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 1392) Enable "sysfs file system support" in "Filesystem" -> "Pseudo 140 filesystems." This is usually enabled by default:: 141 142 CONFIG_SYSFS=y 143 144 Note that "sysfs file system support" might not appear in the "Pseudo 145 filesystems" menu if "Configure standard kernel features (expert users)" 146 is not enabled in "General Setup." In this case, check the .config file 147 itself to ensure that sysfs is turned on, as follows:: 148 149 grep 'CONFIG_SYSFS' .config 150 1513) Enable "Compile the kernel with debug info" in "Kernel hacking.":: 152 153 CONFIG_DEBUG_INFO=Y 154 155 This causes the kernel to be built with debug symbols. The dump 156 analysis tools require a vmlinux with debug symbols in order to read 157 and analyze a dump file. 158 159Dump-capture kernel config options (Arch Independent) 160----------------------------------------------------- 161 1621) Enable "kernel crash dumps" support under "Processor type and 163 features":: 164 165 CONFIG_CRASH_DUMP=y 166 167 And this will select VMCORE_INFO and CRASH_RESERVE:: 168 CONFIG_VMCORE_INFO=y 169 CONFIG_CRASH_RESERVE=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_HIGHMEM4G 184 1852) With CONFIG_SMP=y, usually nr_cpus=1 need specified on the kernel 186 command line when loading the dump-capture kernel because one 187 CPU is enough for kdump kernel to dump vmcore on most of systems. 188 189 However, you can also specify nr_cpus=X to enable multiple processors 190 in kdump kernel. 191 192 With CONFIG_SMP=n, the above things are not related. 193 1943) A relocatable kernel is suggested to be built by default. If not yet, 195 enable "Build a relocatable kernel" support under "Processor type and 196 features":: 197 198 CONFIG_RELOCATABLE=y 199 2004) Use a suitable value for "Physical address where the kernel is 201 loaded" (under "Processor type and features"). This only appears when 202 "kernel crash dumps" is enabled. A suitable value depends upon 203 whether kernel is relocatable or not. 204 205 If you are using a relocatable kernel use CONFIG_PHYSICAL_START=0x100000 206 This will compile the kernel for physical address 1MB, but given the fact 207 kernel is relocatable, it can be run from any physical address hence 208 kexec boot loader will load it in memory region reserved for dump-capture 209 kernel. 210 211 Otherwise it should be the start of memory region reserved for 212 second kernel using boot parameter "crashkernel=Y@X". Here X is 213 start of memory region reserved for dump-capture kernel. 214 Generally X is 16MB (0x1000000). So you can set 215 CONFIG_PHYSICAL_START=0x1000000 216 2175) Make and install the kernel and its modules. DO NOT add this kernel 218 to the boot loader configuration files. 219 220Dump-capture kernel config options (Arch Dependent, ppc64) 221---------------------------------------------------------- 222 2231) Enable "Build a kdump crash kernel" support under "Kernel" options:: 224 225 CONFIG_CRASH_DUMP=y 226 2272) Enable "Build a relocatable kernel" support:: 228 229 CONFIG_RELOCATABLE=y 230 231 Make and install the kernel and its modules. 232 233Dump-capture kernel config options (Arch Dependent, arm) 234---------------------------------------------------------- 235 236- To use a relocatable kernel, 237 Enable "AUTO_ZRELADDR" support under "Boot" options:: 238 239 AUTO_ZRELADDR=y 240 241Dump-capture kernel config options (Arch Dependent, arm64) 242---------------------------------------------------------- 243 244- Please note that kvm of the dump-capture kernel will not be enabled 245 on non-VHE systems even if it is configured. This is because the CPU 246 will not be reset to EL2 on panic. 247 248crashkernel syntax 249=========================== 2501) crashkernel=size@offset 251 252 Here 'size' specifies how much memory to reserve for the dump-capture kernel 253 and 'offset' specifies the beginning of this reserved memory. For example, 254 "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory 255 starting at physical address 0x01000000 (16MB) for the dump-capture kernel. 256 257 The crashkernel region can be automatically placed by the system 258 kernel at run time. This is done by specifying the base address as 0, 259 or omitting it all together:: 260 261 crashkernel=256M@0 262 263 or:: 264 265 crashkernel=256M 266 267 If the start address is specified, note that the start address of the 268 kernel will be aligned to a value (which is Arch dependent), so if the 269 start address is not then any space below the alignment point will be 270 wasted. 271 2722) range1:size1[,range2:size2,...][@offset] 273 274 While the "crashkernel=size[@offset]" syntax is sufficient for most 275 configurations, sometimes it's handy to have the reserved memory dependent 276 on the value of System RAM -- that's mostly for distributors that pre-setup 277 the kernel command line to avoid a unbootable system after some memory has 278 been removed from the machine. 279 280 The syntax is:: 281 282 crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] 283 range=start-[end] 284 285 For example:: 286 287 crashkernel=512M-2G:64M,2G-:128M 288 289 This would mean: 290 291 1) if the RAM is smaller than 512M, then don't reserve anything 292 (this is the "rescue" case) 293 2) if the RAM size is between 512M and 2G (exclusive), then reserve 64M 294 3) if the RAM size is larger than 2G, then reserve 128M 295 2963) crashkernel=size,high and crashkernel=size,low 297 298 If memory above 4G is preferred, crashkernel=size,high can be used to 299 fulfill that. With it, physical memory is allowed to be allocated from top, 300 so could be above 4G if system has more than 4G RAM installed. Otherwise, 301 memory region will be allocated below 4G if available. 302 303 When crashkernel=X,high is passed, kernel could allocate physical memory 304 region above 4G, low memory under 4G is needed in this case. There are 305 three ways to get low memory: 306 307 1) Kernel will allocate at least 256M memory below 4G automatically 308 if crashkernel=Y,low is not specified. 309 2) Let user specify low memory size instead. 310 3) Specified value 0 will disable low memory allocation:: 311 312 crashkernel=0,low 313 3144) crashkernel=size,cma 315 316 Reserve additional crash kernel memory from CMA. This reservation is 317 usable by the first system's userspace memory and kernel movable 318 allocations (memory balloon, zswap). Pages allocated from this memory 319 range will not be included in the vmcore so this should not be used if 320 dumping of userspace memory is intended and it has to be expected that 321 some movable kernel pages may be missing from the dump. 322 323 A standard crashkernel reservation, as described above, is still needed 324 to hold the crash kernel and initrd. 325 326 This option increases the risk of a kdump failure: DMA transfers 327 configured by the first kernel may end up corrupting the second 328 kernel's memory. 329 330 This reservation method is intended for systems that can't afford to 331 sacrifice enough memory for standard crashkernel reservation and where 332 less reliable and possibly incomplete kdump is preferable to no kdump at 333 all. 334 335Boot into System Kernel 336----------------------- 3371) Update the boot loader (such as grub, yaboot, or lilo) configuration 338 files as necessary. 339 3402) Boot the system kernel with the boot parameter "crashkernel=Y@X". 341 342 On x86 and x86_64, use "crashkernel=Y[@X]". Most of the time, the 343 start address 'X' is not necessary, kernel will search a suitable 344 area. Unless an explicit start address is expected. 345 346 On ppc64, use "crashkernel=128M@32M". 347 348 On s390x, typically use "crashkernel=xxM". The value of xx is dependent 349 on the memory consumption of the kdump system. In general this is not 350 dependent on the memory size of the production system. 351 352 On arm, the use of "crashkernel=Y@X" is no longer necessary; the 353 kernel will automatically locate the crash kernel image within the 354 first 512MB of RAM if X is not given. 355 356 On arm64, use "crashkernel=Y[@X]". Note that the start address of 357 the kernel, X if explicitly specified, must be aligned to 2MiB (0x200000). 358 359Load the Dump-capture Kernel 360============================ 361 362After booting to the system kernel, dump-capture kernel needs to be 363loaded. 364 365Based on the architecture and type of image (relocatable or not), one 366can choose to load the uncompressed vmlinux or compressed bzImage/vmlinuz 367of dump-capture kernel. Following is the summary. 368 369For i386 and x86_64: 370 371 - Use bzImage/vmlinuz if kernel is relocatable. 372 - Use vmlinux if kernel is not relocatable. 373 374For ppc64: 375 376 - Use vmlinux 377 378For s390x: 379 380 - Use image or bzImage 381 382For arm: 383 384 - Use zImage 385 386For arm64: 387 388 - Use vmlinux or Image 389 390If you are using an uncompressed vmlinux image then use following command 391to load dump-capture kernel:: 392 393 kexec -p <dump-capture-kernel-vmlinux-image> \ 394 --initrd=<initrd-for-dump-capture-kernel> --args-linux \ 395 --append="root=<root-dev> <arch-specific-options>" 396 397If you are using a compressed bzImage/vmlinuz, then use following command 398to load dump-capture kernel:: 399 400 kexec -p <dump-capture-kernel-bzImage> \ 401 --initrd=<initrd-for-dump-capture-kernel> \ 402 --append="root=<root-dev> <arch-specific-options>" 403 404If you are using a compressed zImage, then use following command 405to load dump-capture kernel:: 406 407 kexec --type zImage -p <dump-capture-kernel-bzImage> \ 408 --initrd=<initrd-for-dump-capture-kernel> \ 409 --dtb=<dtb-for-dump-capture-kernel> \ 410 --append="root=<root-dev> <arch-specific-options>" 411 412If you are using an uncompressed Image, then use following command 413to load dump-capture kernel:: 414 415 kexec -p <dump-capture-kernel-Image> \ 416 --initrd=<initrd-for-dump-capture-kernel> \ 417 --append="root=<root-dev> <arch-specific-options>" 418 419Following are the arch specific command line options to be used while 420loading dump-capture kernel. 421 422For i386 and x86_64: 423 424 "1 irqpoll nr_cpus=1 reset_devices" 425 426For ppc64: 427 428 "1 maxcpus=1 noirqdistrib reset_devices" 429 430For s390x: 431 432 "1 nr_cpus=1 cgroup_disable=memory" 433 434For arm: 435 436 "1 maxcpus=1 reset_devices" 437 438For arm64: 439 440 "1 nr_cpus=1 reset_devices" 441 442Notes on loading the dump-capture kernel: 443 444* By default, the ELF headers are stored in ELF64 format to support 445 systems with more than 4GB memory. On i386, kexec automatically checks if 446 the physical RAM size exceeds the 4 GB limit and if not, uses ELF32. 447 So, on non-PAE systems, ELF32 is always used. 448 449 The --elf32-core-headers option can be used to force the generation of ELF32 450 headers. This is necessary because GDB currently cannot open vmcore files 451 with ELF64 headers on 32-bit systems. 452 453* The "irqpoll" boot parameter reduces driver initialization failures 454 due to shared interrupts in the dump-capture kernel. 455 456* You must specify <root-dev> in the format corresponding to the root 457 device name in the output of mount command. 458 459* Boot parameter "1" boots the dump-capture kernel into single-user 460 mode without networking. If you want networking, use "3". 461 462* We generally don't have to bring up a SMP kernel just to capture the 463 dump. Hence generally it is useful either to build a UP dump-capture 464 kernel or specify maxcpus=1 option while loading dump-capture kernel. 465 Note, though maxcpus always works, you had better replace it with 466 nr_cpus to save memory if supported by the current ARCH, such as x86. 467 468* You should enable multi-cpu support in dump-capture kernel if you intend 469 to use multi-thread programs with it, such as parallel dump feature of 470 makedumpfile. Otherwise, the multi-thread program may have a great 471 performance degradation. To enable multi-cpu support, you should bring up an 472 SMP dump-capture kernel and specify maxcpus/nr_cpus options while loading it. 473 474* For s390x there are two kdump modes: If a ELF header is specified with 475 the elfcorehdr= kernel parameter, it is used by the kdump kernel as it 476 is done on all other architectures. If no elfcorehdr= kernel parameter is 477 specified, the s390x kdump kernel dynamically creates the header. The 478 second mode has the advantage that for CPU and memory hotplug, kdump has 479 not to be reloaded with kexec_load(). 480 481* For s390x systems with many attached devices the "cio_ignore" kernel 482 parameter should be used for the kdump kernel in order to prevent allocation 483 of kernel memory for devices that are not relevant for kdump. The same 484 applies to systems that use SCSI/FCP devices. In that case the 485 "allow_lun_scan" zfcp module parameter should be set to zero before 486 setting FCP devices online. 487 488Kernel Panic 489============ 490 491After successfully loading the dump-capture kernel as previously 492described, the system will reboot into the dump-capture kernel if a 493system crash is triggered. Trigger points are located in panic(), 494die(), die_nmi() and in the sysrq handler (ALT-SysRq-c). 495 496The following conditions will execute a crash trigger point: 497 498If a hard lockup is detected and "NMI watchdog" is configured, the system 499will boot into the dump-capture kernel ( die_nmi() ). 500 501If die() is called, and it happens to be a thread with pid 0 or 1, or die() 502is called inside interrupt context or die() is called and panic_on_oops is set, 503the system will boot into the dump-capture kernel. 504 505On powerpc systems when a soft-reset is generated, die() is called by all cpus 506and the system will boot into the dump-capture kernel. 507 508For testing purposes, you can trigger a crash by using "ALT-SysRq-c", 509"echo c > /proc/sysrq-trigger" or write a module to force the panic. 510 511Write Out the Dump File 512======================= 513 514After the dump-capture kernel is booted, write out the dump file with 515the following command:: 516 517 cp /proc/vmcore <dump-file> 518 519or use scp to write out the dump file between hosts on a network, e.g:: 520 521 scp /proc/vmcore remote_username@remote_ip:<dump-file> 522 523You can also use makedumpfile utility to write out the dump file 524with specified options to filter out unwanted contents, e.g:: 525 526 makedumpfile -l --message-level 1 -d 31 /proc/vmcore <dump-file> 527 528Analysis 529======== 530 531Before analyzing the dump image, you should reboot into a stable kernel. 532 533You can do limited analysis using GDB on the dump file copied out of 534/proc/vmcore. Use the debug vmlinux built with -g and run the following 535command:: 536 537 gdb vmlinux <dump-file> 538 539Stack trace for the task on processor 0, register display, and memory 540display work fine. 541 542Note: GDB cannot analyze core files generated in ELF64 format for x86. 543On systems with a maximum of 4GB of memory, you can generate 544ELF32-format headers using the --elf32-core-headers kernel option on the 545dump kernel. 546 547You can also use the Crash utility to analyze dump files in Kdump 548format. Crash is available at the following URL: 549 550 https://github.com/crash-utility/crash 551 552Crash document can be found at: 553 https://crash-utility.github.io/ 554 555Trigger Kdump on WARN() 556======================= 557 558The kernel parameter, panic_on_warn, calls panic() in all WARN() paths. This 559will cause a kdump to occur at the panic() call. In cases where a user wants 560to specify this during runtime, /proc/sys/kernel/panic_on_warn can be set to 1 561to achieve the same behaviour. 562 563Trigger Kdump on add_taint() 564============================ 565 566The kernel parameter panic_on_taint facilitates a conditional call to panic() 567from within add_taint() whenever the value set in this bitmask matches with the 568bit flag being set by add_taint(). 569This will cause a kdump to occur at the add_taint()->panic() call. 570 571Write the dump file to encrypted disk volume 572============================================ 573 574CONFIG_CRASH_DM_CRYPT can be enabled to support saving the dump file to an 575encrypted disk volume (only x86_64 supported for now). User space can interact 576with /sys/kernel/config/crash_dm_crypt_keys for setup, 577 5781. Tell the first kernel what logon keys are needed to unlock the disk volumes, 579 # Add key #1 580 mkdir /sys/kernel/config/crash_dm_crypt_keys/7d26b7b4-e342-4d2d-b660-7426b0996720 581 # Add key #1's description 582 echo cryptsetup:7d26b7b4-e342-4d2d-b660-7426b0996720 > /sys/kernel/config/crash_dm_crypt_keys/description 583 584 # how many keys do we have now? 585 cat /sys/kernel/config/crash_dm_crypt_keys/count 586 1 587 588 # Add key #2 in the same way 589 590 # how many keys do we have now? 591 cat /sys/kernel/config/crash_dm_crypt_keys/count 592 2 593 594 # To support CPU/memory hot-plugging, re-use keys already saved to reserved 595 # memory 596 echo true > /sys/kernel/config/crash_dm_crypt_key/reuse 597 5982. Load the dump-capture kernel 599 6003. After the dump-capture kerne get booted, restore the keys to user keyring 601 echo yes > /sys/kernel/crash_dm_crypt_keys/restore 602 603Contact 604======= 605 606- kexec@lists.infradead.org 607 608GDB macros 609========== 610 611.. include:: gdbmacros.txt 612 :literal: 613