From: Thomas Gleixner <tglx@linutronix.de>
To: LKML <linux-kernel@vger.kernel.org>
Cc: John Ogness <john.ogness@linutronix.de>,
Petr Mladek <pmladek@suse.com>,
Sergey Senozhatsky <senozhatsky@chromium.org>,
Steven Rostedt <rostedt@goodmis.org>,
Linus Torvalds <torvalds@linuxfoundation.org>,
Peter Zijlstra <peterz@infradead.org>,
"Paul E. McKenney" <paulmck@kernel.org>,
Daniel Vetter <daniel@ffwll.ch>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Helge Deller <deller@gmx.de>,
Jason Wessel <jason.wessel@windriver.com>,
Daniel Thompson <daniel.thompson@linaro.org>,
John Ogness <jogness@linutronix.de>
Subject: [patch RFC 23/29] printk: Add non-BKL console print state functions
Date: Sun, 11 Sep 2022 00:28:07 +0200 (CEST) [thread overview]
Message-ID: <20220910222301.710939648@linutronix.de> (raw)
In-Reply-To: <20220910221947.171557773@linutronix.de>
Provide three functions which are related to the safe handover mechanism
and allow console drivers to denote takeover unsafe sections:
- console_can_proceed()
Invoked by a console driver to check whether a handover request is pending
or whether the console was taken over in a hostile fashion.
- console_enter/exit_unsafe()
Invoked by a console driver to reflect that the driver output function is
about to enter or to leave an critical region where a hostile take over
is unsafe. These functions are also cancelation points.
The unsafe state is reflected in the console state and allows a takeover
attempt to make informed decisions whether to take over and/or output on
such console at all. The unsafe state is also reflected on output to the
driver in the write context to the atomic_write() output function so the
driver can make informed decisions about the required actions or take an
special emergency path.
Co-Developed-by: John Ogness <jogness@linutronix.de>
Signed-off-by: John Ogness <jogness@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
include/linux/console.h | 20 ++++++
kernel/printk/printk_nobkl.c | 134 +++++++++++++++++++++++++++++++++++++++++++
2 files changed, 154 insertions(+)
--- a/include/linux/console.h
+++ b/include/linux/console.h
@@ -300,6 +300,22 @@ struct cons_context {
unsigned int spinwait : 1;
};
+/**
+ * struct cons_write_context - Context handed to the write callbacks
+ * @ctxt: The core console context
+ * @outbuf: Pointer to the text buffer for output
+ * @len: Length to write
+ * @pos: Current write position in @outbuf
+ * @unsafe: Invoked in unsafe state due to force takeover
+ */
+struct cons_write_context {
+ struct cons_context __private ctxt;
+ char *outbuf;
+ unsigned int len;
+ unsigned int pos;
+ bool unsafe;
+};
+
#define CONS_MAX_NEST_LVL 8
/**
@@ -423,6 +439,10 @@ extern void console_list_unlock(void) __
#define for_each_console_kgdb(con) \
hlist_for_each_entry(con, &console_list, node)
+extern bool console_can_proceed(struct cons_write_context *wctxt);
+extern bool console_enter_unsafe(struct cons_write_context *wctxt);
+extern bool console_exit_unsafe(struct cons_write_context *wctxt);
+
extern int console_set_on_cmdline;
extern struct console *early_console;
--- a/kernel/printk/printk_nobkl.c
+++ b/kernel/printk/printk_nobkl.c
@@ -874,6 +874,140 @@ static void cons_free_percpu_data(struct
}
/**
+ * console_can_proceed - Check whether printing can proceed
+ * @wctxt: The write context which was handed to the write function
+ *
+ * Returns: True if the state is correct. False if a handover
+ * has been requested or if the console was taken
+ * over.
+ *
+ * Must be invoked after the record was dumped into the assigned record
+ * buffer and at appropriate safe places in the driver. For unsafe driver
+ * sections see console_enter_unsafe().
+ *
+ * When this function returns false then the calling context is not allowed
+ * to go forward and has to back out immediately and carefully. The buffer
+ * content is not longer trusted either and the console lock is not longer
+ * held.
+ */
+bool console_can_proceed(struct cons_write_context *wctxt)
+{
+ struct cons_context *ctxt = &ACCESS_PRIVATE(wctxt, ctxt);
+ struct console *con = ctxt->console;
+ struct cons_state state;
+
+ cons_state_read(con, STATE_REAL, &state);
+ /* Store it for analyis or reuse */
+ copy_full_state(ctxt->old_state, state);
+
+ /*
+ * If the state aside of req_prio is not longer matching, console
+ * was taken over.
+ */
+ if (!cons_state_full_match(state, ctxt->state))
+ return false;
+
+ /*
+ * Having a safe point for take over and eventually a few
+ * duplicated characters or a full line is way better than a
+ * hostile takeover. Post processing can take care of the garbage.
+ * Continue if the requested priority is not sufficient.
+ */
+ if (state.req_prio <= state.cur_prio)
+ return true;
+
+ /* Release and hand over */
+ cons_release(ctxt);
+ /*
+ * This does not check whether the handover succeeded. The
+ * outermost callsite has to do the final check whether printing
+ * should continue or not. The console is unlocked already so go
+ * back all the way instead of trying to implement heuristics in
+ * tons of places.
+ */
+ return false;
+}
+
+/**
+ * __console_update_unsafe - Update the unsafe bit in @con->atomic_state
+ * @wctxt: The write context which was handed to the write function
+ *
+ * Returns: True if the state is correct. False if a handover
+ * has been requested or if the console was taken
+ * over.
+ *
+ * Must be invoked before a unsafe driver section is entered.
+ *
+ * When this function returns false then the calling context is not allowed
+ * to go forward and has to back out immediately and carefully. The buffer
+ * content is not longer trusted either and the console lock is not longer
+ * held.
+ *
+ * Internal helper to avoid duplicated code
+ */
+static bool __console_update_unsafe(struct cons_write_context *wctxt, bool unsafe)
+{
+ struct cons_context *ctxt = &ACCESS_PRIVATE(wctxt, ctxt);
+ struct console *con = ctxt->console;
+ struct cons_state new;
+
+ do {
+ if (!console_can_proceed(wctxt))
+ return false;
+ /*
+ * console_can_proceed() saved the real state in
+ * ctxt->old_state
+ */
+ copy_full_state(new, ctxt->old_state);
+ new.unsafe = unsafe;
+
+ } while (!cons_state_try_cmpxchg(con, STATE_REAL, &ctxt->old_state, &new));
+
+ copy_full_state(ctxt->state, new);
+ return true;
+}
+
+/**
+ * console_enter_unsafe - Enter an unsafe region in the driver
+ * @wctxt: The write context which was handed to the write function
+ *
+ * Returns: True if the state is correct. False if a handover
+ * has been requested or if the console was taken
+ * over.
+ *
+ * Must be invoked before a unsafe driver section is entered.
+ *
+ * When this function returns false then the calling context is not allowed
+ * to go forward and has to back out immediately and carefully. The buffer
+ * content is not longer trusted either and the console lock is not longer
+ * held.
+ */
+bool console_enter_unsafe(struct cons_write_context *wctxt)
+{
+ return __console_update_unsafe(wctxt, true);
+}
+
+/**
+ * console_exit_unsafe - Exit an unsafe region in the driver
+ * @wctxt: The write context which was handed to the write function
+ *
+ * Returns: True if the state is correct. False if a handover
+ * has been requested or if the console was taken
+ * over.
+ *
+ * Must be invoked before a unsafe driver section is exited.
+ *
+ * When this function returns false then the calling context is not allowed
+ * to go forward and has to back out immediately and carefully. The buffer
+ * content is not longer trusted either and the console lock is not longer
+ * held.
+ */
+bool console_exit_unsafe(struct cons_write_context *wctxt)
+{
+ return __console_update_unsafe(wctxt, false);
+}
+
+/**
* cons_nobkl_init - Initialize the NOBKL console state
* @con: Console to initialize
*/
next prev parent reply other threads:[~2022-09-10 22:30 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-10 22:27 [patch RFC 00/29] printk: A new approach - WIP Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 01/29] printk: Make pr_flush() static Thomas Gleixner
2022-09-14 11:27 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 02/29] printk: Declare log_wait properly Thomas Gleixner
2022-09-14 11:29 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 03/29] printk: Remove write only variable nr_ext_console_drivers Thomas Gleixner
2022-09-14 11:33 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 04/29] printk: Remove bogus comment vs. boot consoles Thomas Gleixner
2022-09-14 11:40 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 05/29] printk: Mark __printk percpu data ready __ro_after_init Thomas Gleixner
2022-09-14 11:41 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 06/29] printk: Protect [un]register_console() with a mutex Thomas Gleixner
2022-09-14 12:05 ` Sergey Senozhatsky
2022-09-14 12:31 ` Sergey Senozhatsky
2022-09-19 12:49 ` John Ogness
2022-09-27 9:56 ` Petr Mladek
2022-09-27 15:19 ` Petr Mladek
2022-09-10 22:27 ` [patch RFC 07/29] printk: Convert console list walks for readers to list lock Thomas Gleixner
2022-09-14 12:46 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 08/29] parisc: Put console abuse into one place Thomas Gleixner
2022-09-14 14:56 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 09/29] serial: kgdboc: Lock consoles in probe function Thomas Gleixner
2022-09-14 14:59 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 10/29] kgbd: Pretend that console list walk is safe Thomas Gleixner
2022-09-14 15:03 ` Sergey Senozhatsky
2022-09-10 22:27 ` [patch RFC 11/29] printk: Convert console_drivers list to hlist Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 12/29] printk: Prepare for SCRU console list protection Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 13/29] printk: Move buffer size defines Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 14/29] printk: Document struct console Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 15/29] printk: Add struct cons_text_buf Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 16/29] printk: Use " Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 17/29] printk: Use an output descriptor struct for emit Thomas Gleixner
2022-09-10 22:27 ` [patch RFC 18/29] printk: Handle dropped message smarter Thomas Gleixner
2022-09-10 22:28 ` [patch RFC 19/29] printk: Add basic infrastructure for non-BKL consoles Thomas Gleixner
2022-11-07 15:58 ` functionality: was: " Petr Mladek
2022-11-07 16:10 ` cosmetic: " Petr Mladek
2022-09-10 22:28 ` [patch RFC 20/29] printk: Add non-BKL console acquire/release logic Thomas Gleixner
2022-09-27 13:49 ` John Ogness
2022-09-10 22:28 ` [patch RFC 21/29] printk: Add buffer management for noBKL consoles Thomas Gleixner
2022-09-10 22:28 ` [patch RFC 22/29] printk: Add sequence handling for non-BKL consoles Thomas Gleixner
2022-09-10 22:28 ` Thomas Gleixner [this message]
2022-09-10 22:28 ` [patch RFC 24/29] printk: Put seq and dropped into cons_text_desc Thomas Gleixner
2022-09-10 22:28 ` [patch RFC 25/29] printk: Provide functions to emit a ringbuffer record on non-BKL consoles Thomas Gleixner
2022-09-10 22:28 ` [patch RFC 26/29] printk: Add threaded printing support Thomas Gleixner
2022-09-10 22:28 ` [patch RFC 27/29] printk: Add write context storage for atomic writes Thomas Gleixner
2022-09-10 22:28 ` [patch RFC 28/29] printk: Provide functions for atomic write enforcement Thomas Gleixner
2022-09-27 13:55 ` John Ogness
2022-09-27 14:40 ` John Ogness
2022-09-27 14:49 ` John Ogness
2022-09-27 15:01 ` John Ogness
2022-09-10 22:28 ` [patch RFC 29/29] printk: Add atomic write enforcement to warn/panic Thomas Gleixner
2022-09-10 22:56 ` [patch RFC 00/29] printk: A new approach - WIP Thomas Gleixner
2022-09-11 9:01 ` Paul E. McKenney
2022-09-11 12:01 ` Linus Torvalds
2022-09-12 16:40 ` printk meeting at LPC 2022 John Ogness
2022-09-15 11:00 ` Sergey Senozhatsky
2022-09-15 11:09 ` Steven Rostedt
2022-09-15 15:25 ` Sergey Senozhatsky
2022-09-23 14:49 ` John Ogness
2022-09-23 15:16 ` Linus Torvalds
2022-09-23 15:20 ` Sebastian Andrzej Siewior
2022-09-23 15:31 ` Steven Rostedt
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=20220910222301.710939648@linutronix.de \
--to=tglx@linutronix.de \
--cc=daniel.thompson@linaro.org \
--cc=daniel@ffwll.ch \
--cc=deller@gmx.de \
--cc=gregkh@linuxfoundation.org \
--cc=jason.wessel@windriver.com \
--cc=jogness@linutronix.de \
--cc=john.ogness@linutronix.de \
--cc=linux-kernel@vger.kernel.org \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=pmladek@suse.com \
--cc=rostedt@goodmis.org \
--cc=senozhatsky@chromium.org \
--cc=torvalds@linuxfoundation.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®