mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Sean Christopherson <seanjc@google.com>
To: Sean Christopherson <seanjc@google.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	 David Woodhouse <dwmw2@infradead.org>,
	Paul Durrant <paul@xen.org>,
	 Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>, Will Deacon <will@kernel.org>,
	 Boqun Feng <boqun@kernel.org>
Cc: Waiman Long <longman@redhat.com>,
	kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
	 David Woodhouse <dwmw@amazon.co.uk>,
	Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
	 syzbot+208f7f3e5f59c11aeb90@syzkaller.appspotmail.com,
	 Carsten Stollmaier <stollmc@amazon.com>
Subject: [PATCH v2 01/20] locking/rt: Use raw_spin_lock_irqsave() in __rwbase_read_unlock()
Date: Fri, 29 May 2026 09:50:55 -0700	[thread overview]
Message-ID: <20260529165114.748639-2-seanjc@google.com> (raw)
In-Reply-To: <20260529165114.748639-1-seanjc@google.com>

From: David Woodhouse <dwmw@amazon.co.uk>

__rwbase_read_unlock() uses raw_spin_lock_irq()/raw_spin_unlock_irq()
which unconditionally disables and re-enables interrupts. When
read_unlock() is called from hardirq context (e.g. after a successful
read_trylock() in a timer callback), the raw_spin_unlock_irq()
incorrectly re-enables interrupts within the hardirq handler.

This causes lockdep warnings ('hardirqs_on_prepare' from hardirq
context) and can lead to IRQ state corruption.

Using read_trylock() in hardirq context on PREEMPT_RT is safe because
it does not record the lock owner. The read_unlock() acquires the
wait_lock which is hardirq safe. This change additionally allows
rwlock_t during early boot.

Switch to raw_spin_lock_irqsave()/raw_spin_unlock_irqrestore() to
preserve the caller's IRQ state.

Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Sean Christopherson <seanjc@google.com>
---
 kernel/locking/rwbase_rt.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/kernel/locking/rwbase_rt.c b/kernel/locking/rwbase_rt.c
index 82e078c0665a..25744862d627 100644
--- a/kernel/locking/rwbase_rt.c
+++ b/kernel/locking/rwbase_rt.c
@@ -153,8 +153,9 @@ static void __sched __rwbase_read_unlock(struct rwbase_rt *rwb,
 	struct rt_mutex_base *rtm = &rwb->rtmutex;
 	struct task_struct *owner;
 	DEFINE_RT_WAKE_Q(wqh);
+	unsigned long flags;
 
-	raw_spin_lock_irq(&rtm->wait_lock);
+	raw_spin_lock_irqsave(&rtm->wait_lock, flags);
 	/*
 	 * Wake the writer, i.e. the rtmutex owner. It might release the
 	 * rtmutex concurrently in the fast path (due to a signal), but to
@@ -167,7 +168,7 @@ static void __sched __rwbase_read_unlock(struct rwbase_rt *rwb,
 
 	/* Pairs with the preempt_enable in rt_mutex_wake_up_q() */
 	preempt_disable();
-	raw_spin_unlock_irq(&rtm->wait_lock);
+	raw_spin_unlock_irqrestore(&rtm->wait_lock, flags);
 	rt_mutex_wake_up_q(&wqh);
 }
 
-- 
2.54.0.823.g6e5bcc1fc9-goog


  reply	other threads:[~2026-05-29 16:51 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-29 16:50 [PATCH v2 00/20] KVM: x86/xen: Fix Xen/GP/PREEMPT_RT issues with rwlock_t Sean Christopherson
2026-05-29 16:50 ` Sean Christopherson [this message]
2026-05-29 19:32   ` [PATCH v2 01/20] locking/rt: Use raw_spin_lock_irqsave() in __rwbase_read_unlock() Peter Zijlstra
2026-05-29 19:34     ` Peter Zijlstra
2026-05-29 20:05       ` Sean Christopherson
2026-05-29 20:13         ` Peter Zijlstra
2026-05-29 20:38           ` Peter Zijlstra
2026-05-30  0:54             ` Sean Christopherson
2026-05-30 10:26               ` Paolo Bonzini
2026-05-30 12:47                 ` David Woodhouse
2026-05-30 14:40                   ` Paolo Bonzini
2026-06-01 10:52                     ` David Woodhouse
2026-06-01 13:01                       ` David Woodhouse
2026-06-01 13:40                         ` Sebastian Andrzej Siewior
2026-06-01 13:53                           ` David Woodhouse
2026-06-01 14:47                             ` Sebastian Andrzej Siewior
2026-06-01 15:11                               ` David Woodhouse
2026-06-01  9:40                   ` Peter Zijlstra
2026-06-01 10:04                     ` David Woodhouse
2026-05-30 13:02                 ` Paolo Bonzini
2026-06-01  8:40                 ` Peter Zijlstra
2026-06-01 11:11                   ` Sebastian Andrzej Siewior
2026-06-01 11:40                     ` Peter Zijlstra
2026-06-01 19:13                     ` Paolo Bonzini
2026-06-02  7:34                       ` Sebastian Andrzej Siewior
2026-06-04 23:58             ` David Woodhouse
2026-05-29 16:50 ` [PATCH v2 02/20] KVM: x86/xen: Use read_trylock() for GPC locks in hardirq/atomic paths Sean Christopherson
2026-05-29 23:28   ` Hillf Danton
2026-05-29 16:50 ` [PATCH v2 03/20] KVM: x86/xen: Remove unnecessary irqsave from GPC lock usage in xen.c Sean Christopherson
2026-05-29 16:50 ` [PATCH v2 04/20] KVM: x86: Remove unnecessary irqsave from kvm_setup_guest_pvclock() Sean Christopherson
2026-05-29 16:50 ` [PATCH v2 05/20] KVM: Remove unnecessary IRQ disabling from GPC lock in pfncache.c Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 06/20] KVM: x86/xen: Use guard() to grab kvm->srcu around gpc critical sections Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 07/20] KVM: x86/xen: Extract delivery of event to vCPU into a separate helper Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 08/20] KVM: x86/xen: Explicitly tag "shared info" page as never being dirty tracked Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 09/20] KVM: x86/xen: Don't dirty track "vCPU info" page Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 10/20] KVM: Move {g,p}fn <=> {g,h}pa conversion helpers to kvm_types.h Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 11/20] KVM: Add CLASS() constructs to automagically handle lock+check of gpc Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 12/20] KVM: x86/xen: Convert kvm_xen_shared_info_init() to gpc's CLASS() APIs Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 13/20] KVM: x86/xen: Don't bother waiting on gpc->lock in SCHEDOP_poll Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 14/20] KVM: x86/xen: Convert wait_pending_event() to gpc's CLASS() APIs Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 15/20] KVM: x86/xen: Convert kvm_xen_set_evtchn_fast() " Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 16/20] KVM: x86/xen: Convert xen_get_guest_pvclock() " Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 17/20] KVM: x86/xen: Drop local "kick_vcpu" from __kvm_xen_set_evtchn_fast() Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 18/20] KVM: x86/xen: Convert event injection to gpc's CLASS() APIs Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 19/20] KVM: Add "extended" gpc CLASS() APIs for sometimes-atomic cases Sean Christopherson
2026-05-29 16:51 ` [PATCH v2 20/20] KVM: x86: Use gfn_to_pfn_cache for record_steal_time Sean Christopherson

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=20260529165114.748639-2-seanjc@google.com \
    --to=seanjc@google.com \
    --cc=bigeasy@linutronix.de \
    --cc=boqun@kernel.org \
    --cc=dwmw2@infradead.org \
    --cc=dwmw@amazon.co.uk \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=longman@redhat.com \
    --cc=mingo@redhat.com \
    --cc=paul@xen.org \
    --cc=pbonzini@redhat.com \
    --cc=peterz@infradead.org \
    --cc=stollmc@amazon.com \
    --cc=syzbot+208f7f3e5f59c11aeb90@syzkaller.appspotmail.com \
    --cc=will@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

Powered by JetHome