linux-kernel - [PATCH 5.15 024/177] s390/entry: fix duplicate tracking of irq nesting level

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20211220143040.896251354@linuxfoundation.org>
Date:   Mon, 20 Dec 2021 15:32:54 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Mark Rutland <mark.rutland@....com>,
        Sven Schnelle <svens@...ux.ibm.com>,
        Heiko Carstens <hca@...ux.ibm.com>
Subject: [PATCH 5.15 024/177] s390/entry: fix duplicate tracking of irq nesting level

From: Sven Schnelle <svens@...ux.ibm.com>

commit c9b12b59e2ea4c3c7cedec7efb071b649652f3a9 upstream.

In the current code, when exiting from idle, rcu_irq_enter() is
called twice during irq entry:

irq_entry_enter()-> rcu_irq_enter()
irq_enter() -> rcu_irq_enter()

This may lead to wrong results from rcu_is_cpu_rrupt_from_idle()
because of a wrong dynticks nmi nesting count. Fix this by only
calling irq_enter_rcu().

Cc: <stable@...r.kernel.org> # 5.12+
Reported-by: Mark Rutland <mark.rutland@....com>
Fixes: 56e62a737028 ("s390: convert to generic entry")
Signed-off-by: Sven Schnelle <svens@...ux.ibm.com>
Signed-off-by: Heiko Carstens <hca@...ux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
 arch/s390/kernel/irq.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -138,7 +138,7 @@ void noinstr do_io_irq(struct pt_regs *r
 	struct pt_regs *old_regs = set_irq_regs(regs);
 	int from_idle;
 
-	irq_enter();
+	irq_enter_rcu();
 
 	if (user_mode(regs))
 		update_timer_sys();
@@ -155,7 +155,8 @@ void noinstr do_io_irq(struct pt_regs *r
 			do_irq_async(regs, IO_INTERRUPT);
 	} while (MACHINE_IS_LPAR && irq_pending(regs));
 
-	irq_exit();
+	irq_exit_rcu();
+
 	set_irq_regs(old_regs);
 	irqentry_exit(regs, state);
 
@@ -169,7 +170,7 @@ void noinstr do_ext_irq(struct pt_regs *
 	struct pt_regs *old_regs = set_irq_regs(regs);
 	int from_idle;
 
-	irq_enter();
+	irq_enter_rcu();
 
 	if (user_mode(regs))
 		update_timer_sys();
@@ -184,7 +185,7 @@ void noinstr do_ext_irq(struct pt_regs *
 
 	do_irq_async(regs, EXT_INTERRUPT);
 
-	irq_exit();
+	irq_exit_rcu();
 	set_irq_regs(old_regs);
 	irqentry_exit(regs, state);