linux-kernel - Re: [patch RFC 28/29] printk: Provide functions for atomic write enforcement

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: <87h70slv4i.fsf@jogness.linutronix.de>
Date:   Tue, 27 Sep 2022 17:07:01 +0206
From:   John Ogness <john.ogness@...utronix.de>
To:     Thomas Gleixner <tglx@...utronix.de>,
        LKML <linux-kernel@...r.kernel.org>
Cc:     Petr Mladek <pmladek@...e.com>,
        Sergey Senozhatsky <senozhatsky@...omium.org>,
        Steven Rostedt <rostedt@...dmis.org>,
        Linus Torvalds <torvalds@...uxfoundation.org>,
        Peter Zijlstra <peterz@...radead.org>,
        "Paul E. McKenney" <paulmck@...nel.org>,
        Daniel Vetter <daniel@...ll.ch>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Helge Deller <deller@....de>,
        Jason Wessel <jason.wessel@...driver.com>,
        Daniel Thompson <daniel.thompson@...aro.org>
Subject: Re: [patch RFC 28/29] printk: Provide functions for atomic write
 enforcement

Below is a fix that was used for the LPC2022 demo to avoid unnecessarily
performing the console_lock/console_unlock dance.

Add a new global boolean @have_bkl_console to be able to quickly
identify if any legacy (bkl) consoles are registered. If there are none,
the console_lock/console_unlock stuff can be skipped. The following
patch does this and can be applied on top.

--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -1830,6 +1830,7 @@ static struct lockdep_map console_owner_dep_map = {
 static DEFINE_RAW_SPINLOCK(console_owner_lock);
 static struct task_struct *console_owner;
 static bool console_waiter;
+static bool have_bkl_console;
 
 /**
  * console_lock_spinning_enable - mark beginning of code where another
@@ -2285,7 +2286,7 @@ asmlinkage int vprintk_emit(int facility, int level,
 	cons_atomic_flush();
 
 	/* If called from the scheduler, we can not call up(). */
-	if (!in_sched) {
+	if (!in_sched && have_bkl_console) {
 		/*
 		 * Try to acquire and then immediately release the console
 		 * semaphore. The release will print out buffers. With the
@@ -2575,7 +2576,7 @@ void resume_console(void)
  */
 static int console_cpu_notify(unsigned int cpu)
 {
-	if (!cpuhp_tasks_frozen) {
+	if (!cpuhp_tasks_frozen && have_bkl_console) {
 		/* If trylock fails, someone else is doing the printing */
 		if (console_trylock())
 			console_unlock();
@@ -3023,6 +3024,9 @@ void console_unblank(void)
 {
 	struct console *c;
 
+	if (!have_bkl_console)
+		return;
+
 	/*
 	 * console_unblank can no longer be called in interrupt context unless
 	 * oops_in_progress is set to 1..
@@ -3052,6 +3056,9 @@ void console_unblank(void)
  */
 void console_flush_on_panic(enum con_flush_mode mode)
 {
+	if (!have_bkl_console)
+		return;
+
 	/*
 	 * If someone else is holding the console lock, trylock will fail
 	 * and may_schedule may be set.  Ignore and proceed to unlock so
@@ -3311,6 +3318,10 @@ void register_console(struct console *newcon)
 	/* Initialize the nobkl data in @newcon */
 	cons_nobkl_init(newcon);
 
+	/* Has a legacy (BKL) console registered? */
+	if (!(newcon->flags & CON_NO_BKL))
+		have_bkl_console = true;
+
 	/*
 	 * Put this console in the list and keep the referred driver at the
 	 * head of the list.
@@ -3603,7 +3613,7 @@ static void wake_up_klogd_work_func(struct irq_work *irq_work)
 {
 	int pending = this_cpu_xchg(printk_pending, 0);
 
-	if (pending & PRINTK_PENDING_OUTPUT) {
+	if (have_bkl_console && (pending & PRINTK_PENDING_OUTPUT)) {
 		/* If trylock fails, someone else is doing the printing */
 		if (console_trylock())
 			console_unlock();

John Ogness