Skip to contentSkip to conversation

Fixes for console command line ordering

PATCH In review Tony Lindgren 3 participants 14 messages Jun 20, 2024 – Jun 25, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

v4 cover Fixes for console command line ordering

Hi,

Recent changes to add support for DEVNAME:0.0 style consoles caused a regression with the preferred console order where the last console on the kernel command line is no longer the preferred console.

The following four changes fix the issue using Petr's suggestion that does not involve calling __add_preferred_console() later on again, and adds the deferred consoles to the console_cmdline[] directly to be updated when the console is ready.

We revert the earlier printk related changes, and then add back the DEVNAME:0.0 functionality based on Petr's code snippet. And we end up reducing the code quite a bit too this way.

And we also revert all the unusable serial core console quirk handling, it does not do anything for the legacy "ttyS" named consoles. And then we add a minimal serial_base_match_and_update_preferred_console().

The reason we want DEVNAME:0.0 style consoles is it helps addressing the console based on the connected serial port controller device rather than using the hardcoded ttyS addressing. And that helps with issues related to the console moving around after togging the HSUART option in the BIOS, or when new ports are enabled in devicetree and aliases are not updated.

Regards,

Tony

Changes since v3:

- Revert the unusable serial core console quirk handling

- Add a minimal patch for serial_base_match_and_update_preferred_console()

Changes since v2:

- Use match_devname_and_update_preferred_console() naming and update the comments

- Add a patch to rename the serial functions to use match and update naming

- Use ttyname instad of chardev in console_setup()

- Split variables per-line in console_setup()

- Initialize idx to -1 for devname in console_setup()

- Add pr_info() statement when a preferred console is associated with a devname

Changes since v1:

- Revert the problem causing printk changes and switch to using the solution based on Petr's suggestion and code snippet

Tony Lindgren (4):
printk: Revert add_preferred_console_match() related commits
printk: Add match_devname_and_update_preferred_console()
serial: core: Revert unusable console quirk handling
serial: core: Add serial_base_match_and_update_preferred_console()

 drivers/tty/serial/8250/8250_core.c  |   5 -
 drivers/tty/serial/serial_base.h     |  22 +---
 drivers/tty/serial/serial_base_bus.c | 116 +++------------------
 drivers/tty/serial/serial_core.c     |   2 +-
 include/linux/printk.h               |   5 +-
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   7 +-
 kernel/printk/printk.c               | 122 ++++++++++++++++------
 9 files changed, 112 insertions(+), 315 deletions(-)
 delete mode 100644 kernel/printk/conopt.c

base-commit: 6ba59ff4227927d3a8530fc2973b80e94b54d58f

Signature · 1 line
2.45.2
#1 lore raw

PATCH v4 1/4 printk: Revert add_preferred_console_match() related commits

Recent changes to allow using DEVNAME:0.0 style console names caused a regression to the kernel command line handling for the console options.

The last preferred console added gets used for init. This is documented in the comments for add_preferred_console(). Now the kernel command line options for console=ttyS0,115200 console=tty0 are wrongly handled and cause the /dev/console to be associated with ttyS0 instead of tty0.

This happens because we are calling __add_preferred_console() later on from serial8250_isa_init_ports() after console_setup() and the console gets treated as the last added preferred console. As the DEVNAME:0.0 style console device is not known at console_setup() time, I added a call to __add_preferred_console() later on when the console is ready.

To fix the issue, let's revert the printk related commits:

f03e8c1060f8 ("printk: Save console options for add_preferred_console_match()") b73c9cbe4f1f ("printk: Flag register_console() if console is set on command line") 8a831c584e6e ("printk: Don't try to parse DEVNAME:0.0 console options")

We need to also drop the call for add_preferred_console_match() from serial_base_add_one_prefcon() added by commit 787a1cabac01 ("serial: core:
Add support for DEVNAME:0.0 style naming for kernel console").

Petr has suggested a better way to handle the deferred consoles that does not rely on calling __add_preferred_console() again.

Fixes f03e8c1060f8 ("printk: Save console options for add_preferred_console_match()")
---
 drivers/tty/serial/serial_base_bus.c |   8 +-
 include/linux/printk.h               |   3 -
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   6 --
 kernel/printk/printk.c               |  23 +----
 6 files changed, 6 insertions(+), 182 deletions(-)
 delete mode 100644 kernel/printk/conopt.c
drivers/tty/serial/serial_base_bus.c +1 -7
index 73c6ee540c83..5ebacb982f9e 100644
@@ -210,13 +210,7 @@ void serial_base_port_device_remove(struct serial_port_device *port_dev)
 static int serial_base_add_one_prefcon(const char *match, const char *dev_name,
 				       int port_id)
 {
-	int ret;
-
-	ret = add_preferred_console_match(match, dev_name, port_id);
-	if (ret == -ENOENT)
-		return 0;
-
-	return ret;
+	return 0;
 }
 
 #ifdef __sparc__
include/linux/printk.h +0 -3
index 40afab23881a..65c5184470f1 100644
@@ -60,9 +60,6 @@ static inline const char *printk_skip_headers(const char *buffer)
 #define CONSOLE_LOGLEVEL_DEFAULT CONFIG_CONSOLE_LOGLEVEL_DEFAULT
 #define CONSOLE_LOGLEVEL_QUIET	 CONFIG_CONSOLE_LOGLEVEL_QUIET
 
-int add_preferred_console_match(const char *match, const char *name,
-				const short idx);
-
 extern int console_printk[];
 
 #define console_loglevel (console_printk[0])
kernel/printk/Makefile +1 -1
index 040fe7d1eda2..39a2b61c7232 100644
@@ -1,5 +1,5 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-y	= printk.o conopt.o
+obj-y	= printk.o
 obj-$(CONFIG_PRINTK)	+= printk_safe.o nbcon.o
 obj-$(CONFIG_A11Y_BRAILLE_CONSOLE)	+= braille.o
 obj-$(CONFIG_PRINTK_INDEX)	+= index.o
kernel/printk/conopt.c deleted +0 -146

146 lines not on this page · raw message

kernel/printk/console_cmdline.h +0 -6
index a125e0235589..3ca74ad391d6 100644
@@ -2,12 +2,6 @@
 #ifndef _CONSOLE_CMDLINE_H
 #define _CONSOLE_CMDLINE_H
 
-#define MAX_CMDLINECONSOLES 8
-
-int console_opt_save(const char *str, const char *brl_opt);
-int console_opt_add_preferred_console(const char *name, const short idx,
-				      char *options, char *brl_options);
-
 struct console_cmdline
 {
 	char	name[16];			/* Name of the driver	    */
kernel/printk/printk.c +4 -19
index 420fd310129d..dddb15f48d59 100644
@@ -383,6 +383,9 @@ static int console_locked;
 /*
  *	Array of consoles built from command line options (console=)
  */
+
+#define MAX_CMDLINECONSOLES 8
+
 static struct console_cmdline console_cmdline[MAX_CMDLINECONSOLES];
 
 static int preferred_console = -1;
@@ -2500,17 +2503,6 @@ static int __init console_setup(char *str)
 	if (_braille_console_setup(&str, &brl_options))
 		return 1;
 
-	/* Save the console for driver subsystem use */
-	if (console_opt_save(str, brl_options))
-		return 1;
-
-	/* Flag register_console() to not call try_enable_default_console() */
-	console_set_on_cmdline = 1;
-
-	/* Don't attempt to parse a DEVNAME:0.0 style console */
-	if (strchr(str, ':'))
-		return 1;
-
 	/*
 	 * Decode str into name, index, options.
 	 */
@@ -2541,13 +2533,6 @@ static int __init console_setup(char *str)
 }
 __setup("console=", console_setup);
 
-/* Only called from add_preferred_console_match() */
-int console_opt_add_preferred_console(const char *name, const short idx,
-				      char *options, char *brl_options)
-{
-	return __add_preferred_console(name, idx, options, brl_options, true);
-}
-
 /**
  * add_preferred_console - add a device to the list of preferred consoles.
  * @name: device name
@@ -3522,7 +3507,7 @@ void register_console(struct console *newcon)
 	 * Note that a console with tty binding will have CON_CONSDEV
 	 * flag set and will be first in the list.
 	 */
-	if (preferred_console < 0 && !console_set_on_cmdline) {
+	if (preferred_console < 0) {
 		if (hlist_empty(&console_list) || !console_first()->device ||
 		    console_first()->flags & CON_BOOT) {
 			try_enable_default_console(newcon);
Signature · 1 line
2.45.2
#2 lore raw

PATCH v4 2/4 printk: Add match_devname_and_update_preferred_console()

In reply to Tony Lindgren

Let's add match_devname_and_update_preferred_console() for driver subsystems to call during init when the console is ready, and it's character device name is known. For now, we use it only for the serial layer to allow console=DEVNAME:0.0 style hardware based addressing for consoles.

The earlier attempt on doing this caused a regression with the kernel command line console order as it added calling __add_preferred_console() again later on during init. A better approach was suggested by Petr where we add the deferred console to the console_cmdline[] and update it later on when the console is ready.

---
 drivers/tty/serial/serial_base_bus.c |   8 ++-
 include/linux/printk.h               |   4 ++
 kernel/printk/console_cmdline.h      |   1 +
 kernel/printk/printk.c               | 103 +++++++++++++++++++++++----
 4 files changed, 100 insertions(+), 16 deletions(-)
drivers/tty/serial/serial_base_bus.c +7 -1
index 5ebacb982f9e..2cf86f1ff298 100644
@@ -210,7 +210,13 @@ void serial_base_port_device_remove(struct serial_port_device *port_dev)
 static int serial_base_add_one_prefcon(const char *match, const char *dev_name,
 				       int port_id)
 {
-	return 0;
+	int ret;
+
+	ret = match_devname_and_update_preferred_console(match, dev_name, port_id);
+	if (ret == -ENOENT)
+		return 0;
+
+	return ret;
 }
 
 #ifdef __sparc__
include/linux/printk.h +4 -0
index 65c5184470f1..7239976698e4 100644
@@ -60,6 +60,10 @@ static inline const char *printk_skip_headers(const char *buffer)
 #define CONSOLE_LOGLEVEL_DEFAULT CONFIG_CONSOLE_LOGLEVEL_DEFAULT
 #define CONSOLE_LOGLEVEL_QUIET	 CONFIG_CONSOLE_LOGLEVEL_QUIET
 
+int match_devname_and_update_preferred_console(const char *match,
+					       const char *name,
+					       const short idx);
+
 extern int console_printk[];
 
 #define console_loglevel (console_printk[0])
kernel/printk/console_cmdline.h +1 -0
index 3ca74ad391d6..0ab573b6d4dc 100644
@@ -6,6 +6,7 @@ struct console_cmdline
 {
 	char	name[16];			/* Name of the driver	    */
 	int	index;				/* Minor dev. to use	    */
+	char	devname[32];			/* DEVNAME:0.0 style device name */
 	bool	user_specified;			/* Specified by command line vs. platform */
 	char	*options;			/* Options for the driver   */
 #ifdef CONFIG_A11Y_BRAILLE_CONSOLE
kernel/printk/printk.c +88 -15

174 lines not on this page · raw message

Signature · 1 line
2.45.2
#3 lore raw

PATCH v4 3/4 serial: core: Revert unusable console quirk handling

In reply to Tony Lindgren

There is no point of trying to handle the console quirks in the serial core instead of console_setup() currently.

With the console_setup() related changes, we are now deferring the DEVNAME:0.0 style consoles based on the ":" naming only. So the serial core console quirk handling would not do anything for the "ttyS" named consoles as they are not deferred.

Also the earlier approach would have depended on further changes to be able to drop the serial port quirk handling from console_setup().

Let's revert the following console quirk handling related serial core commits:

b20172ca6bf4 ("serial: core: Fix ifdef for serial base console functions") 4547cd76f08a ("serial: 8250: Fix add preferred console for serial8250_isa_init_ports()") a8b04cfe7dad ("serial: 8250: Add preferred console in serial8250_isa_init_ports()") a0f32e2dd998 ("serial: core: Handle serial console options")
787a1cabac01 ("serial: core: Add support for DEVNAME:0.0 style naming for kernel console")

Once the console quirk handling is gone, we add back the DEVNAME:0.0 functionality with a minimal patch.

---
 drivers/tty/serial/8250/8250_core.c  |   5 --
 drivers/tty/serial/serial_base.h     |  30 -------
 drivers/tty/serial/serial_base_bus.c | 129 ---------------------------
 drivers/tty/serial/serial_core.c     |   4 -
 4 files changed, 168 deletions(-)
drivers/tty/serial/8250/8250_core.c +0 -5
index ff15022369e4..b0adafc44747 100644
@@ -15,7 +15,6 @@
  */
 
 #include <linux/acpi.h>
-#include <linux/cleanup.h>
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/ioport.h>
@@ -42,8 +41,6 @@
 
 #include <asm/irq.h>
 
-#include "../serial_base.h"	/* For serial_base_add_isa_preferred_console() */
-
 #include "8250.h"
 
 /*
@@ -563,8 +560,6 @@ static void __init serial8250_isa_init_ports(void)
 		port->irqflags |= irqflag;
 		if (serial8250_isa_config != NULL)
 			serial8250_isa_config(i, &up->port, &up->capabilities);
-
-		serial_base_add_isa_preferred_console(serial8250_reg.dev_name, i);
 	}
 }
 
drivers/tty/serial/serial_base.h +0 -30
index 743a72ac34f3..b6c38d2edfd4 100644
@@ -49,33 +49,3 @@ void serial_ctrl_unregister_port(struct uart_driver *drv, struct uart_port *port
 
 int serial_core_register_port(struct uart_driver *drv, struct uart_port *port);
 void serial_core_unregister_port(struct uart_driver *drv, struct uart_port *port);
-
-#ifdef CONFIG_SERIAL_CORE_CONSOLE
-
-int serial_base_add_preferred_console(struct uart_driver *drv,
-				      struct uart_port *port);
-
-#else
-
-static inline
-int serial_base_add_preferred_console(struct uart_driver *drv,
-				      struct uart_port *port)
-{
-	return 0;
-}
-
-#endif
-
-#ifdef CONFIG_SERIAL_8250_CONSOLE
-
-int serial_base_add_isa_preferred_console(const char *name, int idx);
-
-#else
-
-static inline
-int serial_base_add_isa_preferred_console(const char *name, int idx)
-{
-	return 0;
-}
-
-#endif
drivers/tty/serial/serial_base_bus.c +0 -129

141 lines not on this page · raw message

drivers/tty/serial/serial_core.c +0 -4
index 0c4d60976663..2a8006e3d687 100644
@@ -3422,10 +3422,6 @@ int serial_core_register_port(struct uart_driver *drv, struct uart_port *port)
 	if (ret)
 		goto err_unregister_ctrl_dev;
 
-	ret = serial_base_add_preferred_console(drv, port);
-	if (ret)
-		goto err_unregister_port_dev;
-
 	ret = serial_core_add_one_port(drv, port);
 	if (ret)
 		goto err_unregister_port_dev;
Signature · 1 line
2.45.2
maintainer Reviewed-by on v4 3/4
#4 lore raw

Re: [PATCH v4 3/4] serial: core: Revert unusable console quirk handling

Quoted 20 lines · Tony Lindgren

There is no point of trying to handle the console quirks in the serial core instead of console_setup() currently.

With the console_setup() related changes, we are now deferring the DEVNAME:0.0 style consoles based on the ":" naming only. So the serial core console quirk handling would not do anything for the "ttyS" named consoles as they are not deferred.

Also the earlier approach would have depended on further changes to be able to drop the serial port quirk handling from console_setup().

Let's revert the following console quirk handling related serial core commits:

b20172ca6bf4 ("serial: core: Fix ifdef for serial base console functions") 4547cd76f08a ("serial: 8250: Fix add preferred console for serial8250_isa_init_ports()") a8b04cfe7dad ("serial: 8250: Add preferred console in serial8250_isa_init_ports()") a0f32e2dd998 ("serial: core: Handle serial console options")
787a1cabac01 ("serial: core: Add support for DEVNAME:0.0 style naming for kernel console")

Quoted 6 lines

Once the console quirk handling is gone, we add back the DEVNAME:0.0 functionality with a minimal patch.

Best Regards,
Petr

#5 lore raw

PATCH v4 4/4 serial: core: Add serial_base_match_and_update_preferred_console()

In reply to Tony Lindgren

Let's add serial_base_match_and_update_preferred_console() for consoles using DEVNAME:0.0 style naming.

The earlier approach to add it caused issues in the kernel command line ordering as we were calling __add_preferred_console() again for the deferred consoles.

---
 drivers/tty/serial/serial_base.h     | 16 ++++++++++++
 drivers/tty/serial/serial_base_bus.c | 37 ++++++++++++++++++++++++++++
 drivers/tty/serial/serial_core.c     |  4 +++
 3 files changed, 57 insertions(+)
drivers/tty/serial/serial_base.h +16 -0
index b6c38d2edfd4..0d50db5b660b 100644
@@ -49,3 +49,19 @@ void serial_ctrl_unregister_port(struct uart_driver *drv, struct uart_port *port
 
 int serial_core_register_port(struct uart_driver *drv, struct uart_port *port);
 void serial_core_unregister_port(struct uart_driver *drv, struct uart_port *port);
+
+#ifdef CONFIG_SERIAL_CORE_CONSOLE
+
+int serial_base_match_and_update_preferred_console(struct uart_driver *drv,
+						   struct uart_port *port);
+
+#else
+
+static inline
+int serial_base_match_and_update_preferred_console(struct uart_driver *drv,
+						   struct uart_port *port)
+{
+	return 0;
+}
+
+#endif
drivers/tty/serial/serial_base_bus.c +37 -0
index 4df2a4b10445..d822499ba9d6 100644
@@ -8,6 +8,7 @@
  * The serial core bus manages the serial core controller instances.
  */
 
+#include <linux/cleanup.h>
 #include <linux/container_of.h>
 #include <linux/device.h>
 #include <linux/idr.h>
@@ -204,6 +205,42 @@ void serial_base_port_device_remove(struct serial_port_device *port_dev)
 	put_device(&port_dev->dev);
 }
 
+#ifdef CONFIG_SERIAL_CORE_CONSOLE
+
+/**
+ * serial_base_match_and_update_preferred_console - Match and update a preferred console
+ * @drv: Serial port device driver
+ * @port: Serial port instance
+ *
+ * Tries to match and update the preferred console for a serial port for
+ * the kernel command line option console=DEVNAME:0.0.
+ *
+ * Cannot be called early for ISA ports, depends on struct device.
+ *
+ * Return: 0 on success, negative error code on failure.
+ */
+int serial_base_match_and_update_preferred_console(struct uart_driver *drv,
+						   struct uart_port *port)
+{
+	const char *port_match __free(kfree) = NULL;
+	int ret;
+
+	port_match = kasprintf(GFP_KERNEL, "%s:%d.%d", dev_name(port->dev),
+			       port->ctrl_id, port->port_id);
+	if (!port_match)
+		return -ENOMEM;
+
+	ret = match_devname_and_update_preferred_console(port_match,
+							 drv->dev_name,
+							 port->line);
+	if (ret == -ENOENT)
+		return 0;
+
+	return ret;
+}
+
+#endif
+
 static int serial_base_init(void)
 {
 	int ret;
drivers/tty/serial/serial_core.c +4 -0
index 2a8006e3d687..9a18d0b95a41 100644
@@ -3422,6 +3422,10 @@ int serial_core_register_port(struct uart_driver *drv, struct uart_port *port)
 	if (ret)
 		goto err_unregister_ctrl_dev;
 
+	ret = serial_base_match_and_update_preferred_console(drv, port);
+	if (ret)
+		goto err_unregister_port_dev;
+
 	ret = serial_core_add_one_port(drv, port);
 	if (ret)
 		goto err_unregister_port_dev;
Signature · 1 line
2.45.2
maintainer Reviewed-by Tested-by on v4 4/4
#6 lore raw

Re: [PATCH v4 4/4] serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 8 lines · Tony Lindgren

Let's add serial_base_match_and_update_preferred_console() for consoles using DEVNAME:0.0 style naming.

The earlier approach to add it caused issues in the kernel command line ordering as we were calling __add_preferred_console() again for the deferred consoles.

Looks good and seems to work well:

See an idea below.

Quoted 20 lines
drivers/tty/serial/serial_base_bus.c +14 -0
@@ -204,6 +205,42 @@ void serial_base_port_device_remove(struct serial_port_device *port_dev)
 	put_device(&port_dev->dev);
 }
 
+#ifdef CONFIG_SERIAL_CORE_CONSOLE
+
+/**
+ * serial_base_match_and_update_preferred_console - Match and update a preferred console
+ * @drv: Serial port device driver
+ * @port: Serial port instance
+ *
+ * Tries to match and update the preferred console for a serial port for
+ * the kernel command line option console=DEVNAME:0.0.
+ *
+ * Cannot be called early for ISA ports, depends on struct device.
+ *
+ * Return: 0 on success, negative error code on failure.
+ */

drivers/tty/serial/serial_base_bus.c void serial_base_port_device_remove(struct serial_port_device *port_dev) · line 231

Quoted 10 lines
drivers/tty/serial/serial_base_bus.c +10 -0
+int serial_base_match_and_update_preferred_console(struct uart_driver *drv,
+						   struct uart_port *port)
+{
+	const char *port_match __free(kfree) = NULL;
+	int ret;
+
+	port_match = kasprintf(GFP_KERNEL, "%s:%d.%d", dev_name(port->dev),
+			       port->ctrl_id, port->port_id);
+	if (!port_match)
+		return -ENOMEM;

The name is going to be compared with:

struct console_cmdline
{
[...]

	char	devname[32];			/* DEVNAME:0.0 style device name */

It looks like an overkill to allocate such a small buffer. It would be perfectly fine to use a buffer on stack.

Well, we would need to define somewhere (likely in include/linux/console.h):

#define CONSOLE_DEVNAME_LEN 32

and then do

	char port_match[CONSOLE_DEVNAME_LEN];
	int len;

	len = snprintf(port_match, ARRAY_SIZE(port_match), "%s:%d.%d",
		       dev_name(port->dev), port->ctrl_id, port->port_id);
	if (len >= ARRAY_SIZE(port_match)) {
		pr_warn("Console devname does not fit into the buffer: "%s:%d.%d\n",
			 dev_name(port->dev), port->ctrl_id, port->port_id);
		return -ENOMEM;
	}

The advantage is that it would warn when there are longer device names. It would help to catch situations when CONSOLE_DEVNAME_LEN is not big enough.

It might be done in a separate patch.

Quoted 5 lines
drivers/tty/serial/serial_base_bus.c +5 -0
+
+	ret = match_devname_and_update_preferred_console(port_match,
+							 drv->dev_name,
+							 port->line);
+	if (ret == -ENOENT)

drivers/tty/serial/serial_base_bus.c void serial_base_port_device_remove(struct serial_port_device *port_dev) · line 246

Quoted 10 lines
drivers/tty/serial/serial_base_bus.c +7 -0
+		return 0;
+
+	return ret;
+}
+
+#endif
+
 static int serial_base_init(void)
 {
 	int ret;

Best Regards,
Petr

#7 lore raw

Re: [PATCH v4 4/4] serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 8 lines · Petr Mladek
Quoted 7 lines · Tony Lindgren

Let's add serial_base_match_and_update_preferred_console() for consoles using DEVNAME:0.0 style naming.

The earlier approach to add it caused issues in the kernel command line ordering as we were calling __add_preferred_console() again for the deferred consoles.

Quoted 6 lines
Quoted 1 line

Looks good and seems to work well:

OK thanks for testing again.

Quoted 4 lines

See an idea below.

Quoted 2 lines
drivers/tty/serial/serial_base_bus.c +0 -0

...

Quoted 38 lines

drivers/tty/serial/serial_base_bus.c void serial_base_port_device_remove(struct serial_port_device *port_dev) · line 231

Quoted 10 lines
drivers/tty/serial/serial_base_bus.c +10 -0
+int serial_base_match_and_update_preferred_console(struct uart_driver *drv,
+						   struct uart_port *port)
+{
+	const char *port_match __free(kfree) = NULL;
+	int ret;
+
+	port_match = kasprintf(GFP_KERNEL, "%s:%d.%d", dev_name(port->dev),
+			       port->ctrl_id, port->port_id);
+	if (!port_match)
+		return -ENOMEM;

The name is going to be compared with:

struct console_cmdline
{
[...]

	char	devname[32];			/* DEVNAME:0.0 style device name */

It looks like an overkill to allocate such a small buffer. It would be perfectly fine to use a buffer on stack.

Well, we would need to define somewhere (likely in include/linux/console.h):

#define CONSOLE_DEVNAME_LEN 32

and then do

	char port_match[CONSOLE_DEVNAME_LEN];
	int len;

	len = snprintf(port_match, ARRAY_SIZE(port_match), "%s:%d.%d",
		       dev_name(port->dev), port->ctrl_id, port->port_id);
	if (len >= ARRAY_SIZE(port_match)) {
		pr_warn("Console devname does not fit into the buffer: "%s:%d.%d\n",
			 dev_name(port->dev), port->ctrl_id, port->port_id);
		return -ENOMEM;
	}
Quoted 2 lines

The advantage is that it would warn when there are longer device names. It would help to catch situations when CONSOLE_DEVNAME_LEN is not big enough.

Good idea.

Quoted 1 line

It might be done in a separate patch.

Sounds good to me.

Regards,

Tony

maintainer
#8 lore raw
In reply to Tony Lindgren

Added Linus into Cc.

Quoted 31 lines · Tony Lindgren

Hi,

Recent changes to add support for DEVNAME:0.0 style consoles caused a regression with the preferred console order where the last console on the kernel command line is no longer the preferred console.

The following four changes fix the issue using Petr's suggestion that does not involve calling __add_preferred_console() later on again, and adds the deferred consoles to the console_cmdline[] directly to be updated when the console is ready.

We revert the earlier printk related changes, and then add back the DEVNAME:0.0 functionality based on Petr's code snippet. And we end up reducing the code quite a bit too this way.

And we also revert all the unusable serial core console quirk handling, it does not do anything for the legacy "ttyS" named consoles. And then we add a minimal serial_base_match_and_update_preferred_console().

The reason we want DEVNAME:0.0 style consoles is it helps addressing the console based on the connected serial port controller device rather than using the hardcoded ttyS addressing. And that helps with issues related to the console moving around after togging the HSUART option in the BIOS, or when new ports are enabled in devicetree and aliases are not updated.

Tony Lindgren (4):
printk: Revert add_preferred_console_match() related commits
printk: Add match_devname_and_update_preferred_console()
serial: core: Revert unusable console quirk handling
serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 11 lines
 drivers/tty/serial/8250/8250_core.c  |   5 -
 drivers/tty/serial/serial_base.h     |  22 +---
 drivers/tty/serial/serial_base_bus.c | 116 +++------------------
 drivers/tty/serial/serial_core.c     |   2 +-
 include/linux/printk.h               |   5 +-
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   7 +-
 kernel/printk/printk.c               | 122 ++++++++++++++++------
 9 files changed, 112 insertions(+), 315 deletions(-)
 delete mode 100644 kernel/printk/conopt.c

The patchset looks ready for linux-next. And I have pushed it into printk/linux.git, branch for-6.10-register-console-devname.

I am not sure about the mainline. We need to fix the regression in 6.10. The change is not trivial and rc5 is knocking on the doors.

Unfortunately, the patchset intermixes reverts and new code. So that it can't be used for simple revert as is.

I am quite confident that the new code works as expected. It changes tricky code but the logic of the change is quite straightforward.

I see three solutions:

1. Linus could merge the changes directly into rc5.

2. I could send a pull request after it survives few days in linux-next.

3. Or we rework the patchset. And do pure revert for 6.10 and add the feature a clean way for-6.11.

I personally prefer the 3rd solution. But I am super conservative. I guess that most other people would go with the other 2 solutions.

Best Regards,
Petr

#9 lore raw
Quoted 62 lines · Petr Mladek

Added Linus into Cc.

Quoted 31 lines · Tony Lindgren

Hi,

Recent changes to add support for DEVNAME:0.0 style consoles caused a regression with the preferred console order where the last console on the kernel command line is no longer the preferred console.

The following four changes fix the issue using Petr's suggestion that does not involve calling __add_preferred_console() later on again, and adds the deferred consoles to the console_cmdline[] directly to be updated when the console is ready.

We revert the earlier printk related changes, and then add back the DEVNAME:0.0 functionality based on Petr's code snippet. And we end up reducing the code quite a bit too this way.

And we also revert all the unusable serial core console quirk handling, it does not do anything for the legacy "ttyS" named consoles. And then we add a minimal serial_base_match_and_update_preferred_console().

The reason we want DEVNAME:0.0 style consoles is it helps addressing the console based on the connected serial port controller device rather than using the hardcoded ttyS addressing. And that helps with issues related to the console moving around after togging the HSUART option in the BIOS, or when new ports are enabled in devicetree and aliases are not updated.

Tony Lindgren (4):
printk: Revert add_preferred_console_match() related commits
printk: Add match_devname_and_update_preferred_console()
serial: core: Revert unusable console quirk handling
serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 11 lines
 drivers/tty/serial/8250/8250_core.c  |   5 -
 drivers/tty/serial/serial_base.h     |  22 +---
 drivers/tty/serial/serial_base_bus.c | 116 +++------------------
 drivers/tty/serial/serial_core.c     |   2 +-
 include/linux/printk.h               |   5 +-
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   7 +-
 kernel/printk/printk.c               | 122 ++++++++++++++++------
 9 files changed, 112 insertions(+), 315 deletions(-)
 delete mode 100644 kernel/printk/conopt.c

The patchset looks ready for linux-next. And I have pushed it into printk/linux.git, branch for-6.10-register-console-devname.

I am not sure about the mainline. We need to fix the regression in 6.10. The change is not trivial and rc5 is knocking on the doors.

Unfortunately, the patchset intermixes reverts and new code. So that it can't be used for simple revert as is.

I am quite confident that the new code works as expected. It changes tricky code but the logic of the change is quite straightforward.

I see three solutions:

Quoted 7 lines

1. Linus could merge the changes directly into rc5.

2. I could send a pull request after it survives few days in linux-next.

3. Or we rework the patchset. And do pure revert for 6.10 and add the feature a clean way for-6.11.

Pure revert for 6.10 might be good, as it's late in the cycle. Let me know the git ids and I can do that.

Quoted 2 lines

I personally prefer the 3rd solution. But I am super conservative. I guess that most other people would go with the other 2 solutions.

I'll be conservative here as well.

thanks,

greg k-h

#10 lore raw
Quoted 65 lines · Greg Kroah-Hartman
Quoted 64 lines · Petr Mladek

Added Linus into Cc.

Quoted 31 lines · Tony Lindgren

Hi,

Recent changes to add support for DEVNAME:0.0 style consoles caused a regression with the preferred console order where the last console on the kernel command line is no longer the preferred console.

The following four changes fix the issue using Petr's suggestion that does not involve calling __add_preferred_console() later on again, and adds the deferred consoles to the console_cmdline[] directly to be updated when the console is ready.

We revert the earlier printk related changes, and then add back the DEVNAME:0.0 functionality based on Petr's code snippet. And we end up reducing the code quite a bit too this way.

And we also revert all the unusable serial core console quirk handling, it does not do anything for the legacy "ttyS" named consoles. And then we add a minimal serial_base_match_and_update_preferred_console().

The reason we want DEVNAME:0.0 style consoles is it helps addressing the console based on the connected serial port controller device rather than using the hardcoded ttyS addressing. And that helps with issues related to the console moving around after togging the HSUART option in the BIOS, or when new ports are enabled in devicetree and aliases are not updated.

Tony Lindgren (4):
printk: Revert add_preferred_console_match() related commits
printk: Add match_devname_and_update_preferred_console()
serial: core: Revert unusable console quirk handling
serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 11 lines
 drivers/tty/serial/8250/8250_core.c  |   5 -
 drivers/tty/serial/serial_base.h     |  22 +---
 drivers/tty/serial/serial_base_bus.c | 116 +++------------------
 drivers/tty/serial/serial_core.c     |   2 +-
 include/linux/printk.h               |   5 +-
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   7 +-
 kernel/printk/printk.c               | 122 ++++++++++++++++------
 9 files changed, 112 insertions(+), 315 deletions(-)
 delete mode 100644 kernel/printk/conopt.c

The patchset looks ready for linux-next. And I have pushed it into printk/linux.git, branch for-6.10-register-console-devname.

I am not sure about the mainline. We need to fix the regression in 6.10. The change is not trivial and rc5 is knocking on the doors.

Unfortunately, the patchset intermixes reverts and new code. So that it can't be used for simple revert as is.

I am quite confident that the new code works as expected. It changes tricky code but the logic of the change is quite straightforward.

I see three solutions:

1. Linus could merge the changes directly into rc5.

Quoted 8 lines
Quoted 5 lines

2. I could send a pull request after it survives few days in linux-next.

3. Or we rework the patchset. And do pure revert for 6.10 and add the feature a clean way for-6.11.

Pure revert for 6.10 might be good, as it's late in the cycle. Let me know the git ids and I can do that.

Here's the list of git ids to revert:

$ git log --abbrev=12 --pretty=format:"%h (\"%s\")" v6.9..v6.10-rc5 \
--author="Tony Lindgren" kernel/printk drivers/tty/ Documentation/admin-guide/
b20172ca6bf4 ("serial: core: Fix ifdef for serial base console functions")
4547cd76f08a ("serial: 8250: Fix add preferred console for serial8250_isa_init_ports()")
5c3a766e9f05 ("Documentation: kernel-parameters: Add DEVNAME:0.0 format for serial ports")
a8b04cfe7dad ("serial: 8250: Add preferred console in serial8250_isa_init_ports()")
a0f32e2dd998 ("serial: core: Handle serial console options")
787a1cabac01 ("serial: core: Add support for DEVNAME:0.0 style naming for kernel console")
b73c9cbe4f1f ("printk: Flag register_console() if console is set on command line")
8a831c584e6e ("printk: Don't try to parse DEVNAME:0.0 console options")
f03e8c1060f8 ("printk: Save console options for add_preferred_console_match()")
Quoted 4 lines
Quoted 2 lines

I personally prefer the 3rd solution. But I am super conservative. I guess that most other people would go with the other 2 solutions.

I'll be conservative here as well.

Conservative is good.

Regards,

Tony

#11 lore raw
Quoted 77 lines · Tony Lindgren
Quoted 65 lines · Greg Kroah-Hartman
Quoted 64 lines · Petr Mladek

Added Linus into Cc.

Quoted 31 lines · Tony Lindgren

Hi,

Recent changes to add support for DEVNAME:0.0 style consoles caused a regression with the preferred console order where the last console on the kernel command line is no longer the preferred console.

The following four changes fix the issue using Petr's suggestion that does not involve calling __add_preferred_console() later on again, and adds the deferred consoles to the console_cmdline[] directly to be updated when the console is ready.

We revert the earlier printk related changes, and then add back the DEVNAME:0.0 functionality based on Petr's code snippet. And we end up reducing the code quite a bit too this way.

And we also revert all the unusable serial core console quirk handling, it does not do anything for the legacy "ttyS" named consoles. And then we add a minimal serial_base_match_and_update_preferred_console().

The reason we want DEVNAME:0.0 style consoles is it helps addressing the console based on the connected serial port controller device rather than using the hardcoded ttyS addressing. And that helps with issues related to the console moving around after togging the HSUART option in the BIOS, or when new ports are enabled in devicetree and aliases are not updated.

Tony Lindgren (4):
printk: Revert add_preferred_console_match() related commits
printk: Add match_devname_and_update_preferred_console()
serial: core: Revert unusable console quirk handling
serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 11 lines
 drivers/tty/serial/8250/8250_core.c  |   5 -
 drivers/tty/serial/serial_base.h     |  22 +---
 drivers/tty/serial/serial_base_bus.c | 116 +++------------------
 drivers/tty/serial/serial_core.c     |   2 +-
 include/linux/printk.h               |   5 +-
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   7 +-
 kernel/printk/printk.c               | 122 ++++++++++++++++------
 9 files changed, 112 insertions(+), 315 deletions(-)
 delete mode 100644 kernel/printk/conopt.c

The patchset looks ready for linux-next. And I have pushed it into printk/linux.git, branch for-6.10-register-console-devname.

I am not sure about the mainline. We need to fix the regression in 6.10. The change is not trivial and rc5 is knocking on the doors.

Unfortunately, the patchset intermixes reverts and new code. So that it can't be used for simple revert as is.

I am quite confident that the new code works as expected. It changes tricky code but the logic of the change is quite straightforward.

I see three solutions:

1. Linus could merge the changes directly into rc5.

Quoted 8 lines
Quoted 5 lines

2. I could send a pull request after it survives few days in linux-next.

3. Or we rework the patchset. And do pure revert for 6.10 and add the feature a clean way for-6.11.

Pure revert for 6.10 might be good, as it's late in the cycle. Let me know the git ids and I can do that.

Here's the list of git ids to revert:

Quoted 11 lines
$ git log --abbrev=12 --pretty=format:"%h (\"%s\")" v6.9..v6.10-rc5 \
--author="Tony Lindgren" kernel/printk drivers/tty/ Documentation/admin-guide/
b20172ca6bf4 ("serial: core: Fix ifdef for serial base console functions")
4547cd76f08a ("serial: 8250: Fix add preferred console for serial8250_isa_init_ports()")
5c3a766e9f05 ("Documentation: kernel-parameters: Add DEVNAME:0.0 format for serial ports")
a8b04cfe7dad ("serial: 8250: Add preferred console in serial8250_isa_init_ports()")
a0f32e2dd998 ("serial: core: Handle serial console options")
787a1cabac01 ("serial: core: Add support for DEVNAME:0.0 style naming for kernel console")
b73c9cbe4f1f ("printk: Flag register_console() if console is set on command line")
8a831c584e6e ("printk: Don't try to parse DEVNAME:0.0 console options")
f03e8c1060f8 ("printk: Save console options for add_preferred_console_match()")

All now reverted, thanks!

greg k-h

maintainer
#12 lore raw
Quoted 90 lines · Greg Kroah-Hartman
Quoted 77 lines · Tony Lindgren
Quoted 65 lines · Greg Kroah-Hartman
Quoted 64 lines · Petr Mladek

Added Linus into Cc.

Quoted 31 lines · Tony Lindgren

Hi,

Recent changes to add support for DEVNAME:0.0 style consoles caused a regression with the preferred console order where the last console on the kernel command line is no longer the preferred console.

The following four changes fix the issue using Petr's suggestion that does not involve calling __add_preferred_console() later on again, and adds the deferred consoles to the console_cmdline[] directly to be updated when the console is ready.

We revert the earlier printk related changes, and then add back the DEVNAME:0.0 functionality based on Petr's code snippet. And we end up reducing the code quite a bit too this way.

And we also revert all the unusable serial core console quirk handling, it does not do anything for the legacy "ttyS" named consoles. And then we add a minimal serial_base_match_and_update_preferred_console().

The reason we want DEVNAME:0.0 style consoles is it helps addressing the console based on the connected serial port controller device rather than using the hardcoded ttyS addressing. And that helps with issues related to the console moving around after togging the HSUART option in the BIOS, or when new ports are enabled in devicetree and aliases are not updated.

Tony Lindgren (4):
printk: Revert add_preferred_console_match() related commits
printk: Add match_devname_and_update_preferred_console()
serial: core: Revert unusable console quirk handling
serial: core: Add serial_base_match_and_update_preferred_console()

Quoted 11 lines
 drivers/tty/serial/8250/8250_core.c  |   5 -
 drivers/tty/serial/serial_base.h     |  22 +---
 drivers/tty/serial/serial_base_bus.c | 116 +++------------------
 drivers/tty/serial/serial_core.c     |   2 +-
 include/linux/printk.h               |   5 +-
 kernel/printk/Makefile               |   2 +-
 kernel/printk/conopt.c               | 146 ---------------------------
 kernel/printk/console_cmdline.h      |   7 +-
 kernel/printk/printk.c               | 122 ++++++++++++++++------
 9 files changed, 112 insertions(+), 315 deletions(-)
 delete mode 100644 kernel/printk/conopt.c

The patchset looks ready for linux-next. And I have pushed it into printk/linux.git, branch for-6.10-register-console-devname.

I am not sure about the mainline. We need to fix the regression in 6.10. The change is not trivial and rc5 is knocking on the doors.

Unfortunately, the patchset intermixes reverts and new code. So that it can't be used for simple revert as is.

I am quite confident that the new code works as expected. It changes tricky code but the logic of the change is quite straightforward.

I see three solutions:

1. Linus could merge the changes directly into rc5.

Quoted 8 lines
Quoted 5 lines

2. I could send a pull request after it survives few days in linux-next.

3. Or we rework the patchset. And do pure revert for 6.10 and add the feature a clean way for-6.11.

Pure revert for 6.10 might be good, as it's late in the cycle. Let me know the git ids and I can do that.

Here's the list of git ids to revert:

Quoted 11 lines
$ git log --abbrev=12 --pretty=format:"%h (\"%s\")" v6.9..v6.10-rc5 \
--author="Tony Lindgren" kernel/printk drivers/tty/ Documentation/admin-guide/
b20172ca6bf4 ("serial: core: Fix ifdef for serial base console functions")
4547cd76f08a ("serial: 8250: Fix add preferred console for serial8250_isa_init_ports()")
5c3a766e9f05 ("Documentation: kernel-parameters: Add DEVNAME:0.0 format for serial ports")
a8b04cfe7dad ("serial: 8250: Add preferred console in serial8250_isa_init_ports()")
a0f32e2dd998 ("serial: core: Handle serial console options")
787a1cabac01 ("serial: core: Add support for DEVNAME:0.0 style naming for kernel console")
b73c9cbe4f1f ("printk: Flag register_console() if console is set on command line")
8a831c584e6e ("printk: Don't try to parse DEVNAME:0.0 console options")
f03e8c1060f8 ("printk: Save console options for add_preferred_console_match()")
Quoted 1 line

All now reverted, thanks!

Great, thanks Greg!

Tony, could you please send a new patchset which would provide the new solution on top of this revert? It might make sense to wait until the revert reaches mainline.

Best Regards,
Petr

#13 lore raw
Quoted 8 lines · Petr Mladek
Quoted 1 line · Greg Kroah-Hartman

All now reverted, thanks!

Great, thanks Greg!

Tony, could you please send a new patchset which would provide the new solution on top of this revert? It might make sense to wait until the revert reaches mainline.

Yup will do.

Thanks,

Tony