mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Petr Mladek <pmladek@suse.com>
To: Chris Down <chris@chrisdown.name>
Cc: John Ogness <john.ogness@linutronix.de>,
	Sergey Senozhatsky <senozhatsky@chromium.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	Marcos Paulo de Souza <mpdesouza@suse.com>,
	Naveen Kumar Chaudhary <naveen.osdev@gmail.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v5 14/16] printk: Do not set Braille console as preferred_console
Date: Fri, 6 Nov 2026 15:21:19 +0100	[thread overview]
Message-ID: <au3iX_xMl9H-MqJB@pathway.suse.cz> (raw)
In-Reply-To: <auLznwtHktgqlST1@chrisdown.name>

On Thu 2026-10-29 15:19:59, Chris Down wrote:
> Petr Mladek writes:
> >-	if (brl_options)
> >+	if (is_new_entry)
> >+		INIT_LIST_HEAD(&pc->preferred_dev_node);
> 
> is_new_entry is derived from pc->name[0]/pc->devname[0] being empty, and
> console_setup() can store an empty name:
> 
>     console=,115200 console=tty1
> 
> The first one creates an entry with name "" (and links it). The second one 
> stops the lookup loop at the same slot, sees it as new again, and 
> re-initialises a node which is still linked, followed by list_add(). With 
> DEBUG_LIST + BUG_ON_DATA_CORRUPTION:
> 
>     list_add corruption. next->prev should be prev (ffffb010209fc0c0), but was ffffb01023cd1fa0. (next=ffffb01023cd1fa0).
>     kernel BUG at lib/list_debug.c:29!
>     Call trace:
>      __list_add_valid_or_report+0xbc/0x1c0 (P)
>      add_or_update_preferred_console+0x6fc/0x9c0
>      __add_preferred_console+0x11c/0x160
>      console_setup+0x2e4/0x348
>      obsolete_checksetup+0x148/0x190
>      unknown_bootoption+0x150/0x328
>      parse_one+0x108/0x480
>      parse_args+0x158/0x3d0
>      start_kernel+0x134/0x488
> 
> "console=,foo:bar" does the same on the first console= when SPCR added an 
> entry, since it matches the SPCR slot via the empty devname (see 7/16).
> 
> Rejecting empty name/devname in add_or_update_preferred_console() (see my 
> comment on 7/16) should fix that.

Yup, it will be fixed in v6 this way.

> >-		preferred_dev_console = i;
> >+		if (!list_empty(&pc->preferred_dev_node))
> >+			list_del_init(&pc->preferred_dev_node);
> >+	} else if (!is_braille_console_preferred(pc)) {
> >+		if (list_empty(&pc->preferred_dev_node))
> >+			list_add(&pc->preferred_dev_node, &preferred_dev_console);
> >+		else
> >+			list_move(&pc->preferred_dev_node, &preferred_dev_console);
> 
> list_del_init() and list_move() are both fine on an initialised self-linked 
> node (list_move() passes the DEBUG_LIST checks in __list_del_entry_valid()), so 
> the list_empty() checks can go.

I was lazy to check this and wanted to stay on the safe side ;-)
OK, I am going to do the following changes for v6:

--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -2666,13 +2666,9 @@ static int add_or_update_preferred_console(unsigned int i,
 	 */
 	if (brl_options) {
 		want_braille_console = true;
-		if (!list_empty(&pc->preferred_dev_node))
-			list_del_init(&pc->preferred_dev_node);
+		list_del_init(&pc->preferred_dev_node);
 	} else if (!is_braille_console_preferred(pc)) {
-		if (list_empty(&pc->preferred_dev_node))
-			list_add(&pc->preferred_dev_node, &preferred_dev_console);
-		else
-			list_move(&pc->preferred_dev_node, &preferred_dev_console);
+		list_move(&pc->preferred_dev_node, &preferred_dev_console);
 	}
 
 	if (user_specified) {


> With these fixed:
> 
> Acked-by: Chris Down <chris@chrisdown.name>

Thanks.

Best Regards,
Petr

  reply	other threads:[~2026-11-06 14:21 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-29 12:05 [PATCH v5 00/16] printk: Clean up preferred console handling Petr Mladek
2026-09-29 12:05 ` [PATCH v5 01/16] printk: Rename struct console_cmdline to preferred_console Petr Mladek
2026-10-29  7:03   ` Chris Down
2026-11-03 15:16     ` Petr Mladek
2026-09-29 12:05 ` [PATCH v5 02/16] printk: Rename preferred_console to preferred_dev_console Petr Mladek
2026-09-29 12:05 ` [PATCH v5 03/16] printk: Cleanup _braille_(un)register_console() wrappers Petr Mladek
2026-09-29 12:05 ` [PATCH v5 04/16] printk: Separate code for adding/updating preferred console metadata Petr Mladek
2026-10-27 12:05   ` John Ogness
2026-11-03 15:57     ` Petr Mladek
2026-10-29  7:05   ` Chris Down
2026-11-03 16:12     ` Petr Mladek
2026-09-29 12:05 ` [PATCH v5 05/16] printk: Prefer the console either via name or devname exclusively Petr Mladek
2026-10-27 12:14   ` John Ogness
2026-10-29  7:06   ` Chris Down
2026-09-29 12:05 ` [PATCH v5 06/16] printk: Check that index is not defined when a console is preferred via devname Petr Mladek
2026-10-27 18:08   ` John Ogness
2026-10-29  7:07   ` Chris Down
2026-09-29 12:05 ` [PATCH v5 07/16] printk: Ensure consistent name and devname when updating preferred console Petr Mladek
2026-10-29  7:09   ` Chris Down
2026-11-03 17:08     ` Petr Mladek
2026-09-29 12:05 ` [PATCH v5 08/16] printk: Do not quietly ignore adding entry for a preferred console when there is no space Petr Mladek
2026-10-29  7:10   ` Chris Down
2026-10-29 16:23   ` John Ogness
2026-09-29 12:05 ` [PATCH v5 09/16] printk: Allow to override options via the command line Petr Mladek
2026-10-29  7:13   ` Chris Down
2026-10-30 17:04     ` John Ogness
2026-11-04 12:55     ` Petr Mladek
2026-09-29 12:05 ` [PATCH v5 10/16] printk: Allow to debug misuse of add_or_update_preferred_console() Petr Mladek
2026-10-29  7:13   ` Chris Down
2026-09-29 12:05 ` [PATCH v5 11/16] console/braille: Lock console->setup() call during the registration Petr Mladek
2026-10-29  7:15   ` Chris Down
2026-11-04 14:21     ` Petr Mladek
2026-09-29 12:05 ` [PATCH v5 12/16] printk: Separate code for enabling console Petr Mladek
2026-10-29  7:15   ` Chris Down
2026-09-29 12:05 ` [PATCH v5 13/16] printk: Try to register each console as Braille first Petr Mladek
2026-10-29  7:17   ` Chris Down
2026-11-05 17:13     ` Petr Mladek
2026-11-06 11:33   ` Petr Mladek
2026-09-29 12:05 ` [PATCH v5 14/16] printk: Do not set Braille console as preferred_console Petr Mladek
2026-10-29  7:19   ` Chris Down
2026-11-06 14:21     ` Petr Mladek [this message]
2026-09-29 12:05 ` [PATCH v5 15/16] printk: Modify try_enable_default_console() to return error/success Petr Mladek
2026-10-29  7:20   ` Chris Down
2026-09-29 12:05 ` [PATCH v5 16/16] printk: Try enable preferred consoles only when there are any Petr Mladek
2026-10-29  7:20   ` Chris Down

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=au3iX_xMl9H-MqJB@pathway.suse.cz \
    --to=pmladek@suse.com \
    --cc=chris@chrisdown.name \
    --cc=john.ogness@linutronix.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mpdesouza@suse.com \
    --cc=naveen.osdev@gmail.com \
    --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®