transition time in ms */ unsigned int d3cold_delay; /* D3cold->D0 transition time in ms */ @@ -1021,6 +1022,11 @@ bool pci_dev_run_wake(struct pci_dev *dev); bool pci_check_pme_status(struct pci_dev *dev); void pci_pme_wakeup_bus(struct pci_bus *bus); +static inline void pci_ignore_hotplug(struct pci_dev *dev) +{ + dev->ignore_hotplug = 1; +} + static inline int pci_enable_wake(struct pci_dev *dev, pci_power_t state, bool enable) {
Powering off a hot-pluggable device, e.g., with pci_set_power_state(D3cold), normally generates a hot-remove event that unbinds the driver. Some drivers expect to remain bound to a device even while they power it off and back on again. This can be dangerous, because if the device is removed or replaced while it is powered off, the driver doesn't know that anything changed. But some drivers accept that risk. Add pci_ignore_hotplug() for use by drivers that know their device cannot be removed. Using pci_ignore_hotplug() tells the PCI core that hot-plug events for the device should be ignored. The radeon and nouveau drivers use this to switch between a low-power, integrated GPU and a higher-power, higher-performance discrete GPU. They power off the unused GPU, but they want to remain bound to it. This is a reimplementation of f244d8b623da ("ACPIPHP / radeon / nouveau: Fix VGA switcheroo problem related to hotplug") but extends it to work with both acpiphp and pciehp. This fixes a problem where systems with dual GPUs using the radeon drivers become unusable, freezing every few seconds (see bugzillas below). The resume of the radeon device may also fail, e.g., This fixes problems on dual GPU systems where the radeon driver becomes unusable because of problems while suspending the device, as in bug 79701: [drm] radeon: finishing device. radeon 0000:01:00.0: Userspace still has active objects ! radeon 0000:01:00.0: ffff8800cb4ec288 ffff8800cb4ec000 16384 4294967297 force free ... WARNING: CPU: 0 PID: 67 at /home/apw/COD/linux/drivers/gpu/drm/radeon/radeon_gart.c:234 radeon_gart_unbind+0xd2/0xe0 [radeon]() trying to unbind memory from uninitialized GART ! or while resuming it, as in bug 77261: radeon 0000:01:00.0: ring 0 stalled for more than 10158msec radeon 0000:01:00.0: GPU lockup ... radeon 0000:01:00.0: GPU pci config reset pciehp 0000:00:01.0:pcie04: Card not present on Slot(1-1) radeon 0000:01:00.0: GPU reset succeeded, trying to resume *ERROR* radeon: dpm resume failed radeon 0000:01:00.0: Wait for MC idle timedout ! Link: https://bugzilla.kernel.org/show_bug.cgi?id=77261 Link: https://bugzilla.kernel.org/show_bug.cgi?id=79701 Reported-by: Shawn Starr <shawn.starr@rogers.com> Reported-by: Jose P. <lbdkmjdf@sharklasers.com> Tested-by: SpacemanSpiff <a818958@trbvm.com> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> CC: stable@vger.kernel.org # v3.15+ --- drivers/gpu/drm/nouveau/nouveau_drm.c | 1 + drivers/gpu/drm/radeon/radeon_drv.c | 1 + drivers/pci/hotplug/acpiphp_glue.c | 16 ++++++---------- drivers/pci/hotplug/pciehp_hpc.c | 12 ++++++++++++ include/linux/pci.h | 6 ++++++ 5 files changed, 26 insertions(+), 10 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html