[linux-2.6-block.git] / Documentation / ramoops.txt

Ramoops oops/panic logger
=========================

Sergiu Iordache <sergiu@chromium.org>

Updated: 17 November 2011

0. Introduction

Ramoops is an oops/panic logger that writes its logs to RAM before the system
crashes. It works by logging oopses and panics in a circular buffer. Ramoops
needs a system with persistent RAM so that the content of that area can
survive after a restart.

1. Ramoops concepts

Ramoops uses a predefined memory area to store the dump. The start and size
and type of the memory area are set using three variables:
  * "mem_address" for the start
  * "mem_size" for the size. The memory size will be rounded down to a
  power of two.
  * "mem_type" to specifiy if the memory type (default is pgprot_writecombine).

Typically the default value of mem_type=0 should be used as that sets the pstore
mapping to pgprot_writecombine. Setting mem_type=1 attempts to use
pgprot_noncached, which only works on some platforms. This is because pstore
depends on atomic operations. At least on ARM, pgprot_noncached causes the
memory to be mapped strongly ordered, and atomic operations on strongly ordered
memory are implementation defined, and won't work on many ARMs such as omaps.

The memory area is divided into "record_size" chunks (also rounded down to
power of two) and each oops/panic writes a "record_size" chunk of
information.

Dumping both oopses and panics can be done by setting 1 in the "dump_oops"
variable while setting 0 in that variable dumps only the panics.

The module uses a counter to record multiple dumps but the counter gets reset
on restart (i.e. new dumps after the restart will overwrite old ones).

Ramoops also supports software ECC protection of persistent memory regions.
This might be useful when a hardware reset was used to bring the machine back
to life (i.e. a watchdog triggered). In such cases, RAM may be somewhat
corrupt, but usually it is restorable.

2. Setting the parameters

Setting the ramoops parameters can be done in 3 different manners:
 1. Use the module parameters (which have the names of the variables described
 as before).
 For quick debugging, you can also reserve parts of memory during boot
 and then use the reserved memory for ramoops. For example, assuming a machine
 with > 128 MB of memory, the following kernel command line will tell the
 kernel to use only the first 128 MB of memory, and place ECC-protected ramoops
 region at 128 MB boundary:
 "mem=128M ramoops.mem_address=0x8000000 ramoops.ecc=1"
 2. Use Device Tree bindings, as described in
 Documentation/device-tree/bindings/misc/ramoops.txt.
 3. Use a platform device and set the platform data. The parameters can then
 be set through that platform data. An example of doing that is:

#include <linux/pstore_ram.h>
[...]

static struct ramoops_platform_data ramoops_data = {
        .mem_size               = <...>,
        .mem_address            = <...>,
        .mem_type               = <...>,
        .record_size            = <...>,
        .dump_oops              = <...>,
        .ecc                    = <...>,
};

static struct platform_device ramoops_dev = {
        .name = "ramoops",
        .dev = {
                .platform_data = &ramoops_data,
        },
};

[... inside a function ...]
int ret;

ret = platform_device_register(&ramoops_dev);
if (ret) {
	printk(KERN_ERR "unable to register platform device\n");
	return ret;
}

You can specify either RAM memory or peripheral devices' memory. However, when
specifying RAM, be sure to reserve the memory by issuing memblock_reserve()
very early in the architecture code, e.g.:

#include <linux/memblock.h>

memblock_reserve(ramoops_data.mem_address, ramoops_data.mem_size);

3. Dump format

The data dump begins with a header, currently defined as "====" followed by a
timestamp and a new line. The dump then continues with the actual data.

4. Reading the data

The dump data can be read from the pstore filesystem. The format for these
files is "dmesg-ramoops-N", where N is the record number in memory. To delete
a stored record from RAM, simply unlink the respective pstore file.

5. Persistent function tracing

Persistent function tracing might be useful for debugging software or hardware
related hangs. The functions call chain log is stored in a "ftrace-ramoops"
file. Here is an example of usage:

 # mount -t debugfs debugfs /sys/kernel/debug/
 # echo 1 > /sys/kernel/debug/pstore/record_ftrace
 # reboot -f
 [...]
 # mount -t pstore pstore /mnt/
 # tail /mnt/ftrace-ramoops
 0 ffffffff8101ea64  ffffffff8101bcda  native_apic_mem_read <- disconnect_bsp_APIC+0x6a/0xc0
 0 ffffffff8101ea44  ffffffff8101bcf6  native_apic_mem_write <- disconnect_bsp_APIC+0x86/0xc0
 0 ffffffff81020084  ffffffff8101a4b5  hpet_disable <- native_machine_shutdown+0x75/0x90
 0 ffffffff81005f94  ffffffff8101a4bb  iommu_shutdown_noop <- native_machine_shutdown+0x7b/0x90
 0 ffffffff8101a6a1  ffffffff8101a437  native_machine_emergency_restart <- native_machine_restart+0x37/0x40
 0 ffffffff811f9876  ffffffff8101a73a  acpi_reboot <- native_machine_emergency_restart+0xaa/0x1e0
 0 ffffffff8101a514  ffffffff8101a772  mach_reboot_fixups <- native_machine_emergency_restart+0xe2/0x1e0
 0 ffffffff811d9c54  ffffffff8101a7a0  __const_udelay <- native_machine_emergency_restart+0x110/0x1e0
 0 ffffffff811d9c34  ffffffff811d9c80  __delay <- __const_udelay+0x30/0x40
 0 ffffffff811d9d14  ffffffff811d9c3f  delay_tsc <- __delay+0xf/0x20
Commit	Line	Data
4126dacb SI	1	Ramoops oops/panic logger
	2	=========================
	3
	4	Sergiu Iordache <sergiu@chromium.org>
	5
9ba80d99	6	Updated: 17 November 2011
4126dacb SI	7
	8	0. Introduction
	9
	10	Ramoops is an oops/panic logger that writes its logs to RAM before the system
	11	crashes. It works by logging oopses and panics in a circular buffer. Ramoops
	12	needs a system with persistent RAM so that the content of that area can
	13	survive after a restart.
	14
	15	1. Ramoops concepts
	16
027bc8b0 TL	17	Ramoops uses a predefined memory area to store the dump. The start and size
027bc8b0 TL	18	and type of the memory area are set using three variables:
4126dacb SI	19	* "mem_address" for the start
	20	* "mem_size" for the size. The memory size will be rounded down to a
	21	power of two.
027bc8b0 TL	22	* "mem_type" to specifiy if the memory type (default is pgprot_writecombine).
	23
	24	Typically the default value of mem_type=0 should be used as that sets the pstore
	25	mapping to pgprot_writecombine. Setting mem_type=1 attempts to use
	26	pgprot_noncached, which only works on some platforms. This is because pstore
	27	depends on atomic operations. At least on ARM, pgprot_noncached causes the
	28	memory to be mapped strongly ordered, and atomic operations on strongly ordered
	29	memory are implementation defined, and won't work on many ARMs such as omaps.
4126dacb SI	30
	31	The memory area is divided into "record_size" chunks (also rounded down to
	32	power of two) and each oops/panic writes a "record_size" chunk of
	33	information.
	34
	35	Dumping both oopses and panics can be done by setting 1 in the "dump_oops"
	36	variable while setting 0 in that variable dumps only the panics.
	37
	38	The module uses a counter to record multiple dumps but the counter gets reset
	39	on restart (i.e. new dumps after the restart will overwrite old ones).
	40
39eb7e97 AV	41	Ramoops also supports software ECC protection of persistent memory regions.
	42	This might be useful when a hardware reset was used to bring the machine back
	43	to life (i.e. a watchdog triggered). In such cases, RAM may be somewhat
	44	corrupt, but usually it is restorable.
	45
4126dacb SI	46	2. Setting the parameters
4126dacb SI	47
35da6094	48	Setting the ramoops parameters can be done in 3 different manners:
4126dacb SI	49	1. Use the module parameters (which have the names of the variables described
4126dacb SI	50	as before).
958502d8 AV	51	For quick debugging, you can also reserve parts of memory during boot
	52	and then use the reserved memory for ramoops. For example, assuming a machine
	53	with > 128 MB of memory, the following kernel command line will tell the
	54	kernel to use only the first 128 MB of memory, and place ECC-protected ramoops
	55	region at 128 MB boundary:
	56	"mem=128M ramoops.mem_address=0x8000000 ramoops.ecc=1"
35da6094 GH	57	2. Use Device Tree bindings, as described in
	58	Documentation/device-tree/bindings/misc/ramoops.txt.
	59	3. Use a platform device and set the platform data. The parameters can then
4126dacb SI	60	be set through that platform data. An example of doing that is:
4126dacb SI	61
1894a253	62	#include <linux/pstore_ram.h>
4126dacb SI	63	[...]
	64
	65	static struct ramoops_platform_data ramoops_data = {
	66	.mem_size = <...>,
	67	.mem_address = <...>,
027bc8b0	68	.mem_type = <...>,
4126dacb SI	69	.record_size = <...>,
4126dacb SI	70	.dump_oops = <...>,
39eb7e97	71	.ecc = <...>,
4126dacb SI	72	};
	73
	74	static struct platform_device ramoops_dev = {
	75	.name = "ramoops",
	76	.dev = {
	77	.platform_data = &ramoops_data,
	78	},
	79	};
	80
	81	[... inside a function ...]
	82	int ret;
	83
	84	ret = platform_device_register(&ramoops_dev);
	85	if (ret) {
	86	printk(KERN_ERR "unable to register platform device\n");
	87	return ret;
	88	}
	89
958502d8 AV	90	You can specify either RAM memory or peripheral devices' memory. However, when
	91	specifying RAM, be sure to reserve the memory by issuing memblock_reserve()
	92	very early in the architecture code, e.g.:
	93
	94	#include <linux/memblock.h>
	95
	96	memblock_reserve(ramoops_data.mem_address, ramoops_data.mem_size);
	97
4126dacb SI	98	3. Dump format
	99
	100	The data dump begins with a header, currently defined as "====" followed by a
	101	timestamp and a new line. The dump then continues with the actual data.
	102
	103	4. Reading the data
	104
9ba80d99 KC	105	The dump data can be read from the pstore filesystem. The format for these
	106	files is "dmesg-ramoops-N", where N is the record number in memory. To delete
	107	a stored record from RAM, simply unlink the respective pstore file.
a694d1b5 AV	108
	109	5. Persistent function tracing
	110
	111	Persistent function tracing might be useful for debugging software or hardware
	112	related hangs. The functions call chain log is stored in a "ftrace-ramoops"
	113	file. Here is an example of usage:
	114
	115	# mount -t debugfs debugfs /sys/kernel/debug/
65f8c95e	116	# echo 1 > /sys/kernel/debug/pstore/record_ftrace
a694d1b5 AV	117	# reboot -f
	118	[...]
	119	# mount -t pstore pstore /mnt/
	120	# tail /mnt/ftrace-ramoops
	121	0 ffffffff8101ea64 ffffffff8101bcda native_apic_mem_read <- disconnect_bsp_APIC+0x6a/0xc0
	122	0 ffffffff8101ea44 ffffffff8101bcf6 native_apic_mem_write <- disconnect_bsp_APIC+0x86/0xc0
	123	0 ffffffff81020084 ffffffff8101a4b5 hpet_disable <- native_machine_shutdown+0x75/0x90
	124	0 ffffffff81005f94 ffffffff8101a4bb iommu_shutdown_noop <- native_machine_shutdown+0x7b/0x90
	125	0 ffffffff8101a6a1 ffffffff8101a437 native_machine_emergency_restart <- native_machine_restart+0x37/0x40
	126	0 ffffffff811f9876 ffffffff8101a73a acpi_reboot <- native_machine_emergency_restart+0xaa/0x1e0
	127	0 ffffffff8101a514 ffffffff8101a772 mach_reboot_fixups <- native_machine_emergency_restart+0xe2/0x1e0
	128	0 ffffffff811d9c54 ffffffff8101a7a0 __const_udelay <- native_machine_emergency_restart+0x110/0x1e0
	129	0 ffffffff811d9c34 ffffffff811d9c80 __delay <- __const_udelay+0x30/0x40
	130	0 ffffffff811d9d14 ffffffff811d9c3f delay_tsc <- __delay+0xf/0x20