Documentation for Kdump - The kexec-based Crash Dumping Solution

This document includes overview, setup, installation, and analysisinformation.

Overview

Kdump uses kexec to quickly boot to a dump-capture kernel whenever adump of the system kernel’s memory needs to be taken (for example, whenthe system panics). The system kernel’s memory image is preserved acrossthe reboot and is accessible to the dump-capture kernel.

You can use common commands, such as cp, scp or makedumpfile to copythe memory image to a dump file on the local disk, or across the networkto a remote system.

Kdump and kexec are currently supported on the x86, x86_64, ppc64,s390x, arm and arm64 architectures.

When the system kernel boots, it reserves a small section of memory forthe dump-capture kernel. This ensures that ongoing Direct Memory Access(DMA) from the system kernel does not corrupt the dump-capture kernel.The kexec -p command loads the dump-capture kernel into this reservedmemory.

On x86 machines, the first 640 KB of physical memory is needed for boot,regardless of where the kernel loads. For simpler handling, the wholelow 1M is reserved to avoid any later kernel or device driver writingdata into this area. Like this, the low 1M can be reused as system RAMby kdump kernel without extra handling.

On PPC64 machines first 32KB of physical memory is needed for bootingregardless of where the kernel is loaded and to support 64K page sizekexec backs up the first 64KB memory.

For s390x, when kdump is triggered, the crashkernel region is exchangedwith the region [0, crashkernel region size] and then the kdump kernelruns in [0, crashkernel region size]. Therefore no relocatable kernel isneeded for s390x.

All of the necessary information about the system kernel’s core image isencoded in the ELF format, and stored in a reserved area of memorybefore a crash. The physical address of the start of the ELF header ispassed to the dump-capture kernel through the elfcorehdr= bootparameter. Optionally the size of the ELF header can also be passedwhen using the elfcorehdr=[size[KMG]@]offset[KMG] syntax.

With the dump-capture kernel, you can access the memory image through/proc/vmcore. This exports the dump as an ELF-format file that you canwrite out using file copy commands such as cp or scp. You can also usemakedumpfile utility to analyze and write out filtered contents withoptions, e.g with ‘-d 31’ it will only write out kernel data. Further,you can use analysis tools such as the GNU Debugger (GDB) and the Crashtool to debug the dump file. This method ensures that the dump pages arecorrectly ordered.

Setup and Installation

Install kexec-tools

  1. Login as the root user.

  2. Download the kexec-tools user-space package from the following URL:

http://kernel.org/pub/linux/utils/kernel/kexec/kexec-tools.tar.gz

This is a symlink to the latest version.

The latest kexec-tools git tree is available at:

There is also a gitweb interface available athttp://www.kernel.org/git/?p=utils/kernel/kexec/kexec-tools.git

More information about kexec-tools can be found athttp://horms.net/projects/kexec/

  1. Unpack the tarball with the tar command, as follows:

    tar xvpzf kexec-tools.tar.gz

  2. Change to the kexec-tools directory, as follows:

    cd kexec-tools-VERSION

  3. Configure the package, as follows:

    ./configure

  4. Compile the package, as follows:

    make

  5. Install the package, as follows:

    make install

Build the system and dump-capture kernels

There are two possible methods of using Kdump.

  1. Build a separate custom dump-capture kernel for capturing thekernel core dump.

  2. Or use the system kernel binary itself as dump-capture kernel and there isno need to build a separate dump-capture kernel. This is possibleonly with the architectures which support a relocatable kernel. Asof today, i386, x86_64, ppc64, arm and arm64 architectures supportrelocatable kernel.

Building a relocatable kernel is advantageous from the point of view thatone does not have to build a second kernel for capturing the dump. Butat the same time one might want to build a custom dump capture kernelsuitable to his needs.

Following are the configuration setting required for system anddump-capture kernels for enabling kdump support.

System kernel config options

  1. Enable “kexec system call” or “kexec file based system call” in“Processor type and features.”:

    CONFIG_KEXEC=y or CONFIG_KEXEC_FILE=y

    And both of them will select KEXEC_CORE:

    CONFIG_KEXEC_CORE=y

  2. Enable “sysfs file system support” in “Filesystem” -> “Pseudofilesystems.” This is usually enabled by default:

    CONFIG_SYSFS=y

    Note that “sysfs file system support” might not appear in the “Pseudofilesystems” menu if “Configure standard kernel features (expert users)”is not enabled in “General Setup.” In this case, check the .config fileitself to ensure that sysfs is turned on, as follows:

    grep 'CONFIG_SYSFS' .config

  3. Enable “Compile the kernel with debug info” in “Kernel hacking.”:

    CONFIG_DEBUG_INFO=Y

    This causes the kernel to be built with debug symbols. The dumpanalysis tools require a vmlinux with debug symbols in order to readand analyze a dump file.

Dump-capture kernel config options (Arch Independent)

  1. Enable “kernel crash dumps” support under “Processor type andfeatures”:

    CONFIG_CRASH_DUMP=y
    And this will select VMCORE_INFO and CRASH_RESERVE::

    CONFIG_VMCORE_INFO=yCONFIG_CRASH_RESERVE=y

  2. Enable “/proc/vmcore support” under “Filesystems” -> “Pseudo filesystems”:

    CONFIG_PROC_VMCORE=y

    (CONFIG_PROC_VMCORE is set by default when CONFIG_CRASH_DUMP is selected.)

Dump-capture kernel config options (Arch Dependent, i386 and x86_64)

  1. On i386, enable high memory support under “Processor type andfeatures”:

    CONFIG_HIGHMEM4G

  2. With CONFIG_SMP=y, usually nr_cpus=1 need specified on the kernelcommand line when loading the dump-capture kernel because oneCPU is enough for kdump kernel to dump vmcore on most of systems.

    However, you can also specify nr_cpus=X to enable multiple processorsin kdump kernel.

    With CONFIG_SMP=n, the above things are not related.

  3. A relocatable kernel is suggested to be built by default. If not yet,enable “Build a relocatable kernel” support under “Processor type andfeatures”:

    CONFIG_RELOCATABLE=y

  4. Use a suitable value for “Physical address where the kernel isloaded” (under “Processor type and features”). This only appears when“kernel crash dumps” is enabled. A suitable value depends uponwhether kernel is relocatable or not.

    If you are using a relocatable kernel use CONFIG_PHYSICAL_START=0x100000This will compile the kernel for physical address 1MB, but given the factkernel is relocatable, it can be run from any physical address hencekexec boot loader will load it in memory region reserved for dump-capturekernel.

    Otherwise it should be the start of memory region reserved forsecond kernel using boot parameter “crashkernel=Y@X”. Here X isstart of memory region reserved for dump-capture kernel.Generally X is 16MB (0x1000000). So you can setCONFIG_PHYSICAL_START=0x1000000

  5. Make and install the kernel and its modules. DO NOT add this kernelto the boot loader configuration files.

Dump-capture kernel config options (Arch Dependent, ppc64)

  1. Enable “Build a kdump crash kernel” support under “Kernel” options:

    CONFIG_CRASH_DUMP=y

  2. Enable “Build a relocatable kernel” support:

    CONFIG_RELOCATABLE=y

Make and install the kernel and its modules.

Dump-capture kernel config options (Arch Dependent, arm)

  • To use a relocatable kernel,Enable “AUTO_ZRELADDR” support under “Boot” options:

    AUTO_ZRELADDR=y

Dump-capture kernel config options (Arch Dependent, arm64)

  • Please note that kvm of the dump-capture kernel will not be enabledon non-VHE systems even if it is configured. This is because the CPUwill not be reset to EL2 on panic.

crashkernel syntax

  1. crashkernel=size@offset

    Here ‘size’ specifies how much memory to reserve for the dump-capture kerneland ‘offset’ specifies the beginning of this reserved memory. For example,“crashkernel=64M@16M” tells the system kernel to reserve 64 MB of memorystarting at physical address 0x01000000 (16MB) for the dump-capture kernel.

    The crashkernel region can be automatically placed by the systemkernel at run time. This is done by specifying the base address as 0,or omitting it all together:

    crashkernel=256M@0

    or:

    crashkernel=256M

    If the start address is specified, note that the start address of thekernel will be aligned to a value (which is Arch dependent), so if thestart address is not then any space below the alignment point will bewasted.

  2. range1:size1[,range2:size2,...][@offset]

    While the “crashkernel=size[@offset]” syntax is sufficient for mostconfigurations, sometimes it’s handy to have the reserved memory dependenton the value of System RAM -- that’s mostly for distributors that pre-setupthe kernel command line to avoid a unbootable system after some memory hasbeen removed from the machine.

    The syntax is:

    crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset]range=start-[end]

    For example:

    crashkernel=512M-2G:64M,2G-:128M

    This would mean:

    1. if the RAM is smaller than 512M, then don’t reserve anything(this is the “rescue” case)

    2. if the RAM size is between 512M and 2G (exclusive), then reserve 64M

    3. if the RAM size is larger than 2G, then reserve 128M

  3. crashkernel=size,high and crashkernel=size,low

    If memory above 4G is preferred, crashkernel=size,high can be used tofulfill that. With it, physical memory is allowed to be allocated from top,so could be above 4G if system has more than 4G RAM installed. Otherwise,memory region will be allocated below 4G if available.

    When crashkernel=X,high is passed, kernel could allocate physical memoryregion above 4G, low memory under 4G is needed in this case. There arethree ways to get low memory:

    1. Kernel will allocate at least 256M memory below 4G automaticallyif crashkernel=Y,low is not specified.

    2. Let user specify low memory size instead.

    3. Specified value 0 will disable low memory allocation:

      crashkernel=0,low

  4. crashkernel=size,cma

    Reserve additional crash kernel memory from CMA. This reservation isusable by the first system’s userspace memory and kernel movableallocations (memory balloon, zswap). Pages allocated from this memoryrange will not be included in the vmcore so this should not be used ifdumping of userspace memory is intended and it has to be expected thatsome movable kernel pages may be missing from the dump.

    A standard crashkernel reservation, as described above, is still neededto hold the crash kernel and initrd.

    This option increases the risk of a kdump failure: DMA transfersconfigured by the first kernel may end up corrupting the secondkernel’s memory.

    This reservation method is intended for systems that can’t afford tosacrifice enough memory for standard crashkernel reservation and whereless reliable and possibly incomplete kdump is preferable to no kdump atall.

Boot into System Kernel

  1. Update the boot loader (such as grub, yaboot, or lilo) configurationfiles as necessary.

  2. Boot the system kernel with the boot parameter “crashkernel=Y@X”.

    On x86 and x86_64, use “crashkernel=Y[@X]”. Most of the time, thestart address ‘X’ is not necessary, kernel will search a suitablearea. Unless an explicit start address is expected.

    On ppc64, use “crashkernel=128M@32M”.

    On s390x, typically use “crashkernel=xxM”. The value of xx is dependenton the memory consumption of the kdump system. In general this is notdependent on the memory size of the production system.

    On arm, the use of “crashkernel=Y@X” is no longer necessary; thekernel will automatically locate the crash kernel image within thefirst 512MB of RAM if X is not given.

    On arm64, use “crashkernel=Y[@X]”. Note that the start address ofthe kernel, X if explicitly specified, must be aligned to 2MiB (0x200000).

Load the Dump-capture Kernel

After booting to the system kernel, dump-capture kernel needs to beloaded.

Based on the architecture and type of image (relocatable or not), onecan choose to load the uncompressed vmlinux or compressed bzImage/vmlinuzof dump-capture kernel. Following is the summary.

For i386 and x86_64:

  • Use bzImage/vmlinuz if kernel is relocatable.

  • Use vmlinux if kernel is not relocatable.

For ppc64:

  • Use vmlinux

For s390x:

  • Use image or bzImage

For arm:

  • Use zImage

For arm64:

  • Use vmlinux or Image

If you are using an uncompressed vmlinux image then use following commandto load dump-capture kernel:

kexec -p <dump-capture-kernel-vmlinux-image> \--initrd=<initrd-for-dump-capture-kernel> --args-linux \--append="root=<root-dev> <arch-specific-options>"

If you are using a compressed bzImage/vmlinuz, then use following commandto load dump-capture kernel:

kexec -p <dump-capture-kernel-bzImage> \--initrd=<initrd-for-dump-capture-kernel> \--append="root=<root-dev> <arch-specific-options>"

If you are using a compressed zImage, then use following commandto load dump-capture kernel:

kexec --type zImage -p <dump-capture-kernel-bzImage> \--initrd=<initrd-for-dump-capture-kernel> \--dtb=<dtb-for-dump-capture-kernel> \--append="root=<root-dev> <arch-specific-options>"

If you are using an uncompressed Image, then use following commandto load dump-capture kernel:

kexec -p <dump-capture-kernel-Image> \--initrd=<initrd-for-dump-capture-kernel> \--append="root=<root-dev> <arch-specific-options>"

Following are the arch specific command line options to be used whileloading dump-capture kernel.

For i386 and x86_64:

“1 irqpoll nr_cpus=1 reset_devices”

For ppc64:

“1 maxcpus=1 noirqdistrib reset_devices”

For s390x:

“1 nr_cpus=1 cgroup_disable=memory”

For arm:

“1 maxcpus=1 reset_devices”

For arm64:

“1 nr_cpus=1 reset_devices”

Notes on loading the dump-capture kernel:

  • By default, the ELF headers are stored in ELF64 format to supportsystems with more than 4GB memory. On i386, kexec automatically checks ifthe physical RAM size exceeds the 4 GB limit and if not, uses ELF32.So, on non-PAE systems, ELF32 is always used.

    The --elf32-core-headers option can be used to force the generation of ELF32headers. This is necessary because GDB currently cannot open vmcore fileswith ELF64 headers on 32-bit systems.

  • The “irqpoll” boot parameter reduces driver initialization failuresdue to shared interrupts in the dump-capture kernel.

  • You must specify <root-dev> in the format corresponding to the rootdevice name in the output of mount command.

  • Boot parameter “1” boots the dump-capture kernel into single-usermode without networking. If you want networking, use “3”.

  • We generally don’t have to bring up a SMP kernel just to capture thedump. Hence generally it is useful either to build a UP dump-capturekernel or specify maxcpus=1 option while loading dump-capture kernel.Note, though maxcpus always works, you had better replace it withnr_cpus to save memory if supported by the current ARCH, such as x86.

  • You should enable multi-cpu support in dump-capture kernel if you intendto use multi-thread programs with it, such as parallel dump feature ofmakedumpfile. Otherwise, the multi-thread program may have a greatperformance degradation. To enable multi-cpu support, you should bring up anSMP dump-capture kernel and specify maxcpus/nr_cpus options while loading it.

  • For s390x there are two kdump modes: If an ELF header is specified withthe elfcorehdr= kernel parameter, it is used by the kdump kernel as itis done on all other architectures. If no elfcorehdr= kernel parameter isspecified, the s390x kdump kernel dynamically creates the header. Thesecond mode has the advantage that for CPU and memory hotplug, kdump hasnot to be reloaded withkexec_load().

  • For s390x systems with many attached devices the “cio_ignore” kernelparameter should be used for the kdump kernel in order to prevent allocationof kernel memory for devices that are not relevant for kdump. The sameapplies to systems that use SCSI/FCP devices. In that case the“allow_lun_scan” zfcp module parameter should be set to zero beforesetting FCP devices online.

Kernel Panic

After successfully loading the dump-capture kernel as previouslydescribed, the system will reboot into the dump-capture kernel if asystem crash is triggered. Trigger points are located inpanic(),die(),die_nmi() and in the sysrq handler (ALT-SysRq-c).

The following conditions will execute a crash trigger point:

If a hard lockup is detected and “NMI watchdog” is configured, the systemwill boot into the dump-capture kernel (die_nmi() ).

Ifdie() is called, and it happens to be a thread with pid 0 or 1, ordie()is called inside interrupt context ordie() is called and panic_on_oops is set,the system will boot into the dump-capture kernel.

On powerpc systems when a soft-reset is generated,die() is called by all cpusand the system will boot into the dump-capture kernel.

For testing purposes, you can trigger a crash by using “ALT-SysRq-c”,“echo c > /proc/sysrq-trigger” or write a module to force the panic.

Write Out the Dump File

After the dump-capture kernel is booted, write out the dump file withthe following command:

cp /proc/vmcore <dump-file>

or use scp to write out the dump file between hosts on a network, e.g:

scp /proc/vmcore remote_username@remote_ip:<dump-file>

You can also use makedumpfile utility to write out the dump filewith specified options to filter out unwanted contents, e.g:

makedumpfile -l --message-level 1 -d 31 /proc/vmcore <dump-file>

Analysis

Before analyzing the dump image, you should reboot into a stable kernel.

You can do limited analysis using GDB on the dump file copied out of/proc/vmcore. Use the debug vmlinux built with -g and run the followingcommand:

gdb vmlinux <dump-file>

Stack trace for the task on processor 0, register display, and memorydisplay work fine.

Note: GDB cannot analyze core files generated in ELF64 format for x86.On systems with a maximum of 4GB of memory, you can generateELF32-format headers using the --elf32-core-headers kernel option on thedump kernel.

You can also use the Crash utility to analyze dump files in Kdumpformat. Crash is available at the following URL:

Crash document can be found at:

https://crash-utility.github.io/

Trigger Kdump on WARN()

The kernel parameter, panic_on_warn, callspanic() in allWARN() paths. Thiswill cause a kdump to occur at thepanic() call. In cases where a user wantsto specify this during runtime, /proc/sys/kernel/panic_on_warn can be set to 1to achieve the same behaviour.

Trigger Kdump on add_taint()

The kernel parameter panic_on_taint facilitates a conditional call topanic()from withinadd_taint() whenever the value set in this bitmask matches with thebit flag being set byadd_taint().This will cause a kdump to occur at theadd_taint()->panic() call.

Write the dump file to encrypted disk volume

CONFIG_CRASH_DM_CRYPT can be enabled to support saving the dump file to anencrypted disk volume (only x86_64 supported for now). User space can interactwith /sys/kernel/config/crash_dm_crypt_keys for setup,

  1. Tell the first kernel what logon keys are needed to unlock the disk volumes,

    # Add key #1mkdir /sys/kernel/config/crash_dm_crypt_keys/7d26b7b4-e342-4d2d-b660-7426b0996720# Add key #1’s descriptionecho cryptsetup:7d26b7b4-e342-4d2d-b660-7426b0996720 > /sys/kernel/config/crash_dm_crypt_keys/description

    # how many keys do we have now?cat /sys/kernel/config/crash_dm_crypt_keys/count1

    # Add key #2 in the same way

    # how many keys do we have now?cat /sys/kernel/config/crash_dm_crypt_keys/count2

    # To support CPU/memory hot-plugging, re-use keys already saved to reserved# memoryecho true > /sys/kernel/config/crash_dm_crypt_key/reuse

  2. Load the dump-capture kernel

  3. After the dump-capture kerne get booted, restore the keys to user keyringecho yes > /sys/kernel/crash_dm_crypt_keys/restore

Contact

GDB macros

## This file contains a few gdb macros (user defined commands) to extract# useful information from kernel crashdump (kdump) like stack traces of# all the processes or a particular process and trapinfo.## These macros can be used by copying this file in .gdbinit (put in home# directory or current directory) or by invoking gdb command with# --command=<command-file-name> option## Credits:# Alexander Nyberg <alexn@telia.com># V Srivatsa <vatsa@in.ibm.com># Maneesh Soni <maneesh@in.ibm.com>#define bttnobp        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)        set $init_t=&init_task        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)        set var $stacksize = sizeof(union thread_union)        while ($next_t != $init_t)                set $next_t=(struct task_struct *)$next_t                printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm                printf "===================\n"                set var $stackp = $next_t.thread.sp                set var $stack_top = ($stackp & ~($stacksize - 1)) + $stacksize                while ($stackp < $stack_top)                        if (*($stackp) > _stext && *($stackp) < _sinittext)                                info symbol *($stackp)                        end                        set $stackp += 4                end                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)                while ($next_th != $next_t)                        set $next_th=(struct task_struct *)$next_th                        printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm                        printf "===================\n"                        set var $stackp = $next_t.thread.sp                        set var $stack_top = ($stackp & ~($stacksize - 1)) + stacksize                        while ($stackp < $stack_top)                                if (*($stackp) > _stext && *($stackp) < _sinittext)                                        info symbol *($stackp)                                end                                set $stackp += 4                        end                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)                end                set $next_t=(char *)($next_t->tasks.next) - $tasks_off        endenddocument bttnobp        dump all thread stack traces on a kernel compiled with !CONFIG_FRAME_POINTERenddefine btthreadstack        set var $pid_task = $arg0        printf "\npid %d; comm %s:\n", $pid_task.pid, $pid_task.comm        printf "task struct: "        print $pid_task        printf "===================\n"        set var $stackp = $pid_task.thread.sp        set var $stacksize = sizeof(union thread_union)        set var $stack_top = ($stackp & ~($stacksize - 1)) + $stacksize        set var $stack_bot = ($stackp & ~($stacksize - 1))        set $stackp = *((unsigned long *) $stackp)        while (($stackp < $stack_top) && ($stackp > $stack_bot))                set var $addr = *(((unsigned long *) $stackp) + 1)                info symbol $addr                set $stackp = *((unsigned long *) $stackp)        endenddocument btthreadstack         dump a thread stack using the given task structure pointerenddefine btt        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)        set $init_t=&init_task        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)        while ($next_t != $init_t)                set $next_t=(struct task_struct *)$next_t                btthreadstack $next_t                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)                while ($next_th != $next_t)                        set $next_th=(struct task_struct *)$next_th                        btthreadstack $next_th                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)                end                set $next_t=(char *)($next_t->tasks.next) - $tasks_off        endenddocument btt        dump all thread stack traces on a kernel compiled with CONFIG_FRAME_POINTERenddefine btpid        set var $pid = $arg0        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)        set $init_t=&init_task        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)        set var $pid_task = 0        while ($next_t != $init_t)                set $next_t=(struct task_struct *)$next_t                if ($next_t.pid == $pid)                        set $pid_task = $next_t                end                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)                while ($next_th != $next_t)                        set $next_th=(struct task_struct *)$next_th                        if ($next_th.pid == $pid)                                set $pid_task = $next_th                        end                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)                end                set $next_t=(char *)($next_t->tasks.next) - $tasks_off        end        btthreadstack $pid_taskenddocument btpid        backtrace of pidenddefine trapinfo        set var $pid = $arg0        set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)        set $pid_off=((size_t)&((struct task_struct *)0)->thread_group.next)        set $init_t=&init_task        set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)        set var $pid_task = 0        while ($next_t != $init_t)                set $next_t=(struct task_struct *)$next_t                if ($next_t.pid == $pid)                        set $pid_task = $next_t                end                set $next_th=(((char *)$next_t->thread_group.next) - $pid_off)                while ($next_th != $next_t)                        set $next_th=(struct task_struct *)$next_th                        if ($next_th.pid == $pid)                                set $pid_task = $next_th                        end                        set $next_th=(((char *)$next_th->thread_group.next) - $pid_off)                end                set $next_t=(char *)($next_t->tasks.next) - $tasks_off        end        printf "Trapno %ld, cr2 0x%lx, error_code %ld\n", $pid_task.thread.trap_no, \                                $pid_task.thread.cr2, $pid_task.thread.error_codeenddocument trapinfo        Run info threads and lookup pid of thread #1        'trapinfo <pid>' will tell you by which trap & possibly        address the kernel panicked.enddefine dump_record        set var $desc = $arg0        set var $info = $arg1        if ($argc > 2)                set var $prev_flags = $arg2        else                set var $prev_flags = 0        end        set var $prefix = 1        set var $newline = 1        set var $begin = $desc->text_blk_lpos.begin % (1U << prb->text_data_ring.size_bits)        set var $next = $desc->text_blk_lpos.next % (1U << prb->text_data_ring.size_bits)        # handle data-less record        if ($begin & 1)                set var $text_len = 0                set var $log = ""        else                # handle wrapping data block                if ($begin > $next)                        set var $begin = 0                end                # skip over descriptor id                set var $begin = $begin + sizeof(long)                # handle truncated message                if ($next - $begin < $info->text_len)                        set var $text_len = $next - $begin                else                        set var $text_len = $info->text_len                end                set var $log = &prb->text_data_ring.data[$begin]        end        # prev & LOG_CONT && !(info->flags & LOG_PREIX)        if (($prev_flags & 8) && !($info->flags & 4))                set var $prefix = 0        end        # info->flags & LOG_CONT        if ($info->flags & 8)                # (prev & LOG_CONT && !(prev & LOG_NEWLINE))                if (($prev_flags & 8) && !($prev_flags & 2))                        set var $prefix = 0                end                # (!(info->flags & LOG_NEWLINE))                if (!($info->flags & 2))                        set var $newline = 0                end        end        if ($prefix)                printf "[%5lu.%06lu] ", $info->ts_nsec / 1000000000, $info->ts_nsec % 1000000000        end        if ($text_len)                eval "printf \"%%%d.%ds\", $log", $text_len, $text_len        end        if ($newline)                printf "\n"        end        # handle dictionary data        set var $dict = &$info->dev_info.subsystem[0]        set var $dict_len = sizeof($info->dev_info.subsystem)        if ($dict[0] != '\0')                printf " SUBSYSTEM="                set var $idx = 0                while ($idx < $dict_len)                        set var $c = $dict[$idx]                        if ($c == '\0')                                loop_break                        else                                if ($c < ' ' || $c >= 127 || $c == '\\')                                        printf "\\x%02x", $c                                else                                        printf "%c", $c                                end                        end                        set var $idx = $idx + 1                end                printf "\n"        end        set var $dict = &$info->dev_info.device[0]        set var $dict_len = sizeof($info->dev_info.device)        if ($dict[0] != '\0')                printf " DEVICE="                set var $idx = 0                while ($idx < $dict_len)                        set var $c = $dict[$idx]                        if ($c == '\0')                                loop_break                        else                                if ($c < ' ' || $c >= 127 || $c == '\\')                                        printf "\\x%02x", $c                                else                                        printf "%c", $c                                end                        end                        set var $idx = $idx + 1                end                printf "\n"        endenddocument dump_record        Dump a single record. The first parameter is the descriptor,        the second parameter is the info, the third parameter is        optional and specifies the previous record's flags, used for        properly formatting continued lines.enddefine dmesg        # definitions from kernel/printk/printk_ringbuffer.h        set var $desc_committed = 1        set var $desc_finalized = 2        set var $desc_sv_bits = sizeof(long) * 8        set var $desc_flags_shift = $desc_sv_bits - 2        set var $desc_flags_mask = 3 << $desc_flags_shift        set var $id_mask = ~$desc_flags_mask        set var $desc_count = 1U << prb->desc_ring.count_bits        set var $prev_flags = 0        set var $id = prb->desc_ring.tail_id.counter        set var $end_id = prb->desc_ring.head_id.counter        while (1)                set var $desc = &prb->desc_ring.descs[$id % $desc_count]                set var $info = &prb->desc_ring.infos[$id % $desc_count]                # skip non-committed record                set var $state = 3 & ($desc->state_var.counter >> $desc_flags_shift)                if ($state == $desc_committed || $state == $desc_finalized)                        dump_record $desc $info $prev_flags                        set var $prev_flags = $info->flags                end                if ($id == $end_id)                        loop_break                end                set var $id = ($id + 1) & $id_mask        endenddocument dmesg        print the kernel ring bufferend