From: John Ogness <john.ogness@linutronix.de>
To: Petr Mladek <pmladek@suse.com>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>,
Steven Rostedt <rostedt@goodmis.org>,
Thomas Gleixner <tglx@linutronix.de>,
linux-kernel@vger.kernel.org
Subject: [PATCH printk v2 05/11] printk: nbcon: Provide function for atomic flushing
Date: Wed, 20 Sep 2023 01:14:50 +0206 [thread overview]
Message-ID: <20230919230856.661435-6-john.ogness@linutronix.de> (raw)
In-Reply-To: <20230919230856.661435-1-john.ogness@linutronix.de>
From: Thomas Gleixner <tglx@linutronix.de>
Provide nbcon_atomic_flush() to perform atomic write flushing
of all registered nbcon consoles. Like with legacy consoles,
the nbcon consoles are flushed one record per console. This
allows all nbcon consoles to generate pseudo-simultaneously,
rather than one console waiting for the full ringbuffer to
dump to another console before printing anything.
Note that if the current CPU is in a nested elevated priority
state (EMERGENCY/PANIC), nbcon_atomic_flush() does nothing.
This is in case the printing itself generates urgent messages
(OOPS/WARN/PANIC), that those messages are fully stored into
the ringbuffer before any printing resumes.
Co-developed-by: John Ogness <john.ogness@linutronix.de>
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Signed-off-by: Thomas Gleixner (Intel) <tglx@linutronix.de>
---
include/linux/printk.h | 6 +++
kernel/printk/nbcon.c | 101 ++++++++++++++++++++++++++++++++++++++++-
2 files changed, 105 insertions(+), 2 deletions(-)
diff --git a/include/linux/printk.h b/include/linux/printk.h
index 8ef499ab3c1e..58e5f34d6df1 100644
--- a/include/linux/printk.h
+++ b/include/linux/printk.h
@@ -192,6 +192,7 @@ void show_regs_print_info(const char *log_lvl);
extern asmlinkage void dump_stack_lvl(const char *log_lvl) __cold;
extern asmlinkage void dump_stack(void) __cold;
void printk_trigger_flush(void);
+extern void nbcon_atomic_flush_all(void);
#else
static inline __printf(1, 0)
int vprintk(const char *s, va_list args)
@@ -271,6 +272,11 @@ static inline void dump_stack(void)
static inline void printk_trigger_flush(void)
{
}
+
+static inline void nbcon_atomic_flush_all(void)
+{
+}
+
#endif
#ifdef CONFIG_SMP
diff --git a/kernel/printk/nbcon.c b/kernel/printk/nbcon.c
index 9359906b575b..2a9ff18fc78c 100644
--- a/kernel/printk/nbcon.c
+++ b/kernel/printk/nbcon.c
@@ -571,7 +571,6 @@ static struct printk_buffers panic_nbcon_pbufs;
* in an unsafe state. Otherwise, on success the caller may assume
* the console is not in an unsafe state.
*/
-__maybe_unused
static bool nbcon_context_try_acquire(struct nbcon_context *ctxt)
{
unsigned int cpu = smp_processor_id();
@@ -873,7 +872,6 @@ EXPORT_SYMBOL_GPL(nbcon_exit_unsafe);
* When true is returned, @wctxt->ctxt.backlog indicates whether there are
* still records pending in the ringbuffer,
*/
-__maybe_unused
static bool nbcon_emit_next_record(struct nbcon_write_context *wctxt)
{
struct nbcon_context *ctxt = &ACCESS_PRIVATE(wctxt, ctxt);
@@ -988,6 +986,105 @@ static __ref struct nbcon_cpu_state *nbcon_get_cpu_state(void)
return this_cpu_ptr(&nbcon_pcpu_state);
}
+/**
+ * nbcon_atomic_emit_one - Print one record for a console in atomic mode
+ * @wctxt: An initialized write context struct to use
+ * for this context
+ *
+ * Returns false if the given console could not print a record or there are
+ * no more records to print, otherwise true.
+ *
+ * This is an internal helper to handle the locking of the console before
+ * calling nbcon_emit_next_record().
+ */
+static bool nbcon_atomic_emit_one(struct nbcon_write_context *wctxt)
+{
+ struct nbcon_context *ctxt = &ACCESS_PRIVATE(wctxt, ctxt);
+
+ if (!nbcon_context_try_acquire(ctxt))
+ return false;
+
+ /*
+ * nbcon_emit_next_record() returns false when the console was
+ * handed over or taken over. In both cases the context is no
+ * longer valid.
+ */
+ if (!nbcon_emit_next_record(wctxt))
+ return false;
+
+ nbcon_context_release(ctxt);
+
+ return prb_read_valid(prb, ctxt->seq, NULL);
+}
+
+/**
+ * __nbcon_atomic_flush_all - Flush all nbcon consoles in atomic mode
+ * @allow_unsafe_takeover: True, to allow unsafe hostile takeovers
+ */
+static void __nbcon_atomic_flush_all(bool allow_unsafe_takeover)
+{
+ struct nbcon_write_context wctxt = { };
+ struct nbcon_context *ctxt = &ACCESS_PRIVATE(&wctxt, ctxt);
+ struct nbcon_cpu_state *cpu_state;
+ struct console *con;
+ bool any_progress;
+ int cookie;
+
+ cpu_state = nbcon_get_cpu_state();
+
+ /*
+ * Let the outermost flush of this priority print. This avoids
+ * nasty hackery for nested WARN() where the printing itself
+ * generates one and ensures such nested messages are stored to
+ * the ringbuffer before any printing resumes.
+ *
+ * cpu_state->prio <= NBCON_PRIO_NORMAL is not subject to nesting
+ * and can proceed in order to allow any atomic printing for
+ * regular kernel messages.
+ */
+ if (cpu_state->prio > NBCON_PRIO_NORMAL &&
+ cpu_state->nesting[cpu_state->prio] != 1)
+ return;
+
+ do {
+ any_progress = false;
+
+ cookie = console_srcu_read_lock();
+ for_each_console_srcu(con) {
+ short flags = console_srcu_read_flags(con);
+ bool progress;
+
+ if (!(flags & CON_NBCON))
+ continue;
+
+ if (!console_is_usable(con, flags))
+ continue;
+
+ memset(ctxt, 0, sizeof(*ctxt));
+ ctxt->console = con;
+ ctxt->spinwait_max_us = 2000;
+ ctxt->prio = cpu_state->prio;
+ ctxt->allow_unsafe_takeover = allow_unsafe_takeover;
+
+ progress = nbcon_atomic_emit_one(&wctxt);
+ if (!progress)
+ continue;
+ any_progress = true;
+ }
+ console_srcu_read_unlock(cookie);
+ } while (any_progress);
+}
+
+/**
+ * nbcon_atomic_flush_all - Flush all nbcon consoles in atomic mode
+ *
+ * Context: Any context where migration is disabled.
+ */
+void nbcon_atomic_flush_all(void)
+{
+ __nbcon_atomic_flush_all(false);
+}
+
/**
* nbcon_atomic_enter - Enter a context that enforces atomic printing
* @prio: Priority of the context
--
2.39.2
next prev parent reply other threads:[~2023-09-19 23:09 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-19 23:08 [PATCH printk v2 00/11] wire up nbcon atomic printing John Ogness
2023-09-19 23:08 ` [PATCH printk v2 01/11] printk: Make console_is_usable() available to nbcon John Ogness
2023-09-22 8:33 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 02/11] printk: Let console_is_usable() handle nbcon John Ogness
2023-09-22 8:37 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 03/11] printk: Add @flags argument for console_is_usable() John Ogness
2023-09-22 8:41 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 04/11] printk: nbcon: Provide functions to mark atomic write sections John Ogness
2023-09-22 9:33 ` Petr Mladek
2023-09-25 9:25 ` John Ogness
2023-09-25 16:04 ` Petr Mladek
2023-10-05 12:51 ` John Ogness
2023-10-06 12:51 ` panic context: was: " Petr Mladek
2023-10-06 12:53 ` Petr Mladek
2023-10-08 10:13 ` John Ogness
2023-10-09 15:32 ` Petr Mladek
2023-10-10 16:02 ` John Ogness
2023-10-16 8:58 ` Dave Young
2023-10-16 10:09 ` John Ogness
2023-10-06 15:52 ` Petr Mladek
2023-09-19 23:08 ` John Ogness [this message]
2023-09-22 12:32 ` [PATCH printk v2 05/11] printk: nbcon: Provide function for atomic flushing Petr Mladek
2023-09-25 11:11 ` John Ogness
2023-09-19 23:08 ` [PATCH printk v2 06/11] printk: nbcon: Wire up nbcon console " John Ogness
2023-09-22 17:41 ` Petr Mladek
2023-09-25 13:37 ` John Ogness
2023-09-26 12:14 ` Petr Mladek
2023-10-05 13:59 ` John Ogness
2023-09-19 23:08 ` [PATCH printk v2 07/11] printk: nbcon: Wire up nbcon into console_flush_all() John Ogness
2023-09-26 11:34 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 08/11] panic: Add atomic write enforcement to warn/panic John Ogness
2023-09-27 12:02 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 09/11] panic: Add atomic write enforcement to oops John Ogness
2023-09-19 23:36 ` John Ogness
2023-09-20 13:28 ` Andy Shevchenko
2023-09-20 14:20 ` John Ogness
2023-09-20 14:45 ` Andy Shevchenko
2023-09-27 13:15 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 10/11] rcu: Add atomic write enforcement for rcu stalls John Ogness
2023-09-27 15:00 ` Petr Mladek
2023-09-19 23:08 ` [PATCH printk v2 11/11] lockdep: Add atomic write enforcement for lockdep splats John Ogness
2023-09-29 8:31 ` Petr Mladek
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=20230919230856.661435-6-john.ogness@linutronix.de \
--to=john.ogness@linutronix.de \
--cc=linux-kernel@vger.kernel.org \
--cc=pmladek@suse.com \
--cc=rostedt@goodmis.org \
--cc=senozhatsky@chromium.org \
--cc=tglx@linutronix.de \
/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®