mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Petr Mladek <pmladek@suse.com>
To: Xiaochun Li <lixiaochun@open-hieco.net>
Cc: rostedt@goodmis.org, john.ogness@linutronix.de,
	senozhatsky@chromium.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] printk: Remove remaining boot consoles when a real console exists
Date: Thu, 20 Aug 2026 17:13:05 +0200	[thread overview]
Message-ID: <aocZgQNT4abxmOmK@pathway.suse.cz> (raw)
In-Reply-To: <20260805084846.1802042-1-lixiaochun@open-hieco.net>

On Wed 2026-08-05 16:48:46, Xiaochun Li wrote:
> Boot consoles are temporary and should be removed once a real console is
> available. However, the late init cleanup currently only unregisters boot
> consoles that use init section memory. Other boot consoles are expected
> to be removed when the real preferred console is registered.
> 
> This does not cover cases where a real console has registered, but the
> boot console was not removed because the real console did not become the
> preferred console. For example, with multiple console= parameters using
> the same driver, a real 8250 console may be enabled while the early
> console remains registered. The result is duplicate printk output from
> both consoles.
> 
> In the mailing list discussion, two possible approaches were suggested
> to fix this problem [1]. This patch implements the first one: during
> printk_late_init(), check whether at least one real console is already
> registered. If so, unregister all remaining boot consoles. If no real
> console exists yet, keep the existing behavior and unregister only boot
> consoles that reference init section memory, avoiding a period with no
> console output while waiting for a deferred or modular real console.
> 
> Suggested-by: Petr Mladek <pmladek@suse.com>
> Link: https://lore.kernel.org/lkml/ahBsSW2bB2CRfW-k@pathway.suse.cz/ # [1]
> Signed-off-by: Xiaochun Li <lixiaochun@open-hieco.net>
> ---
>  kernel/printk/printk.c | 34 +++++++++++++++++++++++-----------
>  1 file changed, 23 insertions(+), 11 deletions(-)
> 
> diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c
> index 2fe9a963c823..5a5ec7ac99bd 100644
> --- a/kernel/printk/printk.c
> +++ b/kernel/printk/printk.c
> @@ -4420,36 +4420,48 @@ void __init console_init(void)
>   * be a brief interval in which no messages are logged to the console, which
>   * makes it difficult to diagnose problems that occur during this time.
>   *
> - * To mitigate this problem somewhat, only unregister consoles whose memory
> - * intersects with the init section. Note that all other boot consoles will
> - * get unregistered when the real preferred console is registered.
> + * If a real console has already registered, remove all remaining boot consoles.
> + * Otherwise, mitigate the no-console interval by removing only boot consoles
> + * whose memory intersects with the init section.
>   */
>  static int __init printk_late_init(void)
>  {
> +	bool realcon_registered = false;

Nit: I would call the variable "have_real_console" to follow the
     existing naming scheme, see have_boot_console,
     have_legacy_console, ...

     Please, change it in v2 if we need it, ...


>  	struct hlist_node *tmp;
>  	struct console *con;
>  	int ret;
>  
>  	console_list_lock();
> +	for_each_console(con) {
> +		if (!(con->flags & CON_BOOT)) {
> +			realcon_registered = true;
> +			break;
> +		}
> +	}

Sashiko AI worries:

| Can this cause a period of missing console output if a non-visible console
| (like netconsole or pstore) registers early?

I do not agree. Users are able to read the messages on netconsole
or pstore, so I consider them valid real consoles.

| It looks like this checks only for the absence of CON_BOOT to consider a
| real console present. If a real hardware serial console is deferred or
| loaded as a module later, wouldn't we need to check for CON_CONSDEV to
| ensure the preferred hardware console is actually available before dropping
| the boot console?

It does not make much sense. The CON_CONSDEV flag does not
guarantee anything. It is just the best effort to mart console
which will get used by /dev/console.

>  	hlist_for_each_entry_safe(con, tmp, &console_list, node) {
>  		if (!(con->flags & CON_BOOT))
>  			continue;
>  
> -		/* Check addresses that might be used for enabled consoles. */
> -		if (init_section_intersects(con, sizeof(*con)) ||
> -		    init_section_contains(con->write, 0) ||
> -		    init_section_contains(con->read, 0) ||
> -		    init_section_contains(con->device, 0) ||
> -		    init_section_contains(con->unblank, 0) ||
> -		    init_section_contains(con->data, 0)) {
> +		if (!realcon_registered) {
> +			/* Check addresses that might be used for enabled consoles. */
> +			if (!init_section_intersects(con, sizeof(*con)) &&
> +			    !init_section_contains(con->write, 0) &&
> +			    !init_section_contains(con->read, 0) &&
> +			    !init_section_contains(con->device, 0) &&
> +			    !init_section_contains(con->unblank, 0) &&
> +			    !init_section_contains(con->data, 0))
> +				continue;
> +
>  			/*
>  			 * Please, consider moving the reported consoles out
>  			 * of the init section.
>  			 */
>  			pr_warn("bootconsole [%s%d] uses init memory and must be disabled even before the real one is ready\n",
>  				con->name, con->index);
> -			unregister_console_locked(con);
>  		}
> +
> +		unregister_console_locked(con);

Another comment from Sashiko AI:

| Does this unconditional unregistration bypass the keep_bootcon command-line
| parameter?
| 
| If a user boots with keep_bootcon to preserve early boot consoles for
| debugging, it appears this code will bypass the keep_bootcon flag check
| traditionally used in register_console() and unexpectedly unregister the
| debugging consoles once a real console registers.

This one describes a real problem. We have to check the "keep_bootcon"
flag. I suggest something like:

--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -4420,36 +4420,51 @@ void __init console_init(void)
  * be a brief interval in which no messages are logged to the console, which
  * makes it difficult to diagnose problems that occur during this time.
  *
- * To mitigate this problem somewhat, only unregister consoles whose memory
- * intersects with the init section. Note that all other boot consoles will
- * get unregistered when the real preferred console is registered.
+ * If a real console has already registered, remove all remaining boot consoles.
+ * Otherwise, mitigate the no-console interval by removing only boot consoles
+ * whose memory intersects with the init section.
  */
 static int __init printk_late_init(void)
 {
+	bool have_real_console = false;
 	struct hlist_node *tmp;
 	struct console *con;
 	int ret;
 
 	console_list_lock();
+	for_each_console(con) {
+		if (!(con->flags & CON_BOOT)) {
+			have_real_console = true;
+			break;
+		}
+	}
+
 	hlist_for_each_entry_safe(con, tmp, &console_list, node) {
 		if (!(con->flags & CON_BOOT))
 			continue;
 
-		/* Check addresses that might be used for enabled consoles. */
-		if (init_section_intersects(con, sizeof(*con)) ||
-		    init_section_contains(con->write, 0) ||
-		    init_section_contains(con->read, 0) ||
-		    init_section_contains(con->device, 0) ||
-		    init_section_contains(con->unblank, 0) ||
-		    init_section_contains(con->data, 0)) {
+		/*
+		 * Keep the boot console when requested or as a fallback
+		 * unless it is using an init section.
+		 */
+		if (keep_bootcon || !have_real_console) {
+			if (!init_section_intersects(con, sizeof(*con)) &&
+			    !init_section_contains(con->write, 0) &&
+			    !init_section_contains(con->read, 0) &&
+			    !init_section_contains(con->device, 0) &&
+			    !init_section_contains(con->unblank, 0) &&
+			    !init_section_contains(con->data, 0))
+				continue;
+
 			/*
 			 * Please, consider moving the reported consoles out
 			 * of the init section.
 			 */
 			pr_warn("bootconsole [%s%d] uses init memory and must be disabled even before the real one is ready\n",
 				con->name, con->index);
-			unregister_console_locked(con);
 		}
+
+		unregister_console_locked(con);
 	}
 	console_list_unlock();
 

Best Regards,
Petr

  reply	other threads:[~2026-08-20 15:13 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-05  8:48 Xiaochun Li
2026-08-20 15:13 ` Petr Mladek [this message]
2026-08-21  2:10   ` Xiaochun Li

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=aocZgQNT4abxmOmK@pathway.suse.cz \
    --to=pmladek@suse.com \
    --cc=john.ogness@linutronix.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lixiaochun@open-hieco.net \
    --cc=rostedt@goodmis.org \
    --cc=senozhatsky@chromium.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®