The removed code was added to handle the case where the power domain is
already on during the driver's probing. In this use case, the "is_off"
parameter is passed as false to pm_genpd_init() to inform it not to call
the power_on() callback, as it's unnecessary to perform the hardware
power-on procedure since the power domain is already on. Therefore, with
the call to clk_bulk_prepare_enable() by probe(), the system is in the
same operational state as when "is_off" is passed as true after the
power_on() callback execution:
probe() -> is_off == true -> clk_bulk_prepare_enable() called by power_on()
probe() -> is_off == false -> clk_bulk_prepare_enable() called by probe()
Reaching the same logical and operational state, it follows that upon
driver removal, there is no need to perform different actions depending
on the power domain's on/off state during probing.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Link: https://lore.kernel.org/r/20240825143428.556439-2-dario.binacchi@amarulasolutions.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
for_each_available_child_of_node() can help to iterate through the
device_node, and we don't need to use while loop. Besides, the purpose
of the while loop is to find a device_node which fits the condition
"child_req_np == ref_np", we can just read the property of "child_np"
directly in for_each_available_child_of_node(). No functional change
with such conversion.
Signed-off-by: Zhang Zekun <zhangzekun11@huawei.com>
Link: https://lore.kernel.org/r/20240821034022.27394-2-zhangzekun11@huawei.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Merge the pmdomain fixes for v6.11-rc[n] into the next branch, to allow them
to get tested together with the new changes that are targeted for v6.12.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
It has turned out that having _set_required_opps() to recursively call
dev_pm_opp_set_opp() to set the required OPPs, doesn't really work as well
as we expected.
More precisely, at each recursive call to dev_pm_opp_set_opp() we are
changing an OPP for a required_dev that belongs to a required-OPP table.
The problem with this, is that we may have several devices sharing the same
required-OPP table, which leads to an incorrect behaviour in regards to
aggregating the per device votes.
To fix the problem for a required-OPP table belonging to a PM domain, which
is the only existing usecase for now, let's simply replace the call to
dev_pm_opp_set_opp() in _set_required_opps() by a call to _set_opp_level().
Moving forward we may potentially need to add support for other types of
required-OPP tables. In this case, the aggregation needs to be thought of.
Fixes: e37440e7e2 ("OPP: Call dev_pm_opp_set_opp() for required OPPs")
Cc: stable@vger.kernel.org
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://lore.kernel.org/r/20240822224547.385095-2-ulf.hansson@linaro.org
Merge the immutable branch dt into next, to allow the DT bindings to be
tested together with changes that are targeted for v6.12.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Merge the pmdomain fixes for v6.11-rc[n] into the next branch, to allow them
to get tested together with the new changes that are targeted for v6.12.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
With "quiet" set in bootargs, there is power domain failure:
"imx93_power_domain 44462400.power-domain: pd_off timeout: name:
44462400.power-domain, stat: 4"
The current power on opertation takes ISO state as power on finished
flag, but it is wrong. Before powering on operation really finishes,
powering off comes and powering off will never finish because the last
powering on still not finishes, so the following powering off actually
not trigger hardware state machine to run. SSAR is the last step when
powering on a domain, so need to wait SSAR done when powering on.
Since EdgeLock Enclave(ELE) handshake is involved in the flow, enlarge
the waiting time to 10ms for both on and off to avoid timeout.
Cc: stable@vger.kernel.org
Fixes: 0a0f7cc25d ("soc: imx: add i.MX93 SRC power domain driver")
Reviewed-by: Jacky Bai <ping.bai@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Link: https://lore.kernel.org/r/20240814124740.2778952-1-peng.fan@oss.nxp.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
The Raspberry Pi power driver heavily relies on the logging of the
underlying firmware driver. This results in disadvantages like unspecific
error messages or limited debugging options. So implement the logging for
the most important function rpi_firmware_set_power.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Link: https://lore.kernel.org/r/20240728114200.75559-5-wahrenst@gmx.net
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Merge the pmdomain fixes for v6.11-rc[n] into the next branch, to allow them
to get tested together with the new changes that are targeted for v6.12.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
To enable the domain-idle-states to be used during s2idle on a PREEMPT_RT
based configuration, let's allow the re-assignment of the ->enter_s2idle()
callback to psci_enter_s2idle_domain_idle_state().
Similar to s2ram, let's leave the support for CPU hotplug outside
PREEMPT_RT, as it's depending on using runtime PM. For s2idle, this means
that an offline CPU's PM domain will remain powered-on. In practise this
may lead to that a shallower idle-state than necessary gets selected, which
shouldn't be an issue (besides wasting power).
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-8-ulf.hansson@linaro.org
The hierarchical PM domain topology are currently disabled on a PREEMPT_RT
based configuration. As a first step to enable it to be used, let's try to
attach the CPU devices to their PM domains on PREEMPT_RT. In this way the
syscore ops becomes available, allowing the PM domain topology to be
managed during s2ram.
For the moment let's leave the support for CPU hotplug outside PREEMPT_RT,
as it's depending on using runtime PM. For s2ram, this isn't a problem as
all CPUs are managed via the syscore ops.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-7-ulf.hansson@linaro.org
When using the hierarchical topology and PSCI OSI-mode we may end up
overriding the deepest idle-state's ->enter|enter_s2idle() callbacks, but
there is no point to also re-assign the CPUIDLE_FLAG_RCU_IDLE for the
idle-state in question, as that has already been set when parsing the
states from DT. See init_state_node().
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-6-ulf.hansson@linaro.org
The domain-idle-states are currently disabled on a PREEMPT_RT based
configuration for the cpuidle-psci-domain. To enable them to be used for
system-wide suspend and in particular during s2idle, let's set the
GENPD_FLAG_RPM_ALWAYS_ON instead of GENPD_FLAG_ALWAYS_ON for the
corresponding genpd provider.
In this way, the runtime PM path remains disabled in genpd for its attached
devices, while powering-on/off the PM domain during system-wide suspend
becomes allowed.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-5-ulf.hansson@linaro.org
Using kobject_get_path() means a dynamic memory allocation gets done, which
doesn't work on a PREEMPT_RT based configuration while holding genpd's raw
spinlock.
To fix the problem, let's convert into using the simpler dev_name(). This
means the information about the path doesn't get presented in debugfs, but
hopefully this shouldn't be an issue.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-4-ulf.hansson@linaro.org
There is no need to hold the genpd-lock, while assigning the
dev->pm_domain. In fact, it becomes a problem on a PREEMPT_RT based
configuration as the genpd-lock may be a raw spinlock, while the lock
acquired through the call to dev_pm_domain_set() is a regular spinlock.
To fix the problem, let's simply move the calls to dev_pm_domain_set()
outside the genpd-lock.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-3-ulf.hansson@linaro.org
To allow a genpd provider for a CPU PM domain to enter a domain-idle-state
during s2idle on a PREEMPT_RT based configuration, we can't use the regular
spinlock, as they are turned into sleepable locks on PREEMPT_RT.
To address this problem, let's convert into using the raw spinlock, but
only for genpd providers that have the GENPD_FLAG_CPU_DOMAIN bit set. In
this way, the lock can still be acquired/released in atomic context, which
is needed in the idle-path for PREEMPT_RT.
Do note that the genpd power-on/off notifiers may also be fired during
s2idle, but these are already prepared for PREEMPT_RT as they are based on
the raw notifiers. However, consumers of them may need to adopt accordingly
to work properly on PREEMPT_RT.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Raghavendra Kakarla <quic_rkakarla@quicinc.com> # qcm6490 with PREEMPT_RT set
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240527142557.321610-2-ulf.hansson@linaro.org
The kernel sleep profile is no longer working due to a recursive locking
bug introduced by commit 42a20f86dc ("sched: Add wrapper for get_wchan()
to keep task blocked")
Booting with the 'profile=sleep' kernel command line option added or
executing
# echo -n sleep > /sys/kernel/profiling
after boot causes the system to lock up.
Lockdep reports
kthreadd/3 is trying to acquire lock:
ffff93ac82e08d58 (&p->pi_lock){....}-{2:2}, at: get_wchan+0x32/0x70
but task is already holding lock:
ffff93ac82e08d58 (&p->pi_lock){....}-{2:2}, at: try_to_wake_up+0x53/0x370
with the call trace being
lock_acquire+0xc8/0x2f0
get_wchan+0x32/0x70
__update_stats_enqueue_sleeper+0x151/0x430
enqueue_entity+0x4b0/0x520
enqueue_task_fair+0x92/0x6b0
ttwu_do_activate+0x73/0x140
try_to_wake_up+0x213/0x370
swake_up_locked+0x20/0x50
complete+0x2f/0x40
kthread+0xfb/0x180
However, since nobody noticed this regression for more than two years,
let's remove 'profile=sleep' support based on the assumption that nobody
needs this functionality.
Fixes: 42a20f86dc ("sched: Add wrapper for get_wchan() to keep task blocked")
Cc: stable@vger.kernel.org # v5.16+
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull x86 fixes from Thomas Gleixner:
- Prevent a deadlock on cpu_hotplug_lock in the aperf/mperf driver.
A recent change in the ACPI code which consolidated code pathes moved
the invocation of init_freq_invariance_cppc() to be moved to a CPU
hotplug handler. The first invocation on AMD CPUs ends up enabling a
static branch which dead locks because the static branch enable tries
to acquire cpu_hotplug_lock but that lock is already held write by
the hotplug machinery.
Use static_branch_enable_cpuslocked() instead and take the hotplug
lock read for the Intel code path which is invoked from the
architecture code outside of the CPU hotplug operations.
- Fix the number of reserved bits in the sev_config structure bit field
so that the bitfield does not exceed 64 bit.
- Add missing Zen5 model numbers
- Fix the alignment assumptions of pti_clone_pgtable() and
clone_entry_text() on 32-bit:
The code assumes PMD aligned code sections, but on 32-bit the kernel
entry text is not PMD aligned. So depending on the code size and
location, which is configuration and compiler dependent, entry text
can cross a PMD boundary. As the start is not PMD aligned adding PMD
size to the start address is larger than the end address which
results in partially mapped entry code for user space. That causes
endless recursion on the first entry from userspace (usually #PF).
Cure this by aligning the start address in the addition so it ends up
at the next PMD start address.
clone_entry_text() enforces PMD mapping, but on 32-bit the tail might
eventually be PTE mapped, which causes a map fail because the PMD for
the tail is not a large page mapping. Use PTI_LEVEL_KERNEL_IMAGE for
the clone() invocation which resolves to PTE on 32-bit and PMD on
64-bit.
- Zero the 8-byte case for get_user() on range check failure on 32-bit
The recend consolidation of the 8-byte get_user() case broke the
zeroing in the failure case again. Establish it by clearing ECX
before the range check and not afterwards as that obvioulsy can't be
reached when the range check fails
* tag 'x86-urgent-2024-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/uaccess: Zero the 8-byte get_range case on failure on 32-bit
x86/mm: Fix pti_clone_entry_text() for i386
x86/mm: Fix pti_clone_pgtable() alignment assumption
x86/setup: Parse the builtin command line before merging
x86/CPU/AMD: Add models 0x60-0x6f to the Zen5 range
x86/sev: Fix __reserved field in sev_config
x86/aperfmperf: Fix deadlock on cpu_hotplug_lock
Pull timer fixes from Thomas Gleixner:
"Two fixes for the timer/clocksource code:
- The recent fix to make the take over of the broadcast timer more
reliable retrieves a per CPU pointer in preemptible context.
This went unnoticed in testing as some compilers hoist the access
into the non-preemotible section where the pointer is actually
used, but obviously compilers can rightfully invoke it where the
code put it.
Move it into the non-preemptible section right to the actual usage
side to cure it.
- The clocksource watchdog is supposed to emit a warning when the
retry count is greater than one and the number of retries reaches
the limit.
The condition is backwards and warns always when the count is
greater than one. Fixup the condition to prevent spamming dmesg"
* tag 'timers-urgent-2024-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
clocksource: Fix brown-bag boolean thinko in cs_watchdog_read()
tick/broadcast: Move per CPU pointer access into the atomic section
Pull scheduler fixes from Thomas Gleixner:
- When stime is larger than rtime due to accounting imprecision, then
utime = rtime - stime becomes negative. As this is unsigned math, the
result becomes a huge positive number.
Cure it by resetting stime to rtime in that case, so utime becomes 0.
- Restore consistent state when sched_cpu_deactivate() fails.
When offlining a CPU fails in sched_cpu_deactivate() after the SMT
present counter has been decremented, then the function aborts but
fails to increment the SMT present counter and leaves it imbalanced.
Consecutive operations cause it to underflow. Add the missing fixup
for the error path.
For SMT accounting the runqueue needs to marked online again in the
error exit path to restore consistent state.
* tag 'sched-urgent-2024-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/core: Fix unbalance set_rq_online/offline() in sched_cpu_deactivate()
sched/core: Introduce sched_set_rq_on/offline() helper
sched/smt: Fix unbalance sched_smt_present dec/inc
sched/smt: Introduce sched_smt_present_inc/dec() helper
sched/cputime: Fix mul_u64_u64_div_u64() precision for cputime
Pull x86 perf fixes from Thomas Gleixner:
- Move the smp_processor_id() invocation back into the non-preemtible
region, so that the result is valid to use
- Add the missing package C2 residency counters for Sierra Forest CPUs
to make the newly added support actually useful
* tag 'perf-urgent-2024-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86: Fix smp_processor_id()-in-preemptible warnings
perf/x86/intel/cstate: Add pkg C2 residency counter for Sierra Forest
Pull irq fixes from Thomas Gleixner:
"A couple of fixes for interrupt chip drivers:
- Make sure to skip the clear register space in the MBIGEN driver
when calculating the node register index. Otherwise the clear
register is clobbered and the wrong node registers are accessed.
- Fix a signed/unsigned confusion in the loongarch CPU driver which
converts an error code to a huge "valid" interrupt number.
- Convert the mesion GPIO interrupt controller lock to a raw spinlock
so it works on RT.
- Add a missing static to a internal function in the pic32 EVIC
driver"
* tag 'irq-urgent-2024-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip/mbigen: Fix mbigen node address layout
irqchip/meson-gpio: Convert meson_gpio_irq_controller::lock to 'raw_spinlock_t'
irqchip/irq-pic32-evic: Add missing 'static' to internal function
irqchip/loongarch-cpu: Fix return value of lpic_gsi_to_irq()
