mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Will Deacon <will@kernel.org>
To: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org, "Will Deacon" <will@kernel.org>,
	"Thomas Gleixner" <tglx@kernel.org>,
	"Ben Horgan" <ben.horgan@arm.com>,
	"Ștefania Ion" <stefion@google.com>,
	"Catalin Marinas" <catalin.marinas@arm.com>,
	"Pankaj Patil" <pankaj.patil@oss.qualcomm.com>,
	"Borislav Petkov" <bp@alien8.de>,
	"Lorenzo Pieralisi" <lpieralisi@kernel.org>,
	"Jinjie Ruan" <ruanjinjie@huawei.com>,
	"Mark Rutland" <mark.rutland@arm.com>,
	"Tarun Sahu" <tarunsahu@google.com>,
	"Fuad Tabba" <fuad.tabba@linux.dev>,
	"David Woodhouse" <dwmw@amazon.co.uk>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Marc Zyngier" <maz@kernel.org>
Subject: [PATCH v2 13/23] arm64: smp: Defer RCU registration during secondary CPU bringup
Date: Fri,  9 Oct 2026 11:07:24 +0100	[thread overview]
Message-ID: <20261009100738.31288-14-will@kernel.org> (raw)
In-Reply-To: <20261009100738.31288-1-will@kernel.org>

Calling rcutree_report_cpu_starting() early during boot can lead to
livelocks with the generic CPU hotplug mechanism if the boot CPU blocks
on an RCU grace period while the CPU being onlined is spinning in
cpuhp_ap_sync_alive().

In preparation for enabling the generic CPU hotplug code on arm64, split
up the trace_hardirqs_off() call during secondary CPU bringup so that we
update lockdep early but defer the tracing updates until after the new
CPU has been registered with RCU, which itself is deferred until after
the early CPU compatability checks, allowing us to drop the explicit
call to rcutree_report_cpu_dead() entirely.

Signed-off-by: Will Deacon <will@kernel.org>
---
 arch/arm64/kernel/smp.c | 6 +++---
 include/linux/rcutree.h | 2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c
index e1b15cab9ed6..014b163ce658 100644
--- a/arch/arm64/kernel/smp.c
+++ b/arch/arm64/kernel/smp.c
@@ -217,8 +217,7 @@ asmlinkage notrace void secondary_start_kernel(void)
 	if (system_uses_irq_prio_masking())
 		init_gic_priority_masking();
 
-	rcutree_report_cpu_starting(cpu);
-	trace_hardirqs_off();
+	lockdep_hardirqs_off(CALLER_ADDR0);
 
 	/*
 	 * If the system has established the capabilities, make sure
@@ -235,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void)
 	 * Log the CPU info before it is marked online and might get read.
 	 */
 	cpuinfo_store_cpu();
+	rcutree_report_cpu_starting(cpu);
 	update_cpu_features(cpu);
 	store_cpu_topology(cpu);
 
@@ -242,6 +242,7 @@ asmlinkage notrace void secondary_start_kernel(void)
 	 * Enable GIC and timers.
 	 */
 	notify_cpu_starting(cpu);
+	trace_hardirqs_off_finish();
 
 	ipi_setup(cpu);
 
@@ -411,7 +412,6 @@ void __noreturn cpu_die_early(void)
 
 	/* Mark this CPU absent */
 	set_cpu_present(cpu, 0);
-	rcutree_report_cpu_dead();
 
 	if (IS_ENABLED(CONFIG_HOTPLUG_CPU)) {
 		update_cpu_boot_status(CPU_KILL_ME);
diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h
index 16a04202888b..e6ad4f2a475c 100644
--- a/include/linux/rcutree.h
+++ b/include/linux/rcutree.h
@@ -116,7 +116,7 @@ int rcutree_offline_cpu(unsigned int cpu);
 
 void rcutree_migrate_callbacks(int cpu);
 
-/* Called from hotplug and also arm64 early secondary boot failure */
+/* Called from hotplug */
 void rcutree_report_cpu_dead(void);
 
 #endif /* __LINUX_RCUTREE_H */
-- 
2.56.0.385.gd3acb90ef8-goog


  parent reply	other threads:[~2026-10-09 10:08 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-09 10:07 [PATCH v2 00/23] arm64: Implement parallel CPU onlining with PSCI v0.2+ Will Deacon
2026-10-09 10:07 ` [PATCH v2 01/23] cpu/hotplug: Clean up cmpxchg() logic in cpuhp_can_boot_ap() Will Deacon
2026-10-09 10:07 ` [PATCH v2 02/23] cpu/hotplug: Avoid trying to bring up CPUs that are already online Will Deacon
2026-10-09 10:07 ` [PATCH v2 03/23] cpu/hotplug: Avoid busy-polling on archs where cpu_relax() is a no-op Will Deacon
2026-10-09 10:46   ` David Woodhouse
2026-10-09 11:07     ` Will Deacon
2026-10-09 10:07 ` [PATCH v2 04/23] cpu/hotplug: Propagate bring-up status to arch_cpuhp_cleanup_kick_cpu() Will Deacon
2026-10-09 10:07 ` [PATCH v2 05/23] arm64: cpufeature: Check arm64_ftr_regs[] before the first store Will Deacon
2026-10-09 10:07 ` [PATCH v2 06/23] arm64: cpufeature: Add read_cpuid_with_overrides() Will Deacon
2026-10-09 10:07 ` [PATCH v2 07/23] arm64: cpufeature: Read MPAMIDR_EL1 in __cpuinfo_store_cpu() Will Deacon
2026-10-09 15:44   ` Ben Horgan
2026-10-09 10:07 ` [PATCH v2 08/23] arm64: cpufeature: Store every ID register with its overrides applied Will Deacon
2026-10-09 15:51   ` Ben Horgan
2026-10-09 10:07 ` [PATCH v2 09/23] arm64: smp: Tidy up smp_prepare_cpus() Will Deacon
2026-10-09 10:07 ` [PATCH v2 10/23] arm64: smp: Tidy up cpuinfo init and cpufeature updates Will Deacon
2026-10-09 10:07 ` [PATCH v2 11/23] arm64: smp: Defer update of secondary CPU capabilities Will Deacon
2026-10-09 10:07 ` [PATCH v2 12/23] arm64: smp: Don't bother printing the I-cache policy for each CPU Will Deacon
2026-10-09 10:07 ` Will Deacon [this message]
2026-10-09 10:07 ` [PATCH v2 14/23] arm64: smp: Use generic HOTPLUG_CORE_SYNC_FULL machinery for CPU onlining Will Deacon
2026-10-09 10:07 ` [PATCH v2 15/23] arm64: smp: Use generic HOTPLUG_SPLIT_STARTUP " Will Deacon
2026-10-09 10:07 ` [PATCH v2 16/23] arm64: cpu_ops: Make 'cpu_operations' pointer global instead of per-cpu Will Deacon
2026-10-09 10:07 ` [PATCH v2 17/23] arm64: cpu_ops: Introduce get_secondary_cpu_ops() Will Deacon
2026-10-09 10:07 ` [PATCH v2 18/23] firmware/psci: Cache PSCI v0.2+ version number to avoid redundant SMCs Will Deacon
2026-10-09 10:07 ` [PATCH v2 19/23] firmware/psci: Extend ->cpu_on() callback to take an additional argument Will Deacon
2026-10-09 10:07 ` [PATCH v2 20/23] arm64: cpu_ops: Expose optional argument to target cpu in ->cpu_boot() Will Deacon
2026-10-09 10:07 ` [PATCH v2 21/23] arm64: smp: Pass secondary CPU boot parameters via firmware if possible Will Deacon
2026-10-09 10:07 ` [PATCH v2 22/23] arm64: smp: Use generic HOTPLUG_PARALLEL machinery for CPU onlining Will Deacon
2026-10-09 10:07 ` [PATCH v2 23/23] arm64: smp: Harden parallel CPU bringup against broken PSCI firmware Will Deacon

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=20261009100738.31288-14-will@kernel.org \
    --to=will@kernel.org \
    --cc=ben.horgan@arm.com \
    --cc=bp@alien8.de \
    --cc=catalin.marinas@arm.com \
    --cc=dwmw@amazon.co.uk \
    --cc=fuad.tabba@linux.dev \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lpieralisi@kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=maz@kernel.org \
    --cc=pankaj.patil@oss.qualcomm.com \
    --cc=peterz@infradead.org \
    --cc=ruanjinjie@huawei.com \
    --cc=stefion@google.com \
    --cc=tarunsahu@google.com \
    --cc=tglx@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

all inboxes | Powered by JetHome®