1Ramoops oops/panic logger 2========================= 3 4Sergiu Iordache <sergiu@chromium.org> 5 6Updated: 17 November 2011 7 8Introduction 9------------ 10 11Ramoops is an oops/panic logger that writes its logs to RAM before the system 12crashes. It works by logging oopses and panics in a circular buffer. Ramoops 13needs a system with persistent RAM so that the content of that area can 14survive after a restart. 15 16Ramoops concepts 17---------------- 18 19Ramoops uses a predefined memory area to store the dump. The start and size 20and type of the memory area are set using three variables: 21 22 * ``mem_address`` for the start 23 * ``mem_size`` for the size. The memory size will be rounded down to a 24 power of two. 25 * ``mem_type`` to specify if the memory type (default is pgprot_writecombine). 26 27Typically the default value of ``mem_type=0`` should be used as that sets the pstore 28mapping to pgprot_writecombine. Setting ``mem_type=1`` attempts to use 29``pgprot_noncached``, which only works on some platforms. This is because pstore 30depends on atomic operations. At least on ARM, pgprot_noncached causes the 31memory to be mapped strongly ordered, and atomic operations on strongly ordered 32memory are implementation defined, and won't work on many ARMs such as omaps. 33 34The memory area is divided into ``record_size`` chunks (also rounded down to 35power of two) and each kmesg dump writes a ``record_size`` chunk of 36information. 37 38Limiting which kinds of kmsg dumps are stored can be controlled via 39the ``max_reason`` value, as defined in include/linux/kmsg_dump.h's 40``enum kmsg_dump_reason``. For example, to store both Oopses and Panics, 41``max_reason`` should be set to 2 (KMSG_DUMP_OOPS), to store only Panics 42``max_reason`` should be set to 1 (KMSG_DUMP_PANIC). Setting this to 0 43(KMSG_DUMP_UNDEF), means the reason filtering will be controlled by the 44``printk.always_kmsg_dump`` boot param: if unset, it'll be KMSG_DUMP_OOPS, 45otherwise KMSG_DUMP_MAX. 46 47The module uses a counter to record multiple dumps but the counter gets reset 48on restart (i.e. new dumps after the restart will overwrite old ones). 49 50Ramoops also supports software ECC protection of persistent memory regions. 51This might be useful when a hardware reset was used to bring the machine back 52to life (i.e. a watchdog triggered). In such cases, RAM may be somewhat 53corrupt, but usually it is restorable. 54 55Setting the parameters 56---------------------- 57 58Setting the ramoops parameters can be done in several different manners: 59 60 A. Use the module parameters (which have the names of the variables described 61 as before). For quick debugging, you can also reserve parts of memory during 62 boot and then use the reserved memory for ramoops. For example, assuming a 63 machine with > 128 MB of memory, the following kernel command line will tell 64 the kernel to use only the first 128 MB of memory, and place ECC-protected 65 ramoops region at 128 MB boundary:: 66 67 mem=128M ramoops.mem_address=0x8000000 ramoops.ecc=1 68 69 B. Use Device Tree bindings, as described in 70 ``Documentation/devicetree/bindings/reserved-memory/ramoops.txt``. 71 For example:: 72 73 reserved-memory { 74 #address-cells = <2>; 75 #size-cells = <2>; 76 ranges; 77 78 ramoops@8f000000 { 79 compatible = "ramoops"; 80 reg = <0 0x8f000000 0 0x100000>; 81 record-size = <0x4000>; 82 console-size = <0x4000>; 83 }; 84 }; 85 86 C. Use a platform device and set the platform data. The parameters can then 87 be set through that platform data. An example of doing that is: 88 89 .. code-block:: c 90 91 #include <linux/pstore_ram.h> 92 [...] 93 94 static struct ramoops_platform_data ramoops_data = { 95 .mem_size = <...>, 96 .mem_address = <...>, 97 .mem_type = <...>, 98 .record_size = <...>, 99 .max_reason = <...>, 100 .ecc = <...>, 101 }; 102 103 static struct platform_device ramoops_dev = { 104 .name = "ramoops", 105 .dev = { 106 .platform_data = &ramoops_data, 107 }, 108 }; 109 110 [... inside a function ...] 111 int ret; 112 113 ret = platform_device_register(&ramoops_dev); 114 if (ret) { 115 printk(KERN_ERR "unable to register platform device\n"); 116 return ret; 117 } 118 119You can specify either RAM memory or peripheral devices' memory. However, when 120specifying RAM, be sure to reserve the memory by issuing memblock_reserve() 121very early in the architecture code, e.g.:: 122 123 #include <linux/memblock.h> 124 125 memblock_reserve(ramoops_data.mem_address, ramoops_data.mem_size); 126 127Dump format 128----------- 129 130The data dump begins with a header, currently defined as ``====`` followed by a 131timestamp and a new line. The dump then continues with the actual data. 132 133Reading the data 134---------------- 135 136The dump data can be read from the pstore filesystem. The format for these 137files is ``dmesg-ramoops-N``, where N is the record number in memory. To delete 138a stored record from RAM, simply unlink the respective pstore file. 139 140Persistent function tracing 141--------------------------- 142 143Persistent function tracing might be useful for debugging software or hardware 144related hangs. The functions call chain log is stored in a ``ftrace-ramoops`` 145file. Here is an example of usage:: 146 147 # mount -t debugfs debugfs /sys/kernel/debug/ 148 # echo 1 > /sys/kernel/debug/pstore/record_ftrace 149 # reboot -f 150 [...] 151 # mount -t pstore pstore /mnt/ 152 # tail /mnt/ftrace-ramoops 153 0 ffffffff8101ea64 ffffffff8101bcda native_apic_mem_read <- disconnect_bsp_APIC+0x6a/0xc0 154 0 ffffffff8101ea44 ffffffff8101bcf6 native_apic_mem_write <- disconnect_bsp_APIC+0x86/0xc0 155 0 ffffffff81020084 ffffffff8101a4b5 hpet_disable <- native_machine_shutdown+0x75/0x90 156 0 ffffffff81005f94 ffffffff8101a4bb iommu_shutdown_noop <- native_machine_shutdown+0x7b/0x90 157 0 ffffffff8101a6a1 ffffffff8101a437 native_machine_emergency_restart <- native_machine_restart+0x37/0x40 158 0 ffffffff811f9876 ffffffff8101a73a acpi_reboot <- native_machine_emergency_restart+0xaa/0x1e0 159 0 ffffffff8101a514 ffffffff8101a772 mach_reboot_fixups <- native_machine_emergency_restart+0xe2/0x1e0 160 0 ffffffff811d9c54 ffffffff8101a7a0 __const_udelay <- native_machine_emergency_restart+0x110/0x1e0 161 0 ffffffff811d9c34 ffffffff811d9c80 __delay <- __const_udelay+0x30/0x40 162 0 ffffffff811d9d14 ffffffff811d9c3f delay_tsc <- __delay+0xf/0x20 163