July 2025 - Linux-stable-mirror

[PATCH v2] PCI/pwrctrl: Skip creating pwrctrl device unless CONFIG_PCI_PWRCTRL is enabled

by Manivannan Sadhasivam

If devicetree describes power supplies related to a PCI device, we previously created a pwrctrl device even if CONFIG_PCI_PWRCTL was not enabled. When pci_pwrctrl_create_device() creates and returns a pwrctrl device, pci_scan_device() doesn't enumerate the PCI device. It assumes the pwrctrl core will rescan the bus after turning on the power. However, if CONFIG_PCI_PWRCTL is not enabled, the rescan never happens. This may break PCI enumeration on any system that describes power supplies in devicetree but does not use pwrctrl. Jim reported that some brcmstb platforms break this way. While the actual fix would be to convert all the platforms to use pwrctrl framework, we also need to skip creating the pwrctrl device if CONFIG_PCI_PWRCTL is not enabled and let the PCI core scan the device normally (assuming it is already powered on or by the controller driver). Cc: stable(a)vger.kernel.org # 6.15 Fixes: 957f40d039a9 ("PCI/pwrctrl: Move creation of pwrctrl devices to pci_scan_device()") Reported-by: Jim Quinlan <james.quinlan(a)broadcom.com> Closes: https://lore.kernel.org/r/CA+-6iNwgaByXEYD3j=-+H_PKAxXRU78svPMRHDKKci8AGXAU… Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam(a)linaro.org> --- Changes in v2: * Used the stub instead of returning NULL inside the function drivers/pci/probe.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index 4b8693ec9e4c..e6a34db77826 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -2508,6 +2508,7 @@ bool pci_bus_read_dev_vendor_id(struct pci_bus *bus, int devfn, u32 *l, } EXPORT_SYMBOL(pci_bus_read_dev_vendor_id); +#if IS_ENABLED(CONFIG_PCI_PWRCTRL) static struct platform_device *pci_pwrctrl_create_device(struct pci_bus *bus, int devfn) { struct pci_host_bridge *host = pci_find_host_bridge(bus); @@ -2537,6 +2538,12 @@ static struct platform_device *pci_pwrctrl_create_device(struct pci_bus *bus, in return pdev; } +#else +static struct platform_device *pci_pwrctrl_create_device(struct pci_bus *bus, int devfn) +{ + return NULL; +} +#endif /* * Read the config data for a PCI device, sanity-check it, -- 2.43.0

5 months

4
12
0 0

[PATCH 1/3] PCI/pwrctrl: Fix device leak at registration

by Johan Hovold

Make sure to drop the reference to the pwrctrl device taken by of_find_device_by_node() when registering a PCI device. Fixes: b458ff7e8176 ("PCI/pwrctl: Ensure that pwrctl drivers are probed before PCI client drivers") Cc: stable(a)vger.kernel.org # 6.13 Cc: Manivannan Sadhasivam <mani(a)kernel.org> Signed-off-by: Johan Hovold <johan+linaro(a)kernel.org> --- drivers/pci/bus.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c index 69048869ef1c..0394a9c77b38 100644 --- a/drivers/pci/bus.c +++ b/drivers/pci/bus.c @@ -362,11 +362,15 @@ void pci_bus_add_device(struct pci_dev *dev) * before PCI client drivers. */ pdev = of_find_device_by_node(dn); - if (pdev && of_pci_supply_present(dn)) { - if (!device_link_add(&dev->dev, &pdev->dev, - DL_FLAG_AUTOREMOVE_CONSUMER)) - pci_err(dev, "failed to add device link to power control device %s\n", - pdev->name); + if (pdev) { + if (of_pci_supply_present(dn)) { + if (!device_link_add(&dev->dev, &pdev->dev, + DL_FLAG_AUTOREMOVE_CONSUMER)) { + pci_err(dev, "failed to add device link to power control device %s\n", + pdev->name); + } + } + put_device(&pdev->dev); } if (!dn || of_device_is_available(dn)) -- 2.49.1

5 months

4
3
0 0

[PATCH net] net: core: Fix the loop in default_device_exit_net()

by Haiyang Zhang

From: Haiyang Zhang <haiyangz(a)microsoft.com> The loop in default_device_exit_net() won't be able to properly detect the head then stop, and will hit NULL pointer, when a driver, like hv_netvsc, automatically moves the slave device together with the master device. To fix this, add a helper function to return the first migratable netdev correctly, no matter one or two devices were removed from this net's list in the last iteration. Cc: stable(a)vger.kernel.org # 5.4+ Signed-off-by: Haiyang Zhang <haiyangz(a)microsoft.com> --- net/core/dev.c | 31 +++++++++++++++++++++---------- 1 file changed, 21 insertions(+), 10 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 621a639aeba1..d83f5f12cf70 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -12629,19 +12629,11 @@ static struct pernet_operations __net_initdata netdev_net_ops = { .exit = netdev_exit, }; -static void __net_exit default_device_exit_net(struct net *net) +static inline struct net_device *first_migratable_netdev(struct net *net) { - struct netdev_name_node *name_node, *tmp; struct net_device *dev, *aux; - /* - * Push all migratable network devices back to the - * initial network namespace - */ - ASSERT_RTNL(); - for_each_netdev_safe(net, dev, aux) { - int err; - char fb_name[IFNAMSIZ]; + for_each_netdev_safe(net, dev, aux) { /* Ignore unmoveable devices (i.e. loopback) */ if (dev->netns_immutable) continue; @@ -12650,6 +12642,25 @@ static void __net_exit default_device_exit_net(struct net *net) if (dev->rtnl_link_ops && !dev->rtnl_link_ops->netns_refund) continue; + return dev; + } + + return NULL; +} + +static void __net_exit default_device_exit_net(struct net *net) +{ + struct netdev_name_node *name_node, *tmp; + struct net_device *dev; + /* + * Push all migratable network devices back to the + * initial network namespace + */ + ASSERT_RTNL(); + while ((dev = first_migratable_netdev(net)) != NULL) { + int err; + char fb_name[IFNAMSIZ]; + /* Push remaining network devices to init_net */ snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex); if (netdev_name_in_use(&init_net, fb_name)) -- 2.34.1

5 months

3
2
0 0

[PATCH v2] sunvdc: Balance device refcount in vdc_port_mpgroup_check

by Ma Ke

Using device_find_child() to locate a probed virtual-device-port node causes a device refcount imbalance, as device_find_child() internally calls get_device() to increment the device’s reference count before returning its pointer. vdc_port_mpgroup_check() directly returns true upon finding a matching device without releasing the reference via put_device(). We should call put_device() to decrement refcount. As comment of device_find_child() says, 'NOTE: you will need to drop the reference with put_device() after use'. Found by code review. Cc: stable(a)vger.kernel.org Fixes: 3ee70591d6c4 ("sunvdc: prevent sunvdc panic when mpgroup disk added to guest domain") Signed-off-by: Ma Ke <make24(a)iscas.ac.cn> --- Changes in v2: - keep the change style simple as suggestions. --- drivers/block/sunvdc.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/block/sunvdc.c b/drivers/block/sunvdc.c index b5727dea15bd..7af21fe67671 100644 --- a/drivers/block/sunvdc.c +++ b/drivers/block/sunvdc.c @@ -957,8 +957,10 @@ static bool vdc_port_mpgroup_check(struct vio_dev *vdev) dev = device_find_child(vdev->dev.parent, &port_data, vdc_device_probed); - if (dev) + if (dev) { + put_device(dev); return true; + } return false; } -- 2.25.1

5 months

2
1
0 0

[PATCH] comedi: fix race between polling and detaching

by Ian Abbott

syzbot reports a use-after-free in comedi in the below link, which is due to comedi gladly removing the allocated async area even though poll requests are still active on the wait_queue_head inside of it. This can cause a use-after-free when the poll entries are later triggered or removed, as the memory for the wait_queue_head has been freed. We need to check there are no tasks queued on any of the subdevices' wait queues before allowing the device to be detached by the `COMEDI_DEVCONFIG` ioctl. Tasks will read-lock `dev->attach_lock` before adding themselves to the subdevice wait queue, so fix the problem in the `COMEDI_DEVCONFIG` ioctl handler by write-locking `dev->attach_lock` before checking that all of the subdevices are safe to be deleted. This includes testing for any sleepers on the subdevices' wait queues. It remains locked until the device has been detached. This requires the `comedi_device_detach()` function to be refactored slightly, moving the bulk of it into new function `comedi_device_detach_locked()`. Note that the refactor of `comedi_device_detach()` results in `comedi_device_cancel_all()` now being called while `dev->attach_lock` is write-locked, which wasn't the case previously, but that does not matter. Thanks to Jens Axboe for diagnosing the problem and co-developing this patch. Cc: <stable(a)vger.kernel.org> # 5.13+ Fixes: 2f3fdcd7ce93 ("staging: comedi: add rw_semaphore to protect against device detachment") Link: https://lore.kernel.org/all/687bd5fe.a70a0220.693ce.0091.GAE@google.com/ Reported-by: syzbot+01523a0ae5600aef5895(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=01523a0ae5600aef5895 Co-developed-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Ian Abbott <abbotti(a)mev.co.uk> --- Patch does not apply cleanly to longterm release series 5.10 and 5.4. --- drivers/comedi/comedi_fops.c | 31 ++++++++++++++++++++++++------- drivers/comedi/comedi_internal.h | 1 + drivers/comedi/drivers.c | 13 ++++++++++--- 3 files changed, 35 insertions(+), 10 deletions(-) diff --git a/drivers/comedi/comedi_fops.c b/drivers/comedi/comedi_fops.c index 3383a7ce27ff..3007fe946e42 100644 --- a/drivers/comedi/comedi_fops.c +++ b/drivers/comedi/comedi_fops.c @@ -787,6 +787,7 @@ static int is_device_busy(struct comedi_device *dev) struct comedi_subdevice *s; int i; + lockdep_assert_held_write(&dev->attach_lock); lockdep_assert_held(&dev->mutex); if (!dev->attached) return 0; @@ -795,7 +796,16 @@ static int is_device_busy(struct comedi_device *dev) s = &dev->subdevices[i]; if (s->busy) return 1; - if (s->async && comedi_buf_is_mmapped(s)) + if (!s->async) + continue; + if (comedi_buf_is_mmapped(s)) + return 1; + /* + * There may be tasks still waiting on the subdevice's wait + * queue, although they should already be about to be removed + * from it since the subdevice has no active async command. + */ + if (wq_has_sleeper(&s->async->wait_head)) return 1; } @@ -825,15 +835,22 @@ static int do_devconfig_ioctl(struct comedi_device *dev, return -EPERM; if (!arg) { - if (is_device_busy(dev)) - return -EBUSY; + int rc = 0; + if (dev->attached) { - struct module *driver_module = dev->driver->module; + down_write(&dev->attach_lock); + if (is_device_busy(dev)) { + rc = -EBUSY; + } else { + struct module *driver_module = + dev->driver->module; - comedi_device_detach(dev); - module_put(driver_module); + comedi_device_detach_locked(dev); + module_put(driver_module); + } + up_write(&dev->attach_lock); } - return 0; + return rc; } if (copy_from_user(&it, arg, sizeof(it))) diff --git a/drivers/comedi/comedi_internal.h b/drivers/comedi/comedi_internal.h index 9b3631a654c8..cf10ba016ebc 100644 --- a/drivers/comedi/comedi_internal.h +++ b/drivers/comedi/comedi_internal.h @@ -50,6 +50,7 @@ extern struct mutex comedi_drivers_list_lock; int insn_inval(struct comedi_device *dev, struct comedi_subdevice *s, struct comedi_insn *insn, unsigned int *data); +void comedi_device_detach_locked(struct comedi_device *dev); void comedi_device_detach(struct comedi_device *dev); int comedi_device_attach(struct comedi_device *dev, struct comedi_devconfig *it); diff --git a/drivers/comedi/drivers.c b/drivers/comedi/drivers.c index 376130bfba8a..f5c2ee271d0e 100644 --- a/drivers/comedi/drivers.c +++ b/drivers/comedi/drivers.c @@ -158,7 +158,7 @@ static void comedi_device_detach_cleanup(struct comedi_device *dev) int i; struct comedi_subdevice *s; - lockdep_assert_held(&dev->attach_lock); + lockdep_assert_held_write(&dev->attach_lock); lockdep_assert_held(&dev->mutex); if (dev->subdevices) { for (i = 0; i < dev->n_subdevices; i++) { @@ -196,16 +196,23 @@ static void comedi_device_detach_cleanup(struct comedi_device *dev) comedi_clear_hw_dev(dev); } -void comedi_device_detach(struct comedi_device *dev) +void comedi_device_detach_locked(struct comedi_device *dev) { + lockdep_assert_held_write(&dev->attach_lock); lockdep_assert_held(&dev->mutex); comedi_device_cancel_all(dev); - down_write(&dev->attach_lock); dev->attached = false; dev->detach_count++; if (dev->driver) dev->driver->detach(dev); comedi_device_detach_cleanup(dev); +} + +void comedi_device_detach(struct comedi_device *dev) +{ + lockdep_assert_held(&dev->mutex); + down_write(&dev->attach_lock); + comedi_device_detach_locked(dev); up_write(&dev->attach_lock); } -- 2.47.2

5 months

2
1
0 0

FAILED: patch "[PATCH] tracing: Add down_write(trace_event_sem) when adding trace" failed to apply to 5.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.4.y git checkout FETCH_HEAD git cherry-pick -x b5e8acc14dcb314a9b61ff19dcd9fdd0d88f70df # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025072113-perjury-dynamite-23ba@gregkh' --subject-prefix 'PATCH 5.4.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From b5e8acc14dcb314a9b61ff19dcd9fdd0d88f70df Mon Sep 17 00:00:00 2001 From: Steven Rostedt <rostedt(a)goodmis.org> Date: Fri, 18 Jul 2025 22:31:58 -0400 Subject: [PATCH] tracing: Add down_write(trace_event_sem) when adding trace event MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When a module is loaded, it adds trace events defined by the module. It may also need to modify the modules trace printk formats to replace enum names with their values. If two modules are loaded at the same time, the adding of the event to the ftrace_events list can corrupt the walking of the list in the code that is modifying the printk format strings and crash the kernel. The addition of the event should take the trace_event_sem for write while it adds the new event. Also add a lockdep_assert_held() on that semaphore in __trace_add_event_dirs() as it iterates the list. Cc: stable(a)vger.kernel.org Cc: Mathieu Desnoyers <mathieu.desnoyers(a)efficios.com> Acked-by: Masami Hiramatsu (Google) <mhiramat(a)kernel.org> Link: https://lore.kernel.org/20250718223158.799bfc0c@batman.local.home Reported-by: Fusheng Huang(黄富生) <Fusheng.Huang(a)luxshare-ict.com> Closes: https://lore.kernel.org/all/20250717105007.46ccd18f@batman.local.home/ Fixes: 110bf2b764eb6 ("tracing: add protection around module events unload") Signed-off-by: Steven Rostedt (Google) <rostedt(a)goodmis.org> diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 120531268abf..d01e5c910ce1 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -3136,7 +3136,10 @@ __register_event(struct trace_event_call *call, struct module *mod) if (ret < 0) return ret; + down_write(&trace_event_sem); list_add(&call->list, &ftrace_events); + up_write(&trace_event_sem); + if (call->flags & TRACE_EVENT_FL_DYNAMIC) atomic_set(&call->refcnt, 0); else @@ -3750,6 +3753,8 @@ __trace_add_event_dirs(struct trace_array *tr) struct trace_event_call *call; int ret; + lockdep_assert_held(&trace_event_sem); + list_for_each_entry(call, &ftrace_events, list) { ret = __trace_add_new_event(call, tr); if (ret < 0)

5 months

2
1
0 0

FAILED: patch "[PATCH] tracing: Add down_write(trace_event_sem) when adding trace" failed to apply to 5.10-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.10-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y git checkout FETCH_HEAD git cherry-pick -x b5e8acc14dcb314a9b61ff19dcd9fdd0d88f70df # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025072113-nutty-other-f178@gregkh' --subject-prefix 'PATCH 5.10.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From b5e8acc14dcb314a9b61ff19dcd9fdd0d88f70df Mon Sep 17 00:00:00 2001 From: Steven Rostedt <rostedt(a)goodmis.org> Date: Fri, 18 Jul 2025 22:31:58 -0400 Subject: [PATCH] tracing: Add down_write(trace_event_sem) when adding trace event MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When a module is loaded, it adds trace events defined by the module. It may also need to modify the modules trace printk formats to replace enum names with their values. If two modules are loaded at the same time, the adding of the event to the ftrace_events list can corrupt the walking of the list in the code that is modifying the printk format strings and crash the kernel. The addition of the event should take the trace_event_sem for write while it adds the new event. Also add a lockdep_assert_held() on that semaphore in __trace_add_event_dirs() as it iterates the list. Cc: stable(a)vger.kernel.org Cc: Mathieu Desnoyers <mathieu.desnoyers(a)efficios.com> Acked-by: Masami Hiramatsu (Google) <mhiramat(a)kernel.org> Link: https://lore.kernel.org/20250718223158.799bfc0c@batman.local.home Reported-by: Fusheng Huang(黄富生) <Fusheng.Huang(a)luxshare-ict.com> Closes: https://lore.kernel.org/all/20250717105007.46ccd18f@batman.local.home/ Fixes: 110bf2b764eb6 ("tracing: add protection around module events unload") Signed-off-by: Steven Rostedt (Google) <rostedt(a)goodmis.org> diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 120531268abf..d01e5c910ce1 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -3136,7 +3136,10 @@ __register_event(struct trace_event_call *call, struct module *mod) if (ret < 0) return ret; + down_write(&trace_event_sem); list_add(&call->list, &ftrace_events); + up_write(&trace_event_sem); + if (call->flags & TRACE_EVENT_FL_DYNAMIC) atomic_set(&call->refcnt, 0); else @@ -3750,6 +3753,8 @@ __trace_add_event_dirs(struct trace_array *tr) struct trace_event_call *call; int ret; + lockdep_assert_held(&trace_event_sem); + list_for_each_entry(call, &ftrace_events, list) { ret = __trace_add_new_event(call, tr); if (ret < 0)

5 months

2
1
0 0

FAILED: patch "[PATCH] pmdomain: governor: Consider CPU latency tolerance from" failed to apply to 5.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.4.y git checkout FETCH_HEAD git cherry-pick -x 500ba33284416255b9a5b50ace24470b6fe77ea5 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025072114-matriarch-decline-2598@gregkh' --subject-prefix 'PATCH 5.4.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 500ba33284416255b9a5b50ace24470b6fe77ea5 Mon Sep 17 00:00:00 2001 From: Maulik Shah <maulik.shah(a)oss.qualcomm.com> Date: Wed, 9 Jul 2025 14:00:11 +0530 Subject: [PATCH] pmdomain: governor: Consider CPU latency tolerance from pm_domain_cpu_gov pm_domain_cpu_gov is selecting a cluster idle state but does not consider latency tolerance of child CPUs. This results in deeper cluster idle state whose latency does not meet latency tolerance requirement. Select deeper idle state only if global and device latency tolerance of all child CPUs meet. Test results on SM8750 with 300 usec PM-QoS on CPU0 which is less than domain idle state entry (2150) + exit (1983) usec latency mentioned in devicetree, demonstrate the issue. # echo 300 > /sys/devices/system/cpu/cpu0/power/pm_qos_resume_latency_us Before: (Usage is incrementing) ====== # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 29817 537 8 270 0 # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 30348 542 8 271 0 After: (Usage is not incrementing due to latency tolerance) ====== # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 39319 626 14 307 0 # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 39319 626 14 307 0 Signed-off-by: Maulik Shah <maulik.shah(a)oss.qualcomm.com> Fixes: e94999688e3a ("PM / Domains: Add genpd governor for CPUs") Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/r/20250709-pmdomain_qos-v2-1-976b12257899@oss.qualc… Signed-off-by: Ulf Hansson <ulf.hansson(a)linaro.org> diff --git a/drivers/pmdomain/governor.c b/drivers/pmdomain/governor.c index c1e148657c87..39359811a930 100644 --- a/drivers/pmdomain/governor.c +++ b/drivers/pmdomain/governor.c @@ -8,6 +8,7 @@ #include <linux/pm_domain.h> #include <linux/pm_qos.h> #include <linux/hrtimer.h> +#include <linux/cpu.h> #include <linux/cpuidle.h> #include <linux/cpumask.h> #include <linux/ktime.h> @@ -349,6 +350,8 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) struct cpuidle_device *dev; ktime_t domain_wakeup, next_hrtimer; ktime_t now = ktime_get(); + struct device *cpu_dev; + s64 cpu_constraint, global_constraint; s64 idle_duration_ns; int cpu, i; @@ -359,6 +362,7 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) if (!(genpd->flags & GENPD_FLAG_CPU_DOMAIN)) return true; + global_constraint = cpu_latency_qos_limit(); /* * Find the next wakeup for any of the online CPUs within the PM domain * and its subdomains. Note, we only need the genpd->cpus, as it already @@ -372,8 +376,16 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) if (ktime_before(next_hrtimer, domain_wakeup)) domain_wakeup = next_hrtimer; } + + cpu_dev = get_cpu_device(cpu); + if (cpu_dev) { + cpu_constraint = dev_pm_qos_raw_resume_latency(cpu_dev); + if (cpu_constraint < global_constraint) + global_constraint = cpu_constraint; + } } + global_constraint *= NSEC_PER_USEC; /* The minimum idle duration is from now - until the next wakeup. */ idle_duration_ns = ktime_to_ns(ktime_sub(domain_wakeup, now)); if (idle_duration_ns <= 0) @@ -389,8 +401,10 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) */ i = genpd->state_idx; do { - if (idle_duration_ns >= (genpd->states[i].residency_ns + - genpd->states[i].power_off_latency_ns)) { + if ((idle_duration_ns >= (genpd->states[i].residency_ns + + genpd->states[i].power_off_latency_ns)) && + (global_constraint >= (genpd->states[i].power_on_latency_ns + + genpd->states[i].power_off_latency_ns))) { genpd->state_idx = i; genpd->gd->last_enter = now; genpd->gd->reflect_residency = true;

5 months

2
1
0 0

FAILED: patch "[PATCH] pmdomain: governor: Consider CPU latency tolerance from" failed to apply to 5.10-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.10-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y git checkout FETCH_HEAD git cherry-pick -x 500ba33284416255b9a5b50ace24470b6fe77ea5 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025072113-grill-thimble-5d5c@gregkh' --subject-prefix 'PATCH 5.10.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 500ba33284416255b9a5b50ace24470b6fe77ea5 Mon Sep 17 00:00:00 2001 From: Maulik Shah <maulik.shah(a)oss.qualcomm.com> Date: Wed, 9 Jul 2025 14:00:11 +0530 Subject: [PATCH] pmdomain: governor: Consider CPU latency tolerance from pm_domain_cpu_gov pm_domain_cpu_gov is selecting a cluster idle state but does not consider latency tolerance of child CPUs. This results in deeper cluster idle state whose latency does not meet latency tolerance requirement. Select deeper idle state only if global and device latency tolerance of all child CPUs meet. Test results on SM8750 with 300 usec PM-QoS on CPU0 which is less than domain idle state entry (2150) + exit (1983) usec latency mentioned in devicetree, demonstrate the issue. # echo 300 > /sys/devices/system/cpu/cpu0/power/pm_qos_resume_latency_us Before: (Usage is incrementing) ====== # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 29817 537 8 270 0 # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 30348 542 8 271 0 After: (Usage is not incrementing due to latency tolerance) ====== # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 39319 626 14 307 0 # cat /sys/kernel/debug/pm_genpd/power-domain-cluster0/idle_states State Time Spent(ms) Usage Rejected Above Below S0 39319 626 14 307 0 Signed-off-by: Maulik Shah <maulik.shah(a)oss.qualcomm.com> Fixes: e94999688e3a ("PM / Domains: Add genpd governor for CPUs") Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/r/20250709-pmdomain_qos-v2-1-976b12257899@oss.qualc… Signed-off-by: Ulf Hansson <ulf.hansson(a)linaro.org> diff --git a/drivers/pmdomain/governor.c b/drivers/pmdomain/governor.c index c1e148657c87..39359811a930 100644 --- a/drivers/pmdomain/governor.c +++ b/drivers/pmdomain/governor.c @@ -8,6 +8,7 @@ #include <linux/pm_domain.h> #include <linux/pm_qos.h> #include <linux/hrtimer.h> +#include <linux/cpu.h> #include <linux/cpuidle.h> #include <linux/cpumask.h> #include <linux/ktime.h> @@ -349,6 +350,8 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) struct cpuidle_device *dev; ktime_t domain_wakeup, next_hrtimer; ktime_t now = ktime_get(); + struct device *cpu_dev; + s64 cpu_constraint, global_constraint; s64 idle_duration_ns; int cpu, i; @@ -359,6 +362,7 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) if (!(genpd->flags & GENPD_FLAG_CPU_DOMAIN)) return true; + global_constraint = cpu_latency_qos_limit(); /* * Find the next wakeup for any of the online CPUs within the PM domain * and its subdomains. Note, we only need the genpd->cpus, as it already @@ -372,8 +376,16 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) if (ktime_before(next_hrtimer, domain_wakeup)) domain_wakeup = next_hrtimer; } + + cpu_dev = get_cpu_device(cpu); + if (cpu_dev) { + cpu_constraint = dev_pm_qos_raw_resume_latency(cpu_dev); + if (cpu_constraint < global_constraint) + global_constraint = cpu_constraint; + } } + global_constraint *= NSEC_PER_USEC; /* The minimum idle duration is from now - until the next wakeup. */ idle_duration_ns = ktime_to_ns(ktime_sub(domain_wakeup, now)); if (idle_duration_ns <= 0) @@ -389,8 +401,10 @@ static bool cpu_power_down_ok(struct dev_pm_domain *pd) */ i = genpd->state_idx; do { - if (idle_duration_ns >= (genpd->states[i].residency_ns + - genpd->states[i].power_off_latency_ns)) { + if ((idle_duration_ns >= (genpd->states[i].residency_ns + + genpd->states[i].power_off_latency_ns)) && + (global_constraint >= (genpd->states[i].power_on_latency_ns + + genpd->states[i].power_off_latency_ns))) { genpd->state_idx = i; genpd->gd->last_enter = now; genpd->gd->reflect_residency = true;

5 months

2
1
0 0

[PATCH 6.1] crypto: qat - fix ring to service map for QAT GEN4

by Giovanni Cabiddu

[ Upstream commit a238487f7965d102794ed9f8aff0b667cd2ae886 ] The 4xxx drivers hardcode the ring to service mapping. However, when additional configurations where added to the driver, the mappings were not updated. This implies that an incorrect mapping might be reported through pfvf for certain configurations. This is a backport of the upstream commit with modifications, as the original patch does not apply cleanly to kernel v6.1.x. The logic has been simplified to reflect the limited configurations of the QAT driver in this version: crypto-only and compression. Instead of dynamically computing the ring to service mappings, these are now hardcoded to simplify the backport. Fixes: 0cec19c761e5 ("crypto: qat - add support for compression for 4xxx") Signed-off-by: Giovanni Cabiddu <giovanni.cabiddu(a)intel.com> Reviewed-by: Damian Muszynski <damian.muszynski(a)intel.com> Reviewed-by: Tero Kristo <tero.kristo(a)linux.intel.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Cc: <stable(a)vger.kernel.org> # 6.1.x Reviewed-by: Ahsan Atta <ahsan.atta(a)intel.com> Tested-by: Ahsan Atta <ahsan.atta(a)intel.com> --- drivers/crypto/qat/qat_4xxx/adf_4xxx_hw_data.c | 13 +++++++++++++ drivers/crypto/qat/qat_common/adf_accel_devices.h | 1 + drivers/crypto/qat/qat_common/adf_gen4_hw_data.h | 6 ++++++ drivers/crypto/qat/qat_common/adf_init.c | 3 +++ 4 files changed, 23 insertions(+) diff --git a/drivers/crypto/qat/qat_4xxx/adf_4xxx_hw_data.c b/drivers/crypto/qat/qat_4xxx/adf_4xxx_hw_data.c index fda5f699ff57..65b52c692add 100644 --- a/drivers/crypto/qat/qat_4xxx/adf_4xxx_hw_data.c +++ b/drivers/crypto/qat/qat_4xxx/adf_4xxx_hw_data.c @@ -297,6 +297,18 @@ static char *uof_get_name(struct adf_accel_dev *accel_dev, u32 obj_num) return NULL; } +static u16 get_ring_to_svc_map(struct adf_accel_dev *accel_dev) +{ + switch (get_service_enabled(accel_dev)) { + case SVC_CY: + return ADF_GEN4_DEFAULT_RING_TO_SRV_MAP; + case SVC_DC: + return ADF_GEN4_DEFAULT_RING_TO_SRV_MAP_DC; + } + + return 0; +} + static u32 uof_get_ae_mask(struct adf_accel_dev *accel_dev, u32 obj_num) { switch (get_service_enabled(accel_dev)) { @@ -353,6 +365,7 @@ void adf_init_hw_data_4xxx(struct adf_hw_device_data *hw_data) hw_data->uof_get_ae_mask = uof_get_ae_mask; hw_data->set_msix_rttable = set_msix_default_rttable; hw_data->set_ssm_wdtimer = adf_gen4_set_ssm_wdtimer; + hw_data->get_ring_to_svc_map = get_ring_to_svc_map; hw_data->disable_iov = adf_disable_sriov; hw_data->ring_pair_reset = adf_gen4_ring_pair_reset; hw_data->enable_pm = adf_gen4_enable_pm; diff --git a/drivers/crypto/qat/qat_common/adf_accel_devices.h b/drivers/crypto/qat/qat_common/adf_accel_devices.h index ad01d99e6e2b..7993d0f82dea 100644 --- a/drivers/crypto/qat/qat_common/adf_accel_devices.h +++ b/drivers/crypto/qat/qat_common/adf_accel_devices.h @@ -176,6 +176,7 @@ struct adf_hw_device_data { void (*get_arb_info)(struct arb_info *arb_csrs_info); void (*get_admin_info)(struct admin_info *admin_csrs_info); enum dev_sku_info (*get_sku)(struct adf_hw_device_data *self); + u16 (*get_ring_to_svc_map)(struct adf_accel_dev *accel_dev); int (*alloc_irq)(struct adf_accel_dev *accel_dev); void (*free_irq)(struct adf_accel_dev *accel_dev); void (*enable_error_correction)(struct adf_accel_dev *accel_dev); diff --git a/drivers/crypto/qat/qat_common/adf_gen4_hw_data.h b/drivers/crypto/qat/qat_common/adf_gen4_hw_data.h index 4fb4b3df5a18..5e653ec755e6 100644 --- a/drivers/crypto/qat/qat_common/adf_gen4_hw_data.h +++ b/drivers/crypto/qat/qat_common/adf_gen4_hw_data.h @@ -95,6 +95,12 @@ do { \ ADF_RING_BUNDLE_SIZE * (bank) + \ ADF_RING_CSR_RING_SRV_ARB_EN, (value)) +#define ADF_GEN4_DEFAULT_RING_TO_SRV_MAP_DC \ + (COMP << ADF_CFG_SERV_RING_PAIR_0_SHIFT | \ + COMP << ADF_CFG_SERV_RING_PAIR_1_SHIFT | \ + COMP << ADF_CFG_SERV_RING_PAIR_2_SHIFT | \ + COMP << ADF_CFG_SERV_RING_PAIR_3_SHIFT) + /* Default ring mapping */ #define ADF_GEN4_DEFAULT_RING_TO_SRV_MAP \ (ASYM << ADF_CFG_SERV_RING_PAIR_0_SHIFT | \ diff --git a/drivers/crypto/qat/qat_common/adf_init.c b/drivers/crypto/qat/qat_common/adf_init.c index 2e3481270c4b..49f07584f8c9 100644 --- a/drivers/crypto/qat/qat_common/adf_init.c +++ b/drivers/crypto/qat/qat_common/adf_init.c @@ -95,6 +95,9 @@ int adf_dev_init(struct adf_accel_dev *accel_dev) return -EFAULT; } + if (hw_data->get_ring_to_svc_map) + hw_data->ring_to_svc_map = hw_data->get_ring_to_svc_map(accel_dev); + if (adf_ae_init(accel_dev)) { dev_err(&GET_DEV(accel_dev), "Failed to initialise Acceleration Engine\n"); -- 2.50.0

5 months

3
6
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror July 2025