1================================================================ 2Documentation for Kdump - The kexec-based Crash Dumping Solution 3================================================================ 4 5This document includes overview, setup, installation, and analysis 6information. 7 8Overview 9======== 10 11Kdump uses kexec to quickly boot to a dump-capture kernel whenever a 12dump of the system kernel's memory needs to be taken (for example, when 13the system panics). The system kernel's memory image is preserved across 14the reboot and is accessible to the dump-capture kernel. 15 16You can use common commands, such as cp, scp or makedumpfile to copy 17the memory image to a dump file on the local disk, or across the network 18to a remote system. 19 20Kdump and kexec are currently supported on the x86, x86_64, ppc64, 21s390x, arm and arm64 architectures. 22 23When the system kernel boots, it reserves a small section of memory for 24the dump-capture kernel. This ensures that ongoing Direct Memory Access 25(DMA) from the system kernel does not corrupt the dump-capture kernel. 26The kexec -p command loads the dump-capture kernel into this reserved 27memory. 28 29On x86 machines, the first 640 KB of physical memory is needed for boot, 30regardless of where the kernel loads. For simpler handling, the whole 31low 1M is reserved to avoid any later kernel or device driver writing 32data into this area. Like this, the low 1M can be reused as system RAM 33by kdump kernel without extra handling. 34 35On PPC64 machines first 32KB of physical memory is needed for booting 36regardless of where the kernel is loaded and to support 64K page size 37kexec backs up the first 64KB memory. 38 39For s390x, when kdump is triggered, the crashkernel region is exchanged 40with the region [0, crashkernel region size] and then the kdump kernel 41runs in [0, crashkernel region size]. Therefore no relocatable kernel is 42needed for s390x. 43 44All of the necessary information about the system kernel's core image is 45encoded in the ELF format, and stored in a reserved area of memory 46before a crash. The physical address of the start of the ELF header is 47passed to the dump-capture kernel through the elfcorehdr= boot 48parameter. Optionally the size of the ELF header can also be passed 49when using the elfcorehdr=[size[KMG]@]offset[KMG] syntax. 50 51With the dump-capture kernel, you can access the memory image through 52/proc/vmcore. This exports the dump as an ELF-format file that you can 53write out using file copy commands such as cp or scp. You can also use 54makedumpfile utility to analyze and write out filtered contents with 55options, e.g with '-d 31' it will only write out kernel data. Further, 56you can use analysis tools such as the GNU Debugger (GDB) and the Crash 57tool to debug the dump file. This method ensures that the dump pages are 58correctly ordered. 59 60Setup and Installation 61====================== 62 63Install kexec-tools 64------------------- 65 661) Login as the root user. 67 682) Download the kexec-tools user-space package from the following URL: 69 70http://kernel.org/pub/linux/utils/kernel/kexec/kexec-tools.tar.gz 71 72This is a symlink to the latest version. 73 74The latest kexec-tools git tree is available at: 75 76- git://git.kernel.org/pub/scm/utils/kernel/kexec/kexec-tools.git 77- http://www.kernel.org/pub/scm/utils/kernel/kexec/kexec-tools.git 78 79There is also a gitweb interface available at 80http://www.kernel.org/git/?p=utils/kernel/kexec/kexec-tools.git 81 82More information about kexec-tools can be found at 83http://horms.net/projects/kexec/ 84 853) Unpack the tarball with the tar command, as follows:: 86 87 tar xvpzf kexec-tools.tar.gz 88 894) Change to the kexec-tools directory, as follows:: 90 91 cd kexec-tools-VERSION 92 935) Configure the package, as follows:: 94 95 ./configure 96 976) Compile the package, as follows:: 98 99 make 100 1017) Install the package, as follows:: 102 103 make install 104 105 106Build the system and dump-capture kernels 107----------------------------------------- 108There are two possible methods of using Kdump. 109 1101) Build a separate custom dump-capture kernel for capturing the 111 kernel core dump. 112 1132) Or use the system kernel binary itself as dump-capture kernel and there is 114 no need to build a separate dump-capture kernel. This is possible 115 only with the architectures which support a relocatable kernel. As 116 of today, i386, x86_64, ppc64, arm and arm64 architectures support 117 relocatable kernel. 118 119Building a relocatable kernel is advantageous from the point of view that 120one does not have to build a second kernel for capturing the dump. But 121at the same time one might want to build a custom dump capture kernel 122suitable to his needs. 123 124Following are the configuration setting required for system and 125dump-capture kernels for enabling kdump support. 126 127System kernel config options 128---------------------------- 129 1301) Enable "kexec system call" or "kexec file based system call" in 131 "Processor type and features.":: 132 133 CONFIG_KEXEC=y or CONFIG_KEXEC_FILE=y 134 135 And both of them will select KEXEC_CORE:: 136 137 CONFIG_KEXEC_CORE=y 138 139 Subsequently, CRASH_CORE is selected by KEXEC_CORE:: 140 141 CONFIG_CRASH_CORE=y 142 1432) Enable "sysfs file system support" in "Filesystem" -> "Pseudo 144 filesystems." This is usually enabled by default:: 145 146 CONFIG_SYSFS=y 147 148 Note that "sysfs file system support" might not appear in the "Pseudo 149 filesystems" menu if "Configure standard kernel features (expert users)" 150 is not enabled in "General Setup." In this case, check the .config file 151 itself to ensure that sysfs is turned on, as follows:: 152 153 grep 'CONFIG_SYSFS' .config 154 1553) Enable "Compile the kernel with debug info" in "Kernel hacking.":: 156 157 CONFIG_DEBUG_INFO=Y 158 159 This causes the kernel to be built with debug symbols. The dump 160 analysis tools require a vmlinux with debug symbols in order to read 161 and analyze a dump file. 162 163Dump-capture kernel config options (Arch Independent) 164----------------------------------------------------- 165 1661) Enable "kernel crash dumps" support under "Processor type and 167 features":: 168 169 CONFIG_CRASH_DUMP=y 170 1712) Enable "/proc/vmcore support" under "Filesystems" -> "Pseudo filesystems":: 172 173 CONFIG_PROC_VMCORE=y 174 175 (CONFIG_PROC_VMCORE is set by default when CONFIG_CRASH_DUMP is selected.) 176 177Dump-capture kernel config options (Arch Dependent, i386 and x86_64) 178-------------------------------------------------------------------- 179 1801) On i386, enable high memory support under "Processor type and 181 features":: 182 183 CONFIG_HIGHMEM64G=y 184 185 or:: 186 187 CONFIG_HIGHMEM4G 188 1892) With CONFIG_SMP=y, usually nr_cpus=1 need specified on the kernel 190 command line when loading the dump-capture kernel because one 191 CPU is enough for kdump kernel to dump vmcore on most of systems. 192 193 However, you can also specify nr_cpus=X to enable multiple processors 194 in kdump kernel. In this case, "disable_cpu_apicid=" is needed to 195 tell kdump kernel which cpu is 1st kernel's BSP. Please refer to 196 admin-guide/kernel-parameters.txt for more details. 197 198 With CONFIG_SMP=n, the above things are not related. 199 2003) A relocatable kernel is suggested to be built by default. If not yet, 201 enable "Build a relocatable kernel" support under "Processor type and 202 features":: 203 204 CONFIG_RELOCATABLE=y 205 2064) Use a suitable value for "Physical address where the kernel is 207 loaded" (under "Processor type and features"). This only appears when 208 "kernel crash dumps" is enabled. A suitable value depends upon 209 whether kernel is relocatable or not. 210 211 If you are using a relocatable kernel use CONFIG_PHYSICAL_START=0x100000 212 This will compile the kernel for physical address 1MB, but given the fact 213 kernel is relocatable, it can be run from any physical address hence 214 kexec boot loader will load it in memory region reserved for dump-capture 215 kernel. 216 217 Otherwise it should be the start of memory region reserved for 218 second kernel using boot parameter "crashkernel=Y@X". Here X is 219 start of memory region reserved for dump-capture kernel. 220 Generally X is 16MB (0x1000000). So you can set 221 CONFIG_PHYSICAL_START=0x1000000 222 2235) Make and install the kernel and its modules. DO NOT add this kernel 224 to the boot loader configuration files. 225 226Dump-capture kernel config options (Arch Dependent, ppc64) 227---------------------------------------------------------- 228 2291) Enable "Build a kdump crash kernel" support under "Kernel" options:: 230 231 CONFIG_CRASH_DUMP=y 232 2332) Enable "Build a relocatable kernel" support:: 234 235 CONFIG_RELOCATABLE=y 236 237 Make and install the kernel and its modules. 238 239Dump-capture kernel config options (Arch Dependent, arm) 240---------------------------------------------------------- 241 242- To use a relocatable kernel, 243 Enable "AUTO_ZRELADDR" support under "Boot" options:: 244 245 AUTO_ZRELADDR=y 246 247Dump-capture kernel config options (Arch Dependent, arm64) 248---------------------------------------------------------- 249 250- Please note that kvm of the dump-capture kernel will not be enabled 251 on non-VHE systems even if it is configured. This is because the CPU 252 will not be reset to EL2 on panic. 253 254crashkernel syntax 255=========================== 2561) crashkernel=size@offset 257 258 Here 'size' specifies how much memory to reserve for the dump-capture kernel 259 and 'offset' specifies the beginning of this reserved memory. For example, 260 "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory 261 starting at physical address 0x01000000 (16MB) for the dump-capture kernel. 262 263 The crashkernel region can be automatically placed by the system 264 kernel at run time. This is done by specifying the base address as 0, 265 or omitting it all together:: 266 267 crashkernel=256M@0 268 269 or:: 270 271 crashkernel=256M 272 273 If the start address is specified, note that the start address of the 274 kernel will be aligned to a value (which is Arch dependent), so if the 275 start address is not then any space below the alignment point will be 276 wasted. 277 2782) range1:size1[,range2:size2,...][@offset] 279 280 While the "crashkernel=size[@offset]" syntax is sufficient for most 281 configurations, sometimes it's handy to have the reserved memory dependent 282 on the value of System RAM -- that's mostly for distributors that pre-setup 283 the kernel command line to avoid a unbootable system after some memory has 284 been removed from the machine. 285 286 The syntax is:: 287 288 crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] 289 range=start-[end] 290 291 For example:: 292 293 crashkernel=512M-2G:64M,2G-:128M 294 295 This would mean: 296 297 1) if the RAM is smaller than 512M, then don't reserve anything 298 (this is the "rescue" case) 299 2) if the RAM size is between 512M and 2G (exclusive), then reserve 64M 300 3) if the RAM size is larger than 2G, then reserve 128M 301 3023) crashkernel=size,high and crashkernel=size,low 303 304 If memory above 4G is preferred, crashkernel=size,high can be used to 305 fulfill that. With it, physical memory is allowed to be allocated from top, 306 so could be above 4G if system has more than 4G RAM installed. Otherwise, 307 memory region will be allocated below 4G if available. 308 309 When crashkernel=X,high is passed, kernel could allocate physical memory 310 region above 4G, low memory under 4G is needed in this case. There are 311 three ways to get low memory: 312 313 1) Kernel will allocate at least 256M memory below 4G automatically 314 if crashkernel=Y,low is not specified. 315 2) Let user specify low memory size instead. 316 3) Specified value 0 will disable low memory allocation:: 317 318 crashkernel=0,low 319 320Boot into System Kernel 321----------------------- 3221) Update the boot loader (such as grub, yaboot, or lilo) configuration 323 files as necessary. 324 3252) Boot the system kernel with the boot parameter "crashkernel=Y@X". 326 327 On x86 and x86_64, use "crashkernel=Y[@X]". Most of the time, the 328 start address 'X' is not necessary, kernel will search a suitable 329 area. Unless an explicit start address is expected. 330 331 On ppc64, use "crashkernel=128M@32M". 332 333 On s390x, typically use "crashkernel=xxM". The value of xx is dependent 334 on the memory consumption of the kdump system. In general this is not 335 dependent on the memory size of the production system. 336 337 On arm, the use of "crashkernel=Y@X" is no longer necessary; the 338 kernel will automatically locate the crash kernel image within the 339 first 512MB of RAM if X is not given. 340 341 On arm64, use "crashkernel=Y[@X]". Note that the start address of 342 the kernel, X if explicitly specified, must be aligned to 2MiB (0x200000). 343 344Load the Dump-capture Kernel 345============================ 346 347After booting to the system kernel, dump-capture kernel needs to be 348loaded. 349 350Based on the architecture and type of image (relocatable or not), one 351can choose to load the uncompressed vmlinux or compressed bzImage/vmlinuz 352of dump-capture kernel. Following is the summary. 353 354For i386 and x86_64: 355 356 - Use bzImage/vmlinuz if kernel is relocatable. 357 - Use vmlinux if kernel is not relocatable. 358 359For ppc64: 360 361 - Use vmlinux 362 363For s390x: 364 365 - Use image or bzImage 366 367For arm: 368 369 - Use zImage 370 371For arm64: 372 373 - Use vmlinux or Image 374 375If you are using an uncompressed vmlinux image then use following command 376to load dump-capture kernel:: 377 378 kexec -p <dump-capture-kernel-vmlinux-image> \ 379 --initrd=<initrd-for-dump-capture-kernel> --args-linux \ 380 --append="root=<root-dev> <arch-specific-options>" 381 382If you are using a compressed bzImage/vmlinuz, then use following command 383to load dump-capture kernel:: 384 385 kexec -p <dump-capture-kernel-bzImage> \ 386 --initrd=<initrd-for-dump-capture-kernel> \ 387 --append="root=<root-dev> <arch-specific-options>" 388 389If you are using a compressed zImage, then use following command 390to load dump-capture kernel:: 391 392 kexec --type zImage -p <dump-capture-kernel-bzImage> \ 393 --initrd=<initrd-for-dump-capture-kernel> \ 394 --dtb=<dtb-for-dump-capture-kernel> \ 395 --append="root=<root-dev> <arch-specific-options>" 396 397If you are using an uncompressed Image, then use following command 398to load dump-capture kernel:: 399 400 kexec -p <dump-capture-kernel-Image> \ 401 --initrd=<initrd-for-dump-capture-kernel> \ 402 --append="root=<root-dev> <arch-specific-options>" 403 404Following are the arch specific command line options to be used while 405loading dump-capture kernel. 406 407For i386 and x86_64: 408 409 "1 irqpoll nr_cpus=1 reset_devices" 410 411For ppc64: 412 413 "1 maxcpus=1 noirqdistrib reset_devices" 414 415For s390x: 416 417 "1 nr_cpus=1 cgroup_disable=memory" 418 419For arm: 420 421 "1 maxcpus=1 reset_devices" 422 423For arm64: 424 425 "1 nr_cpus=1 reset_devices" 426 427Notes on loading the dump-capture kernel: 428 429* By default, the ELF headers are stored in ELF64 format to support 430 systems with more than 4GB memory. On i386, kexec automatically checks if 431 the physical RAM size exceeds the 4 GB limit and if not, uses ELF32. 432 So, on non-PAE systems, ELF32 is always used. 433 434 The --elf32-core-headers option can be used to force the generation of ELF32 435 headers. This is necessary because GDB currently cannot open vmcore files 436 with ELF64 headers on 32-bit systems. 437 438* The "irqpoll" boot parameter reduces driver initialization failures 439 due to shared interrupts in the dump-capture kernel. 440 441* You must specify <root-dev> in the format corresponding to the root 442 device name in the output of mount command. 443 444* Boot parameter "1" boots the dump-capture kernel into single-user 445 mode without networking. If you want networking, use "3". 446 447* We generally don't have to bring up a SMP kernel just to capture the 448 dump. Hence generally it is useful either to build a UP dump-capture 449 kernel or specify maxcpus=1 option while loading dump-capture kernel. 450 Note, though maxcpus always works, you had better replace it with 451 nr_cpus to save memory if supported by the current ARCH, such as x86. 452 453* You should enable multi-cpu support in dump-capture kernel if you intend 454 to use multi-thread programs with it, such as parallel dump feature of 455 makedumpfile. Otherwise, the multi-thread program may have a great 456 performance degradation. To enable multi-cpu support, you should bring up an 457 SMP dump-capture kernel and specify maxcpus/nr_cpus, disable_cpu_apicid=[X] 458 options while loading it. 459 460* For s390x there are two kdump modes: If a ELF header is specified with 461 the elfcorehdr= kernel parameter, it is used by the kdump kernel as it 462 is done on all other architectures. If no elfcorehdr= kernel parameter is 463 specified, the s390x kdump kernel dynamically creates the header. The 464 second mode has the advantage that for CPU and memory hotplug, kdump has 465 not to be reloaded with kexec_load(). 466 467* For s390x systems with many attached devices the "cio_ignore" kernel 468 parameter should be used for the kdump kernel in order to prevent allocation 469 of kernel memory for devices that are not relevant for kdump. The same 470 applies to systems that use SCSI/FCP devices. In that case the 471 "allow_lun_scan" zfcp module parameter should be set to zero before 472 setting FCP devices online. 473 474Kernel Panic 475============ 476 477After successfully loading the dump-capture kernel as previously 478described, the system will reboot into the dump-capture kernel if a 479system crash is triggered. Trigger points are located in panic(), 480die(), die_nmi() and in the sysrq handler (ALT-SysRq-c). 481 482The following conditions will execute a crash trigger point: 483 484If a hard lockup is detected and "NMI watchdog" is configured, the system 485will boot into the dump-capture kernel ( die_nmi() ). 486 487If die() is called, and it happens to be a thread with pid 0 or 1, or die() 488is called inside interrupt context or die() is called and panic_on_oops is set, 489the system will boot into the dump-capture kernel. 490 491On powerpc systems when a soft-reset is generated, die() is called by all cpus 492and the system will boot into the dump-capture kernel. 493 494For testing purposes, you can trigger a crash by using "ALT-SysRq-c", 495"echo c > /proc/sysrq-trigger" or write a module to force the panic. 496 497Write Out the Dump File 498======================= 499 500After the dump-capture kernel is booted, write out the dump file with 501the following command:: 502 503 cp /proc/vmcore <dump-file> 504 505or use scp to write out the dump file between hosts on a network, e.g:: 506 507 scp /proc/vmcore remote_username@remote_ip:<dump-file> 508 509You can also use makedumpfile utility to write out the dump file 510with specified options to filter out unwanted contents, e.g:: 511 512 makedumpfile -l --message-level 1 -d 31 /proc/vmcore <dump-file> 513 514Analysis 515======== 516 517Before analyzing the dump image, you should reboot into a stable kernel. 518 519You can do limited analysis using GDB on the dump file copied out of 520/proc/vmcore. Use the debug vmlinux built with -g and run the following 521command:: 522 523 gdb vmlinux <dump-file> 524 525Stack trace for the task on processor 0, register display, and memory 526display work fine. 527 528Note: GDB cannot analyze core files generated in ELF64 format for x86. 529On systems with a maximum of 4GB of memory, you can generate 530ELF32-format headers using the --elf32-core-headers kernel option on the 531dump kernel. 532 533You can also use the Crash utility to analyze dump files in Kdump 534format. Crash is available at the following URL: 535 536 https://github.com/crash-utility/crash 537 538Crash document can be found at: 539 https://crash-utility.github.io/ 540 541Trigger Kdump on WARN() 542======================= 543 544The kernel parameter, panic_on_warn, calls panic() in all WARN() paths. This 545will cause a kdump to occur at the panic() call. In cases where a user wants 546to specify this during runtime, /proc/sys/kernel/panic_on_warn can be set to 1 547to achieve the same behaviour. 548 549Trigger Kdump on add_taint() 550============================ 551 552The kernel parameter panic_on_taint facilitates a conditional call to panic() 553from within add_taint() whenever the value set in this bitmask matches with the 554bit flag being set by add_taint(). 555This will cause a kdump to occur at the add_taint()->panic() call. 556 557Contact 558======= 559 560- kexec@lists.infradead.org 561 562GDB macros 563========== 564 565.. include:: gdbmacros.txt 566 :literal: 567