linux-kernel - Re: [PATCH 2/2] kdb: Adapt kdb_msg_write to work with NBCON consoles

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: <84ldorx1z2.fsf@jogness.linutronix.de>
Date: Sun, 13 Jul 2025 22:42:17 +0206
From: John Ogness <john.ogness@...utronix.de>
To: Marcos Paulo de Souza <mpdesouza@...e.com>, Greg Kroah-Hartman
 <gregkh@...uxfoundation.org>, Petr Mladek <pmladek@...e.com>, Steven
 Rostedt <rostedt@...dmis.org>, Sergey Senozhatsky
 <senozhatsky@...omium.org>, Jason Wessel <jason.wessel@...driver.com>,
 Daniel Thompson <danielt@...nel.org>, Douglas Anderson
 <dianders@...omium.org>
Cc: linux-kernel@...r.kernel.org, kgdb-bugreport@...ts.sourceforge.net,
 Marcos Paulo de Souza <mpdesouza@...e.com>
Subject: Re: [PATCH 2/2] kdb: Adapt kdb_msg_write to work with NBCON consoles

On 2025-07-13, Marcos Paulo de Souza <mpdesouza@...e.com> wrote:
> diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c
> index 9b11b10b120cf07e451a7a4d92ce50f9a6c066b2..3b7365c11d06b01d487767fd89f1081da10dd2ed 100644
> --- a/kernel/debug/kdb/kdb_io.c
> +++ b/kernel/debug/kdb/kdb_io.c
> @@ -558,6 +558,25 @@ static int kdb_search_string(char *searched, char *searchfor)
>  	return 0;
>  }
>  
> +static struct nbcon_context *nbcon_acquire_ctxt(struct console *con,
> +					struct nbcon_write_context *wctxt,
> +					char *msg, int msg_len)
> +{
> +	struct nbcon_context *ctxt = &ACCESS_PRIVATE(wctxt, ctxt);
> +
> +	ctxt->console		    = con;
> +	ctxt->spinwait_max_us	    = 0;
> +	ctxt->prio		    = NBCON_PRIO_EMERGENCY;
> +	ctxt->allow_unsafe_takeover = false;
> +	wctxt->outbuf		    = msg;
> +	wctxt->len		    = msg_len;
> +
> +	if (!nbcon_context_try_acquire(ctxt))
> +		return NULL;
> +
> +	return ctxt;

This function is grabbing a reference to a private member and returning
it, thus exposing internals. Can we instead create a proper API in
kernel/printk/nbcon.c for kdb?

For example, take a look at:

nbcon_device_try_acquire() and nbcon_device_release()

We could have something similar for kdb, such as:

bool *nbcon_kdb_try_acquire(struct nbcon_write_context *wctxt,
			    struct console *con, char *msg, int msg_len);

void nbcon_kdb_release(struct nbcon_write_context *wctxt);

> +}
> +
>  static void kdb_msg_write(const char *msg, int msg_len)
>  {
>  	struct console *c;
> @@ -589,12 +608,26 @@ static void kdb_msg_write(const char *msg, int msg_len)
>  	 */
>  	cookie = console_srcu_read_lock();
>  	for_each_console_srcu(c) {
> -		if (!(console_srcu_read_flags(c) & CON_ENABLED))
> +		struct nbcon_write_context wctxt = { };
> +		struct nbcon_context *ctxt;

With the above suggestion we do not need @ctxt.

> +		short flags = console_srcu_read_flags(c);
> +
> +		if (!console_is_usable(c, flags, true))
>  			continue;
>  		if (c == dbg_io_ops->cons)
>  			continue;
> -		if (!c->write)
> -			continue;
> +
> +		/*
> +		 * Do not continue if the console is NBCON and the context
> +		 * can't be acquired.
> +		 */
> +		if (flags & CON_NBCON) {
> +			ctxt = nbcon_acquire_ctxt(c, &wctxt, (char *)msg,
> +						  msg_len);
> +			if (!ctxt)
> +				continue;

And this becomes:

			if (!nbcon_kdb_try_acquire(&wctxt, c, (char *)msg, msg_len))
				continue;
> +		}
> +
>  		/*
>  		 * Set oops_in_progress to encourage the console drivers to
>  		 * disregard their internal spin locks: in the current calling
> @@ -605,7 +638,12 @@ static void kdb_msg_write(const char *msg, int msg_len)
>  		 * for this calling context.
>  		 */
>  		++oops_in_progress;
> -		c->write(c, msg, msg_len);
> +		if (flags & CON_NBCON) {
> +			c->write_atomic(c, &wctxt);
> +			nbcon_context_release(ctxt);

And this becomes:

			nbcon_kdb_release(&wctxt);

> +		} else {
> +			c->write(c, msg, msg_len);
> +		}
>  		--oops_in_progress;
>  		touch_nmi_watchdog();
>  	}

John Ogness