From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Qian Cai <cai@lca.pw>, Borislav Petkov <bp@suse.de>,
"Rafael J . Wysocki" <rafael.j.wysocki@intel.com>,
Sasha Levin <sashal@kernel.org>,
linux-pm@vger.kernel.org, linux-acpi@vger.kernel.org,
devel@acpica.org
Subject: [PATCH AUTOSEL 5.4 58/84] x86: ACPI: fix CPU hotplug deadlock
Date: Wed, 15 Apr 2020 07:44:15 -0400 [thread overview]
Message-ID: <20200415114442.14166-58-sashal@kernel.org> (raw)
In-Reply-To: <20200415114442.14166-1-sashal@kernel.org>
From: Qian Cai <cai@lca.pw>
[ Upstream commit 696ac2e3bf267f5a2b2ed7d34e64131f2287d0ad ]
Similar to commit 0266d81e9bf5 ("acpi/processor: Prevent cpu hotplug
deadlock") except this is for acpi_processor_ffh_cstate_probe():
"The problem is that the work is scheduled on the current CPU from the
hotplug thread associated with that CPU.
It's not required to invoke these functions via the workqueue because
the hotplug thread runs on the target CPU already.
Check whether current is a per cpu thread pinned on the target CPU and
invoke the function directly to avoid the workqueue."
WARNING: possible circular locking dependency detected
------------------------------------------------------
cpuhp/1/15 is trying to acquire lock:
ffffc90003447a28 ((work_completion)(&wfc.work)){+.+.}-{0:0}, at: __flush_work+0x4c6/0x630
but task is already holding lock:
ffffffffafa1c0e8 (cpuidle_lock){+.+.}-{3:3}, at: cpuidle_pause_and_lock+0x17/0x20
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (cpu_hotplug_lock){++++}-{0:0}:
cpus_read_lock+0x3e/0xc0
irq_calc_affinity_vectors+0x5f/0x91
__pci_enable_msix_range+0x10f/0x9a0
pci_alloc_irq_vectors_affinity+0x13e/0x1f0
pci_alloc_irq_vectors_affinity at drivers/pci/msi.c:1208
pqi_ctrl_init+0x72f/0x1618 [smartpqi]
pqi_pci_probe.cold.63+0x882/0x892 [smartpqi]
local_pci_probe+0x7a/0xc0
work_for_cpu_fn+0x2e/0x50
process_one_work+0x57e/0xb90
worker_thread+0x363/0x5b0
kthread+0x1f4/0x220
ret_from_fork+0x27/0x50
-> #0 ((work_completion)(&wfc.work)){+.+.}-{0:0}:
__lock_acquire+0x2244/0x32a0
lock_acquire+0x1a2/0x680
__flush_work+0x4e6/0x630
work_on_cpu+0x114/0x160
acpi_processor_ffh_cstate_probe+0x129/0x250
acpi_processor_evaluate_cst+0x4c8/0x580
acpi_processor_get_power_info+0x86/0x740
acpi_processor_hotplug+0xc3/0x140
acpi_soft_cpu_online+0x102/0x1d0
cpuhp_invoke_callback+0x197/0x1120
cpuhp_thread_fun+0x252/0x2f0
smpboot_thread_fn+0x255/0x440
kthread+0x1f4/0x220
ret_from_fork+0x27/0x50
other info that might help us debug this:
Chain exists of:
(work_completion)(&wfc.work) --> cpuhp_state-up --> cpuidle_lock
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(cpuidle_lock);
lock(cpuhp_state-up);
lock(cpuidle_lock);
lock((work_completion)(&wfc.work));
*** DEADLOCK ***
3 locks held by cpuhp/1/15:
#0: ffffffffaf51ab10 (cpu_hotplug_lock){++++}-{0:0}, at: cpuhp_thread_fun+0x69/0x2f0
#1: ffffffffaf51ad40 (cpuhp_state-up){+.+.}-{0:0}, at: cpuhp_thread_fun+0x69/0x2f0
#2: ffffffffafa1c0e8 (cpuidle_lock){+.+.}-{3:3}, at: cpuidle_pause_and_lock+0x17/0x20
Call Trace:
dump_stack+0xa0/0xea
print_circular_bug.cold.52+0x147/0x14c
check_noncircular+0x295/0x2d0
__lock_acquire+0x2244/0x32a0
lock_acquire+0x1a2/0x680
__flush_work+0x4e6/0x630
work_on_cpu+0x114/0x160
acpi_processor_ffh_cstate_probe+0x129/0x250
acpi_processor_evaluate_cst+0x4c8/0x580
acpi_processor_get_power_info+0x86/0x740
acpi_processor_hotplug+0xc3/0x140
acpi_soft_cpu_online+0x102/0x1d0
cpuhp_invoke_callback+0x197/0x1120
cpuhp_thread_fun+0x252/0x2f0
smpboot_thread_fn+0x255/0x440
kthread+0x1f4/0x220
ret_from_fork+0x27/0x50
Signed-off-by: Qian Cai <cai@lca.pw>
Tested-by: Borislav Petkov <bp@suse.de>
[ rjw: Subject ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
arch/x86/kernel/acpi/cstate.c | 3 ++-
drivers/acpi/processor_throttling.c | 7 -------
include/acpi/processor.h | 8 ++++++++
3 files changed, 10 insertions(+), 8 deletions(-)
diff --git a/arch/x86/kernel/acpi/cstate.c b/arch/x86/kernel/acpi/cstate.c
index caf2edccbad2e..49ae4e1ac9cd8 100644
--- a/arch/x86/kernel/acpi/cstate.c
+++ b/arch/x86/kernel/acpi/cstate.c
@@ -161,7 +161,8 @@ int acpi_processor_ffh_cstate_probe(unsigned int cpu,
/* Make sure we are running on right CPU */
- retval = work_on_cpu(cpu, acpi_processor_ffh_cstate_probe_cpu, cx);
+ retval = call_on_cpu(cpu, acpi_processor_ffh_cstate_probe_cpu, cx,
+ false);
if (retval == 0) {
/* Use the hint in CST */
percpu_entry->states[cx->index].eax = cx->address;
diff --git a/drivers/acpi/processor_throttling.c b/drivers/acpi/processor_throttling.c
index 532a1ae3595a7..a0bd56ece3ff5 100644
--- a/drivers/acpi/processor_throttling.c
+++ b/drivers/acpi/processor_throttling.c
@@ -897,13 +897,6 @@ static long __acpi_processor_get_throttling(void *data)
return pr->throttling.acpi_processor_get_throttling(pr);
}
-static int call_on_cpu(int cpu, long (*fn)(void *), void *arg, bool direct)
-{
- if (direct || (is_percpu_thread() && cpu == smp_processor_id()))
- return fn(arg);
- return work_on_cpu(cpu, fn, arg);
-}
-
static int acpi_processor_get_throttling(struct acpi_processor *pr)
{
if (!pr)
diff --git a/include/acpi/processor.h b/include/acpi/processor.h
index 47805172e73d8..683e124ad517d 100644
--- a/include/acpi/processor.h
+++ b/include/acpi/processor.h
@@ -297,6 +297,14 @@ static inline void acpi_processor_ffh_cstate_enter(struct acpi_processor_cx
}
#endif
+static inline int call_on_cpu(int cpu, long (*fn)(void *), void *arg,
+ bool direct)
+{
+ if (direct || (is_percpu_thread() && cpu == smp_processor_id()))
+ return fn(arg);
+ return work_on_cpu(cpu, fn, arg);
+}
+
/* in processor_perflib.c */
#ifdef CONFIG_CPU_FREQ
--
2.20.1
next prev parent reply other threads:[~2020-04-15 12:20 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20200415114442.14166-1-sashal@kernel.org>
2020-04-15 11:43 ` [PATCH AUTOSEL 5.4 09/84] power: supply: bq27xxx_battery: Silence deferred-probe error Sasha Levin
2020-04-15 11:44 ` [PATCH AUTOSEL 5.4 55/84] drm/nouveau: workaround runpm fail by disabling PCI power management on certain intel bridges Sasha Levin
2020-04-15 11:44 ` Sasha Levin [this message]
2020-04-15 11:44 ` [PATCH AUTOSEL 5.4 79/84] power: supply: axp288_fuel_gauge: Broaden vendor check for Intel Compute Sticks Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200415114442.14166-58-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=bp@suse.de \
--cc=cai@lca.pw \
--cc=devel@acpica.org \
--cc=linux-acpi@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=rafael.j.wysocki@intel.com \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).