Documentation: kdump: update kdump guide
Subsystems (4)
PATCH 1/1 Documentation: kdump: update kdump guide
Some parts of the guide are aged, hence need be updated.
1) The backup area of the 1st 640K on X86_64 has been removed by below commits, update the description accordingly.
commit 7c321eb2b843 ("x86/kdump: Remove the backup region handling")
commit 6f599d84231f ("x86/kdump: Always reserve the low 1M when the crashkernel option is specified")
2) Sort out the descripiton of "crashkernel syntax" part.
3) And some other minor cleanups.
--- Documentation/admin-guide/kdump/kdump.rst | 150 ++++++++++++++-------- 1 file changed, 97 insertions(+), 53 deletions(-)
Documentation/admin-guide/kdump/kdump.rst +97 -53
@@ -2,7 +2,7 @@ Documentation for Kdump - The kexec-based Crash Dumping Solution ================================================================ -This document includes overview, setup and installation, and analysis +This document includes overview, setup, installation, and analysis information. Overview
@@ -13,12 +13,12 @@ dump of the system kernel's memory needs to be taken (for example, when the system panics). The system kernel's memory image is preserved across the reboot and is accessible to the dump-capture kernel. -You can use common commands, such as cp and scp, to copy the -memory image to a dump file on the local disk, or across the network to -a remote system. +You can use common commands, such as cp, scp or makedumpfile to copy +the memory image to a dump file on the local disk, or across the network +to a remote system. -Kdump and kexec are currently supported on the x86, x86_64, ppc64, ia64, -s390x, arm and arm64 architectures. +Kdump and kexec are currently supported on the x86/64, ppc64, ia64, +s390x, arm/64 architectures. When the system kernel boots, it reserves a small section of memory for the dump-capture kernel. This ensures that ongoing Direct Memory Access
@@ -27,12 +27,14 @@ The kexec -p command loads the dump-capture kernel into this reserved memory. On x86 machines, the first 640 KB of physical memory is needed to boot, -regardless of where the kernel loads. Therefore, kexec backs up this -region just before rebooting into the dump-capture kernel. +regardless of where the kernel loads. For simpler handling, the whole +low 1M is reserved to avoid any later kernel or device driver writing +data into this area. Like this, the low 1M can be reused as system RAM +by kdump kernel without extra handling. -Similarly on PPC64 machines first 32KB of physical memory is needed for -booting regardless of where the kernel is loaded and to support 64K page -size kexec backs up the first 64KB memory. +On PPC64 machines first 32KB of physical memory is needed for booting +regardless of where the kernel is loaded and to support 64K page size +kexec backs up the first 64KB memory. For s390x, when kdump is triggered, the crashkernel region is exchanged with the region [0, crashkernel region size] and then the kdump kernel
@@ -46,14 +48,14 @@ passed to the dump-capture kernel through the elfcorehdr= boot parameter. Optionally the size of the ELF header can also be passed when 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 can -write out using file copy commands such as cp or scp. Further, you can -use analysis tools such as the GNU Debugger (GDB) and the Crash tool to -debug the dump file. This method ensures that the dump pages are correctly -ordered. - +write out using file copy commands such as cp or scp. You can also use +makedumpfile utility to analyze and write out filtered contents with +options, 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 Crash +tool to debug the dump file. This method ensures that the dump pages are +correctly ordered. Setup and Installation ======================
@@ -111,7 +113,7 @@ There are two possible methods of using Kdump. 2) Or use the system kernel binary itself as dump-capture kernel and there is no need to build a separate dump-capture kernel. This is possible only with the architectures which support a relocatable kernel. As - of today, i386, x86_64, ppc64, ia64, arm and arm64 architectures support + of today, x86/64, ppc64, ia64, arm/64 architectures support relocatable kernel. Building a relocatable kernel is advantageous from the point of view that
@@ -125,9 +127,18 @@ dump-capture kernels for enabling kdump support. System kernel config options ---------------------------- -1) Enable "kexec system call" in "Processor type and features.":: +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=y + CONFIG_KEXEC_CORE=y + + Subsequently, CRASH_CORE is selected by KEXEC_CORE:: + + CONFIG_CRASH_CORE=y 2) Enable "sysfs file system support" in "Filesystem" -> "Pseudo filesystems." This is usually enabled by default::
@@ -180,7 +191,7 @@ Dump-capture kernel config options (Arch Dependent, i386 and x86_64) CONFIG_SMP=n - (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line + (If CONFIG_SMP=y, then specify nr_cpus=1 on the kernel command line when loading the dump-capture kernel, see section "Load the Dump-capture Kernel".)
@@ -241,10 +252,6 @@ Dump-capture kernel config options (Arch Dependent, ia64) crashkernel=256M - If the start address is specified, note that the start address of the - kernel will be aligned to 64Mb, so if the start address is not then - any space below the alignment point will be wasted. - Dump-capture kernel config options (Arch Dependent, arm) ----------------------------------------------------------
@@ -260,44 +267,77 @@ Dump-capture kernel config options (Arch Dependent, arm64) on non-VHE systems even if it is configured. This is because the CPU will not be reset to EL2 on panic. -Extended crashkernel syntax +crashkernel syntax =========================== +1) crashkernel=size@offset -While the "crashkernel=size[@offset]" syntax is sufficient for most -configurations, sometimes it's handy to have the reserved memory dependent -on the value of System RAM -- that's mostly for distributors that pre-setup -the kernel command line to avoid a unbootable system after some memory has -been removed from the machine. + Here 'size' specifies how much memory to reserve for the dump-capture kernel + and 'offset' specifies the beginning of this reserved memory. For example, + "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory + starting at physical address 0x01000000 (16MB) for the dump-capture kernel. + + The crashkernel region can be automatically placed by the system + kernel at run time. This is done by specifying the base address as 0, + or omitting it all together:: -The syntax is:: + crashkernel=256M@0 + or:: - crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] - range=start-[end] + crashkernel=256M -For example:: + If the start address is specified, note that the start address of the + kernel will be aligned to a value (which is Arch dependent), so if the + start address is not then any space below the alignment point will be + wasted. - crashkernel=512M-2G:64M,2G-:128M +2) range1:size1[,range2:size2,...][@offset] -This would mean: + While the "crashkernel=size[@offset]" syntax is sufficient for most + configurations, sometimes it's handy to have the reserved memory dependent + on the value of System RAM -- that's mostly for distributors that pre-setup + the kernel command line to avoid a unbootable system after some memory has + been removed from the machine. - 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 + The syntax is:: + crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] + range=start-[end] + For example:: -Boot into System Kernel -======================= + 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 to + fulfill that. With it, physical memory is allowed to allocate 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 memory + region above 4G, low memory under 4G is needed in this case. There are + three ways to get low memory: + + 1) Kernel will allocate at least 256M memory below 4G automatically + if crashkernel=Y,low is not specified. + 2) Let user specify low memory size instread. + 3) Specified value 0 will disable low memory allocation:: + + crashkernel=0,low + +Boot into System Kernel +----------------------- 1) Update the boot loader (such as grub, yaboot, or lilo) configuration files as necessary. -2) Boot the system kernel with the boot parameter "crashkernel=Y@X", - where Y specifies how much memory to reserve for the dump-capture kernel - and X specifies the beginning of this reserved memory. For example, - "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory - starting at physical address 0x01000000 (16MB) for the dump-capture kernel. +2) Boot the system kernel with the boot parameter "crashkernel=Y@X". On x86 and x86_64, use "crashkernel=64M@16M".
@@ -392,7 +432,7 @@ loading dump-capture kernel. For i386, x86_64 and ia64: - "1 irqpoll maxcpus=1 reset_devices" + "1 irqpoll nr_cpus=1 reset_devices" For ppc64:
@@ -400,7 +440,7 @@ For ppc64: For s390x: - "1 maxcpus=1 cgroup_disable=memory" + "1 nr_cpus=1 cgroup_disable=memory" For arm:
@@ -408,7 +448,7 @@ For arm: For arm64: - "1 maxcpus=1 reset_devices" + "1 nr_cpus=1 reset_devices" Notes on loading the dump-capture kernel:
@@ -487,7 +527,12 @@ After the dump-capture kernel is booted, write out the dump file with the following command:: cp /proc/vmcore <dump-file> + scp /proc/vmcore to <user@server>:<path>/%HOST-%DATE/ + +You can also use makedumpfile utility to write out the dump file +with specified options to filter out unwanted contents, e.g:: + core_collector makedumpfile -l --message-level 1 -d 31 Analysis ========
@@ -535,8 +580,7 @@ This will cause a kdump to occur at the add_taint()->panic() call. Contact ======= -- Vivek Goyal ([email protected]) -- Maneesh Soni ([email protected]) +- [email protected] GDB macros ==========
Signature · 7 lines
2.17.2 _______________________________________________ kexec mailing list [email protected] http://lists.infradead.org/mailman/listinfo/kexec
Hi Baoquan,
Thanks for the update! Since we are updating it I added arch maintainers to see if they have any comments about the archtectures part.
I added a few comments inline, but still want more inputs from other people :)
Quoted 120 lines · Baoquan He
Some parts of the guide are aged, hence need be updated.
1) The backup area of the 1st 640K on
X86_64has been removed by below commits, update the description accordingly.commit
7c321eb2b843("x86/kdump: Remove the backup region handling")
commit6f599d84231f("x86/kdump: Always reserve the low 1M when the crashkernel option is specified")2) Sort out the descripiton of "crashkernel syntax" part.
3) And some other minor cleanups.
Signed-off-by Baoquan He <[email protected]>--- Documentation/admin-guide/kdump/kdump.rst | 150 ++++++++++++++-------- 1 file changed, 97 insertions(+), 53 deletions(-)Documentation/admin-guide/kdump/kdump.rst +31
-20@@ -2,7 +2,7 @@ Documentation for Kdump - The kexec-based Crash Dumping Solution ================================================================ -This document includes overview, setup and installation, and analysis +This document includes overview, setup, installation, and analysis information. Overview@@ -13,12 +13,12 @@ dump of the system kernel's memory needs to be taken (for example, when the system panics). The system kernel's memory image is preserved across the reboot and is accessible to the dump-capture kernel. -You can use common commands, such as cp and scp, to copy the -memory image to a dump file on the local disk, or across the network to -a remote system. +You can use common commands, such as cp, scp or makedumpfile to copy +the memory image to a dump file on the local disk, or across the network +to a remote system. -Kdump and kexec are currently supported on the x86, x86_64, ppc64, ia64, -s390x, arm and arm64 architectures. +Kdump and kexec are currently supported on the x86/64, ppc64, ia64, +s390x, arm/64 architectures. When the system kernel boots, it reserves a small section of memory for the dump-capture kernel. This ensures that ongoing Direct Memory Access@@ -27,12 +27,14 @@ The kexec -p command loads the dump-capture kernel into this reserved memory. On x86 machines, the first 640 KB of physical memory is needed to boot, -regardless of where the kernel loads. Therefore, kexec backs up this -region just before rebooting into the dump-capture kernel. +regardless of where the kernel loads. For simpler handling, the whole +low 1M is reserved to avoid any later kernel or device driver writing +data into this area. Like this, the low 1M can be reused as system RAM +by kdump kernel without extra handling. -Similarly on PPC64 machines first 32KB of physical memory is needed for -booting regardless of where the kernel is loaded and to support 64K page -size kexec backs up the first 64KB memory. +On PPC64 machines first 32KB of physical memory is needed for booting +regardless of where the kernel is loaded and to support 64K page size +kexec backs up the first 64KB memory. For s390x, when kdump is triggered, the crashkernel region is exchanged with the region [0, crashkernel region size] and then the kdump kernel@@ -46,14 +48,14 @@ passed to the dump-capture kernel through the elfcorehdr= boot parameter. Optionally the size of the ELF header can also be passed when 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 can -write out using file copy commands such as cp or scp. Further, you can -use analysis tools such as the GNU Debugger (GDB) and the Crash tool to -debug the dump file. This method ensures that the dump pages are correctly -ordered. - +write out using file copy commands such as cp or scp. You can also use +makedumpfile utility to analyze and write out filtered contents with +options, 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 Crash +tool to debug the dump file. This method ensures that the dump pages are +correctly ordered. Setup and Installation ======================@@ -111,7 +113,7 @@ There are two possible methods of using Kdump. 2) Or use the system kernel binary itself as dump-capture kernel and there is no need to build a separate dump-capture kernel. This is possible only with the architectures which support a relocatable kernel. As - of today, i386, x86_64, ppc64, ia64, arm and arm64 architectures support + of today, x86/64, ppc64, ia64, arm/64 architectures support relocatable kernel. Building a relocatable kernel is advantageous from the point of view that@@ -125,9 +127,18 @@ dump-capture kernels for enabling kdump support. System kernel config options ---------------------------- -1) Enable "kexec system call" in "Processor type and features.":: +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=y + CONFIG_KEXEC_CORE=y + + Subsequently, CRASH_CORE is selected by KEXEC_CORE:: + + CONFIG_CRASH_CORE=y 2) Enable "sysfs file system support" in "Filesystem" -> "Pseudo filesystems." This is usually enabled by default::
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, i386 and x86_64) · line 196
Quoted 7 lines
Documentation/admin-guide/kdump/kdump.rst +1
-1@@ -180,7 +191,7 @@ Dump-capture kernel config options (Arch Dependent, i386 and x86_64) CONFIG_SMP=n - (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line + (If CONFIG_SMP=y, then specify nr_cpus=1 on the kernel command line when loading the dump-capture kernel, see section "Load the Dump-capture Kernel".)
This part should be obsolete? Since for X86_64 we can enable smp boot with disable_cpu_apicid=X set (see the Notes on loading the dump-capture kernel part) So I think no need to disable CONFIG_SMP at all. The current RHEL use of nr_cpus=1 is just to save 2nd kernel memory use.
Ditto for the text for other arches, not sure if they need update though, see if other maintainers can provide inputs..
Otherwise for the CONFIG_RELOCATABLE related part, it may be better to update as well?
''' quote:
3) If one wants to build and use a relocatable kernel, Enable "Build a relocatable kernel" support under "Processor type and features"::
CONFIG_RELOCATABLE=y
4) Use a suitable value for "Physical address where the kernel is loaded" (under "Processor type and features"). This only appears when "kernel crash dumps" is enabled. A suitable value depends upon whether kernel is relocatable or not.
If you are using a relocatable kernel use CONFIG_PHYSICAL_START=0x100000 This will compile the kernel for physical address 1MB, but given the fact kernel is relocatable, it can be run from any physical address hence kexec boot loader will load it in memory region reserved for dump-capture kernel.
Otherwise it should be the start of memory region reserved for second kernel using boot parameter "crashkernel=Y@X". Here X is start of memory region reserved for dump-capture kernel. Generally X is 16MB (0x1000000). So you can set CONFIG_PHYSICAL_START=0x1000000
''' end quote
Since relocatable kernel is used by default now so we may just not describe it as "If one want to build with it =y", I feel it should be a corner case instead of
the default use case. Maybe HPA, Vivek, Eric can provide more opinions since
they may know more about the background.
Quoted 102 lines
Documentation/admin-guide/kdump/kdump.rst +55
-22@@ -241,10 +252,6 @@ Dump-capture kernel config options (Arch Dependent, ia64) crashkernel=256M - If the start address is specified, note that the start address of the - kernel will be aligned to 64Mb, so if the start address is not then - any space below the alignment point will be wasted. - Dump-capture kernel config options (Arch Dependent, arm) ----------------------------------------------------------@@ -260,44 +267,77 @@ Dump-capture kernel config options (Arch Dependent, arm64) on non-VHE systems even if it is configured. This is because the CPU will not be reset to EL2 on panic. -Extended crashkernel syntax +crashkernel syntax =========================== +1) crashkernel=size@offset -While the "crashkernel=size[@offset]" syntax is sufficient for most -configurations, sometimes it's handy to have the reserved memory dependent -on the value of System RAM -- that's mostly for distributors that pre-setup -the kernel command line to avoid a unbootable system after some memory has -been removed from the machine. + Here 'size' specifies how much memory to reserve for the dump-capture kernel + and 'offset' specifies the beginning of this reserved memory. For example, + "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory + starting at physical address 0x01000000 (16MB) for the dump-capture kernel. + + The crashkernel region can be automatically placed by the system + kernel at run time. This is done by specifying the base address as 0, + or omitting it all together:: -The syntax is:: + crashkernel=256M@0 + or:: - crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] - range=start-[end] + crashkernel=256M -For example:: + If the start address is specified, note that the start address of the + kernel will be aligned to a value (which is Arch dependent), so if the + start address is not then any space below the alignment point will be + wasted. - crashkernel=512M-2G:64M,2G-:128M +2) range1:size1[,range2:size2,...][@offset] -This would mean: + While the "crashkernel=size[@offset]" syntax is sufficient for most + configurations, sometimes it's handy to have the reserved memory dependent + on the value of System RAM -- that's mostly for distributors that pre-setup + the kernel command line to avoid a unbootable system after some memory has + been removed from the machine. - 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 + The syntax is:: + crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] + range=start-[end] + For example:: -Boot into System Kernel -======================= + 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 to + fulfill that. With it, physical memory is allowed to allocate 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 memory + region above 4G, low memory under 4G is needed in this case. There are + three ways to get low memory: + + 1) Kernel will allocate at least 256M memory below 4G automatically + if crashkernel=Y,low is not specified. + 2) Let user specify low memory size instread. + 3) Specified value 0 will disable low memory allocation:: + + crashkernel=0,low + +Boot into System Kernel +----------------------- 1) Update the boot loader (such as grub, yaboot, or lilo) configuration
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, arm64) · line 342
Quoted 10 lines
Documentation/admin-guide/kdump/kdump.rst +1
-5files as necessary. -2) Boot the system kernel with the boot parameter "crashkernel=Y@X", - where Y specifies how much memory to reserve for the dump-capture kernel - and X specifies the beginning of this reserved memory. For example, - "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory - starting at physical address 0x01000000 (16MB) for the dump-capture kernel. +2) Boot the system kernel with the boot parameter "crashkernel=Y@X". On x86 and x86_64, use "crashkernel=64M@16M".
For the recommendation of crashkernel it would be better to drop the @16M since most people do not need it?
Quoted 29 lines
Documentation/admin-guide/kdump/kdump.rst +3
-3@@ -392,7 +432,7 @@ loading dump-capture kernel. For i386, x86_64 and ia64: - "1 irqpoll maxcpus=1 reset_devices" + "1 irqpoll nr_cpus=1 reset_devices" For ppc64:@@ -400,7 +440,7 @@ For ppc64: For s390x: - "1 maxcpus=1 cgroup_disable=memory" + "1 nr_cpus=1 cgroup_disable=memory" For arm:@@ -408,7 +448,7 @@ For arm: For arm64: - "1 maxcpus=1 reset_devices" + "1 nr_cpus=1 reset_devices" Notes on loading the dump-capture kernel:
Documentation/admin-guide/kdump/kdump.rst After the dump-capture kernel is booted, write out the dump file with · line 535
Quoted 9 lines
Documentation/admin-guide/kdump/kdump.rst +5
-0@@ -487,7 +527,12 @@ After the dump-capture kernel is booted, write out the dump file with the following command:: cp /proc/vmcore <dump-file> + scp /proc/vmcore to <user@server>:<path>/%HOST-%DATE/ + +You can also use makedumpfile utility to write out the dump file +with specified options to filter out unwanted contents, e.g:: + core_collector makedumpfile -l --message-level 1 -d 31
Looks like scp and core_collector usage are based on Fedora/RHEL, but since this doc is for generic upstream, it might be better to describe it in generic way, eg. (maybe just drop scp)
makedumpfile -l --message-level 1 -d 31 /proc/vmcore <dump-file> scp /proc/vmcore <user@server>:<path>/<dump-file>
Quoted 16 lines
Documentation/admin-guide/kdump/kdump.rst +1
-2Analysis ========@@ -535,8 +580,7 @@ This will cause a kdump to occur at the add_taint()->panic() call. Contact ======= -- Vivek Goyal ([email protected]) -- Maneesh Soni ([email protected]) +- [email protected] GDB macros ==========Signature · 1 line
2.17.2
Thanks
Dave
Quoted 5 lines · Dave Young
Hi Baoquan,
Quoted 4 lines
Documentation/admin-guide/kdump/kdump.rst +0
-0@@ -180,7 +191,7 @@ Dump-capture kernel config options (Arch Dependent, i386 and x86_64) CONFIG_SMP=n
Quoted 9 lines
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, i386 and x86_64) · line 196
Quoted 4 lines
Documentation/admin-guide/kdump/kdump.rst +1
-1- (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line + (If CONFIG_SMP=y, then specify nr_cpus=1 on the kernel command line when loading the dump-capture kernel, see section "Load the Dump-capture Kernel".)This part should be obsolete? Since for
X86_64we can enable smp boot with disable_cpu_apicid=X set (see the Notes on loading the dump-capture kernel part) So I think no need to disableCONFIG_SMPat all. The current RHEL use of nr_cpus=1 is just to save 2nd kernel memory use.
Keeping them because they are not wrong. Talking about default config, currently we only care about x86_64 mostly, not sure if we should remove i386 part too. Anyway, I am fine to remove them and the below relocatable thing.
Quoted 38 lines
Ditto for the text for other arches, not sure if they need update though, see if other maintainers can provide inputs..
Otherwise for the
CONFIG_RELOCATABLErelated part, it may be better to update as well?
''' quote:
3) If one wants to build and use a relocatable kernel, Enable "Build a relocatable kernel" support under "Processor type and features"::
CONFIG_RELOCATABLE=y4) Use a suitable value for "Physical address where the kernel is loaded" (under "Processor type and features"). This only appears when "kernel crash dumps" is enabled. A suitable value depends upon whether 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 fact kernel is relocatable, it can be run from any physical address hence kexec boot loader will load it in memory region reserved for dump-capture kernel.Otherwise it should be the start of memory region reserved for second kernel using boot parameter "crashkernel=Y@X". Here X is start of memory region reserved for dump-capture kernel. Generally X is 16MB (
0x1000000). So you can setCONFIG_PHYSICAL_START=0x1000000
''' end quoteSince relocatable kernel is used by default now so we may just not describe it as "If one want to build with it =y", I feel it should be a corner case instead of
the default use case. Maybe HPA, Vivek, Eric can provide more opinions since
they may know more about the background.Quoted 1 line
...
Quoted 41 lines
Quoted 41 lines
Documentation/admin-guide/kdump/kdump.rst +29
-7-Boot into System Kernel -======================= + 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 to + fulfill that. With it, physical memory is allowed to allocate 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 memory + region above 4G, low memory under 4G is needed in this case. There are + three ways to get low memory: + + 1) Kernel will allocate at least 256M memory below 4G automatically + if crashkernel=Y,low is not specified. + 2) Let user specify low memory size instread. + 3) Specified value 0 will disable low memory allocation:: + + crashkernel=0,low + +Boot into System Kernel +----------------------- 1) Update the boot loader (such as grub, yaboot, or lilo) configuration files as necessary. -2) Boot the system kernel with the boot parameter "crashkernel=Y@X", - where Y specifies how much memory to reserve for the dump-capture kernel - and X specifies the beginning of this reserved memory. For example, - "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory - starting at physical address 0x01000000 (16MB) for the dump-capture kernel. +2) Boot the system kernel with the boot parameter "crashkernel=Y@X".
Quoted 4 lines
Quoted 1 line
On x86 and x86_64, use "crashkernel=64M@16M".For the recommendation of crashkernel it would be better to drop the @16M since most people do not need it?
It's only an example? Change it as "crashkernel=128M" instead to make it more helpful?
Quoted 38 lines
Quoted 38 lines
Documentation/admin-guide/kdump/kdump.rst +7
-3@@ -392,7 +432,7 @@ loading dump-capture kernel. For i386, x86_64 and ia64: - "1 irqpoll maxcpus=1 reset_devices" + "1 irqpoll nr_cpus=1 reset_devices" For ppc64:@@ -400,7 +440,7 @@ For ppc64: For s390x: - "1 maxcpus=1 cgroup_disable=memory" + "1 nr_cpus=1 cgroup_disable=memory" For arm:@@ -408,7 +448,7 @@ For arm: For arm64: - "1 maxcpus=1 reset_devices" + "1 nr_cpus=1 reset_devices" Notes on loading the dump-capture kernel:@@ -487,7 +527,12 @@ After the dump-capture kernel is booted, write out the dump file with the following command:: cp /proc/vmcore <dump-file> + scp /proc/vmcore to <user@server>:<path>/%HOST-%DATE/ + +You can also use makedumpfile utility to write out the dump file +with specified options to filter out unwanted contents, e.g::
Quoted 7 lines
Documentation/admin-guide/kdump/kdump.rst After the dump-capture kernel is booted, write out the dump file with · line 535
Quoted 1 line
Documentation/admin-guide/kdump/kdump.rst +1
-0+ core_collector makedumpfile -l --message-level 1 -d 31Looks like scp and
core_collectorusage are based on Fedora/RHEL, but since this doc is for generic upstream, it might be better to describe it in generic way, eg. (maybe just drop scp)
makedumpfile -l --message-level 1 -d 31/proc/vmcore<dump-file> scp/proc/vmcore<user@server>:<path>/<dump-file>
Right, forgot removing core_collector. While scp is also a generic tool?
Thanks for reviewing.
Hi Baoquan,
Quoted 16 lines · Baoquan He
Quoted 5 lines · Dave Young
Hi Baoquan,
Quoted 4 lines
Documentation/admin-guide/kdump/kdump.rst +0
-0@@ -180,7 +191,7 @@ Dump-capture kernel config options (Arch Dependent, i386 and x86_64) CONFIG_SMP=nQuoted 9 lines
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, i386 and x86_64) · line 196
Quoted 4 lines
Documentation/admin-guide/kdump/kdump.rst +1
-1- (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line + (If CONFIG_SMP=y, then specify nr_cpus=1 on the kernel command line when loading the dump-capture kernel, see section "Load the Dump-capture Kernel".)This part should be obsolete? Since for
X86_64we can enable smp boot with disable_cpu_apicid=X set (see the Notes on loading the dump-capture kernel part) So I think no need to disableCONFIG_SMPat all. The current RHEL use of nr_cpus=1 is just to save 2nd kernel memory use.
Quoted 4 lines
Keeping them because they are not wrong. Talking about default config, currently we only care about
x86_64mostly, not sure if we should remove i386 part too. Anyway, I am fine to remove them and the below relocatable thing.
I also agree it is not wrong :) But I personally think the doc should target for the most common use cases. If CONFIG_SMP=n is not common then we may just describe the default words for CONFIG_SMP=y, and we may add some words for exection cases
for example:
Specify nr_cpus=1 blabla
Note: if CONFIG_SMP is not set then nr_cpus=1 is not needed ...
Quoted 81 lines
Quoted 39 lines
Ditto for the text for other arches, not sure if they need update though, see if other maintainers can provide inputs..
Otherwise for the
CONFIG_RELOCATABLErelated part, it may be better to update as well?
''' quote:
3) If one wants to build and use a relocatable kernel, Enable "Build a relocatable kernel" support under "Processor type and features"::
CONFIG_RELOCATABLE=y4) Use a suitable value for "Physical address where the kernel is loaded" (under "Processor type and features"). This only appears when "kernel crash dumps" is enabled. A suitable value depends upon whether 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 fact kernel is relocatable, it can be run from any physical address hence kexec boot loader will load it in memory region reserved for dump-capture kernel.Otherwise it should be the start of memory region reserved for second kernel using boot parameter "crashkernel=Y@X". Here X is start of memory region reserved for dump-capture kernel. Generally X is 16MB (
0x1000000). So you can setCONFIG_PHYSICAL_START=0x1000000
''' end quoteSince relocatable kernel is used by default now so we may just not describe it as "If one want to build with it =y", I feel it should be a corner case instead of
the default use case. Maybe HPA, Vivek, Eric can provide more opinions since
they may know more about the background.Quoted 1 line
...
Quoted 41 lines
Quoted 41 lines
Documentation/admin-guide/kdump/kdump.rst +29
-7-Boot into System Kernel -======================= + 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 to + fulfill that. With it, physical memory is allowed to allocate 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 memory + region above 4G, low memory under 4G is needed in this case. There are + three ways to get low memory: + + 1) Kernel will allocate at least 256M memory below 4G automatically + if crashkernel=Y,low is not specified. + 2) Let user specify low memory size instread. + 3) Specified value 0 will disable low memory allocation:: + + crashkernel=0,low + +Boot into System Kernel +----------------------- 1) Update the boot loader (such as grub, yaboot, or lilo) configuration files as necessary. -2) Boot the system kernel with the boot parameter "crashkernel=Y@X", - where Y specifies how much memory to reserve for the dump-capture kernel - and X specifies the beginning of this reserved memory. For example, - "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory - starting at physical address 0x01000000 (16MB) for the dump-capture kernel. +2) Boot the system kernel with the boot parameter "crashkernel=Y@X".
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
On x86 and x86_64, use "crashkernel=64M@16M".For the recommendation of crashkernel it would be better to drop the @16M since most people do not need it?
It's only an example? Change it as "crashkernel=128M" instead to make it more helpful?
Hmm, it would be better to make the most common use case in default example, and put the corner cases on corner :)
Quoted 47 lines
Quoted 39 lines
Quoted 39 lines
Documentation/admin-guide/kdump/kdump.rst +7
-3@@ -392,7 +432,7 @@ loading dump-capture kernel. For i386, x86_64 and ia64: - "1 irqpoll maxcpus=1 reset_devices" + "1 irqpoll nr_cpus=1 reset_devices" For ppc64:@@ -400,7 +440,7 @@ For ppc64: For s390x: - "1 maxcpus=1 cgroup_disable=memory" + "1 nr_cpus=1 cgroup_disable=memory" For arm:@@ -408,7 +448,7 @@ For arm: For arm64: - "1 maxcpus=1 reset_devices" + "1 nr_cpus=1 reset_devices" Notes on loading the dump-capture kernel:@@ -487,7 +527,12 @@ After the dump-capture kernel is booted, write out the dump file with the following command:: cp /proc/vmcore <dump-file> + scp /proc/vmcore to <user@server>:<path>/%HOST-%DATE/ + +You can also use makedumpfile utility to write out the dump file +with specified options to filter out unwanted contents, e.g::Quoted 7 lines
Documentation/admin-guide/kdump/kdump.rst After the dump-capture kernel is booted, write out the dump file with · line 535
Quoted 1 line
Documentation/admin-guide/kdump/kdump.rst +1
-0+ core_collector makedumpfile -l --message-level 1 -d 31Looks like scp and
core_collectorusage are based on Fedora/RHEL, but since this doc is for generic upstream, it might be better to describe it in generic way, eg. (maybe just drop scp)
makedumpfile -l --message-level 1 -d 31/proc/vmcore<dump-file> scp/proc/vmcore<user@server>:<path>/<dump-file>
Quoted 1 line
Right, forgot removing
core_collector. While scp is also a generic tool?
scp is good to have, just it is similar to cp, as these are just examples I'm fine for either.
If taking scp then just drop the " to" and %HOST_%DATE/ etc.
Quoted 3 lines
Thanks for reviewing.
Thanks
Dave
_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec
PATCH v2 1/1 Documentation: kdump: update kdump guide
Some parts of the guide are aged, hence need be updated.
1) The backup area of the 1st 640K on X86_64 has been removed by below commits, update the description accordingly.
commit 7c321eb2b843 ("x86/kdump: Remove the backup region handling")
commit 6f599d84231f ("x86/kdump: Always reserve the low 1M when the crashkernel option is specified")
2) Sort out the descripiton of "crashkernel syntax" part.
3) And some other minor cleanups.
---v1->v2:
Update the obsolete descriptions about SMP and RELOCATABLE according to Dave's comment.
Documentation/admin-guide/kdump/kdump.rst | 165 ++++++++++++++-------- 1 file changed, 106 insertions(+), 59 deletions(-)
Documentation/admin-guide/kdump/kdump.rst +106 -59
@@ -2,7 +2,7 @@ Documentation for Kdump - The kexec-based Crash Dumping Solution ================================================================ -This document includes overview, setup and installation, and analysis +This document includes overview, setup, installation, and analysis information. Overview
@@ -13,9 +13,9 @@ dump of the system kernel's memory needs to be taken (for example, when the system panics). The system kernel's memory image is preserved across the reboot and is accessible to the dump-capture kernel. -You can use common commands, such as cp and scp, to copy the -memory image to a dump file on the local disk, or across the network to -a remote system. +You can use common commands, such as cp, scp or makedumpfile to copy +the memory image to a dump file on the local disk, or across the network +to a remote system. Kdump and kexec are currently supported on the x86, x86_64, ppc64, ia64, s390x, arm and arm64 architectures.
@@ -27,12 +27,14 @@ The kexec -p command loads the dump-capture kernel into this reserved memory. On x86 machines, the first 640 KB of physical memory is needed to boot, -regardless of where the kernel loads. Therefore, kexec backs up this -region just before rebooting into the dump-capture kernel. +regardless of where the kernel loads. For simpler handling, the whole +low 1M is reserved to avoid any later kernel or device driver writing +data into this area. Like this, the low 1M can be reused as system RAM +by kdump kernel without extra handling. -Similarly on PPC64 machines first 32KB of physical memory is needed for -booting regardless of where the kernel is loaded and to support 64K page -size kexec backs up the first 64KB memory. +On PPC64 machines first 32KB of physical memory is needed for booting +regardless of where the kernel is loaded and to support 64K page size +kexec backs up the first 64KB memory. For s390x, when kdump is triggered, the crashkernel region is exchanged with the region [0, crashkernel region size] and then the kdump kernel
@@ -46,14 +48,14 @@ passed to the dump-capture kernel through the elfcorehdr= boot parameter. Optionally the size of the ELF header can also be passed when 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 can -write out using file copy commands such as cp or scp. Further, you can -use analysis tools such as the GNU Debugger (GDB) and the Crash tool to -debug the dump file. This method ensures that the dump pages are correctly -ordered. - +write out using file copy commands such as cp or scp. You can also use +makedumpfile utility to analyze and write out filtered contents with +options, 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 Crash +tool to debug the dump file. This method ensures that the dump pages are +correctly ordered. Setup and Installation ======================
@@ -125,9 +127,18 @@ dump-capture kernels for enabling kdump support. System kernel config options ---------------------------- -1) Enable "kexec system call" in "Processor type and features.":: +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=y + CONFIG_KEXEC_CORE=y + + Subsequently, CRASH_CORE is selected by KEXEC_CORE:: + + CONFIG_CRASH_CORE=y 2) Enable "sysfs file system support" in "Filesystem" -> "Pseudo filesystems." This is usually enabled by default::
@@ -175,17 +186,19 @@ Dump-capture kernel config options (Arch Dependent, i386 and x86_64) CONFIG_HIGHMEM4G -2) On i386 and x86_64, disable symmetric multi-processing support - under "Processor type and features":: +2) With CONFIG_SMP=y, usually nr_cpus=1 need specified on the kernel + command line when loading the dump-capture kernel because one + CPU is enough for kdump kernel to dump vmcore on most of systems. - CONFIG_SMP=n + However, you can also specify nr_cpus=X to enable multiple processors + in kdump kernel. In this case, "disable_cpu_apicid=" is needed to + tell kdump kernel which cpu is 1st kernel's BSP. Please refer to + admin-guide/kernel-parameters.txt for more details. - (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line - when loading the dump-capture kernel, see section "Load the Dump-capture - Kernel".) + With CONFIG_SMP=n, the above things are not related. -3) If one wants to build and use a relocatable kernel, - Enable "Build a relocatable kernel" support under "Processor type and +3) A relocatable kernel is suggested to be built by default. If not yet, + enable "Build a relocatable kernel" support under "Processor type and features":: CONFIG_RELOCATABLE=y
@@ -241,10 +254,6 @@ Dump-capture kernel config options (Arch Dependent, ia64) crashkernel=256M - If the start address is specified, note that the start address of the - kernel will be aligned to 64Mb, so if the start address is not then - any space below the alignment point will be wasted. - Dump-capture kernel config options (Arch Dependent, arm) ----------------------------------------------------------
@@ -260,46 +269,81 @@ Dump-capture kernel config options (Arch Dependent, arm64) on non-VHE systems even if it is configured. This is because the CPU will not be reset to EL2 on panic. -Extended crashkernel syntax +crashkernel syntax =========================== +1) crashkernel=size@offset -While the "crashkernel=size[@offset]" syntax is sufficient for most -configurations, sometimes it's handy to have the reserved memory dependent -on the value of System RAM -- that's mostly for distributors that pre-setup -the kernel command line to avoid a unbootable system after some memory has -been removed from the machine. + Here 'size' specifies how much memory to reserve for the dump-capture kernel + and 'offset' specifies the beginning of this reserved memory. For example, + "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory + starting at physical address 0x01000000 (16MB) for the dump-capture kernel. -The syntax is:: + The crashkernel region can be automatically placed by the system + kernel at run time. This is done by specifying the base address as 0, + or omitting it all together:: - crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] - range=start-[end] + crashkernel=256M@0 + or:: -For example:: + crashkernel=256M - crashkernel=512M-2G:64M,2G-:128M + If the start address is specified, note that the start address of the + kernel will be aligned to a value (which is Arch dependent), so if the + start address is not then any space below the alignment point will be + wasted. -This would mean: +2) range1:size1[,range2:size2,...][@offset] - 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 + While the "crashkernel=size[@offset]" syntax is sufficient for most + configurations, sometimes it's handy to have the reserved memory dependent + on the value of System RAM -- that's mostly for distributors that pre-setup + the kernel command line to avoid a unbootable system after some memory has + been removed from the machine. + The syntax is:: + crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] + range=start-[end] -Boot into System Kernel -======================= + 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 to + fulfill that. With it, physical memory is allowed to allocate 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 memory + region above 4G, low memory under 4G is needed in this case. There are + three ways to get low memory: + + 1) Kernel will allocate at least 256M memory below 4G automatically + if crashkernel=Y,low is not specified. + 2) Let user specify low memory size instread. + 3) Specified value 0 will disable low memory allocation:: + + crashkernel=0,low + +Boot into System Kernel +----------------------- 1) Update the boot loader (such as grub, yaboot, or lilo) configuration files as necessary. -2) Boot the system kernel with the boot parameter "crashkernel=Y@X", - where Y specifies how much memory to reserve for the dump-capture kernel - and X specifies the beginning of this reserved memory. For example, - "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory - starting at physical address 0x01000000 (16MB) for the dump-capture kernel. +2) Boot the system kernel with the boot parameter "crashkernel=Y@X". - On x86 and x86_64, use "crashkernel=64M@16M". + On x86 and x86_64, use "crashkernel=Y[@X]". Most of the time, the + start address 'X' is not necessary, kernel will search a suitable + area. Unless an explicit start address is expected. On ppc64, use "crashkernel=128M@32M".
@@ -331,8 +375,8 @@ of dump-capture kernel. Following is the summary. For i386 and x86_64: - - Use vmlinux if kernel is not relocatable. - Use bzImage/vmlinuz if kernel is relocatable. + - Use vmlinux if kernel is not relocatable. For ppc64:
@@ -392,7 +436,7 @@ loading dump-capture kernel. For i386, x86_64 and ia64: - "1 irqpoll maxcpus=1 reset_devices" + "1 irqpoll nr_cpus=1 reset_devices" For ppc64:
@@ -400,7 +444,7 @@ For ppc64: For s390x: - "1 maxcpus=1 cgroup_disable=memory" + "1 nr_cpus=1 cgroup_disable=memory" For arm:
@@ -408,7 +452,7 @@ For arm: For arm64: - "1 maxcpus=1 reset_devices" + "1 nr_cpus=1 reset_devices" Notes on loading the dump-capture kernel:
@@ -488,6 +532,10 @@ the following command:: cp /proc/vmcore <dump-file> +You can also use makedumpfile utility to write out the dump file +with specified options to filter out unwanted contents, e.g:: + + makedumpfile -l --message-level 1 -d 31 /proc/vmcore <dump-file> Analysis ========
@@ -535,8 +583,7 @@ This will cause a kdump to occur at the add_taint()->panic() call. Contact ======= -- Vivek Goyal ([email protected]) -- Maneesh Soni ([email protected]) +- [email protected] GDB macros ==========
Signature · 1 line
2.17.2
Hi Baoquan,
Just some spell checking found issues please see comments inline. Otherwise looks good to me:
Quoted 48 lines · Baoquan He
Some parts of the guide are aged, hence need be updated.
1) The backup area of the 1st 640K on
X86_64has been removed by below commits, update the description accordingly.commit
7c321eb2b843("x86/kdump: Remove the backup region handling")
commit6f599d84231f("x86/kdump: Always reserve the low 1M when the crashkernel option is specified")2) Sort out the descripiton of "crashkernel syntax" part.
3) And some other minor cleanups.
Signed-off-by Baoquan He <[email protected]>---
v1->v2:
Update the obsolete descriptions about SMP and RELOCATABLE according to Dave's comment.Documentation/admin-guide/kdump/kdump.rst | 165 ++++++++++++++-------- 1 file changed, 106 insertions(+), 59 deletions(-)Documentation/admin-guide/kdump/kdump.rst +4
-4@@ -2,7 +2,7 @@ Documentation for Kdump - The kexec-based Crash Dumping Solution ================================================================ -This document includes overview, setup and installation, and analysis +This document includes overview, setup, installation, and analysis information. Overview@@ -13,9 +13,9 @@ dump of the system kernel's memory needs to be taken (for example, when the system panics). The system kernel's memory image is preserved across the reboot and is accessible to the dump-capture kernel. -You can use common commands, such as cp and scp, to copy the -memory image to a dump file on the local disk, or across the network to -a remote system. +You can use common commands, such as cp, scp or makedumpfile to copy +the memory image to a dump file on the local disk, or across the network +to a remote system. Kdump and kexec are currently supported on the x86, x86_64, ppc64, ia64, s390x, arm and arm64 architectures.
Documentation/admin-guide/kdump/kdump.rst The kexec -p command loads the dump-capture kernel into this reserved · line 29
Quoted 3 lines
Documentation/admin-guide/kdump/kdump.rst +0
-0@@ -27,12 +27,14 @@ The kexec -p command loads the dump-capture kernel into this reserved memory. On x86 machines, the first 640 KB of physical memory is needed to boot,
s/to boot/for boot
Quoted 109 lines
Documentation/admin-guide/kdump/kdump.rst +36
-29-regardless of where the kernel loads. Therefore, kexec backs up this -region just before rebooting into the dump-capture kernel. +regardless of where the kernel loads. For simpler handling, the whole +low 1M is reserved to avoid any later kernel or device driver writing +data into this area. Like this, the low 1M can be reused as system RAM +by kdump kernel without extra handling. -Similarly on PPC64 machines first 32KB of physical memory is needed for -booting regardless of where the kernel is loaded and to support 64K page -size kexec backs up the first 64KB memory. +On PPC64 machines first 32KB of physical memory is needed for booting +regardless of where the kernel is loaded and to support 64K page size +kexec backs up the first 64KB memory. For s390x, when kdump is triggered, the crashkernel region is exchanged with the region [0, crashkernel region size] and then the kdump kernel@@ -46,14 +48,14 @@ passed to the dump-capture kernel through the elfcorehdr= boot parameter. Optionally the size of the ELF header can also be passed when 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 can -write out using file copy commands such as cp or scp. Further, you can -use analysis tools such as the GNU Debugger (GDB) and the Crash tool to -debug the dump file. This method ensures that the dump pages are correctly -ordered. - +write out using file copy commands such as cp or scp. You can also use +makedumpfile utility to analyze and write out filtered contents with +options, 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 Crash +tool to debug the dump file. This method ensures that the dump pages are +correctly ordered. Setup and Installation ======================@@ -125,9 +127,18 @@ dump-capture kernels for enabling kdump support. System kernel config options ---------------------------- -1) Enable "kexec system call" in "Processor type and features.":: +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=y + CONFIG_KEXEC_CORE=y + + Subsequently, CRASH_CORE is selected by KEXEC_CORE:: + + CONFIG_CRASH_CORE=y 2) Enable "sysfs file system support" in "Filesystem" -> "Pseudo filesystems." This is usually enabled by default::@@ -175,17 +186,19 @@ Dump-capture kernel config options (Arch Dependent, i386 and x86_64) CONFIG_HIGHMEM4G -2) On i386 and x86_64, disable symmetric multi-processing support - under "Processor type and features":: +2) With CONFIG_SMP=y, usually nr_cpus=1 need specified on the kernel + command line when loading the dump-capture kernel because one + CPU is enough for kdump kernel to dump vmcore on most of systems. - CONFIG_SMP=n + However, you can also specify nr_cpus=X to enable multiple processors + in kdump kernel. In this case, "disable_cpu_apicid=" is needed to + tell kdump kernel which cpu is 1st kernel's BSP. Please refer to + admin-guide/kernel-parameters.txt for more details. - (If CONFIG_SMP=y, then specify maxcpus=1 on the kernel command line - when loading the dump-capture kernel, see section "Load the Dump-capture - Kernel".) + With CONFIG_SMP=n, the above things are not related. -3) If one wants to build and use a relocatable kernel, - Enable "Build a relocatable kernel" support under "Processor type and +3) A relocatable kernel is suggested to be built by default. If not yet, + enable "Build a relocatable kernel" support under "Processor type and features":: CONFIG_RELOCATABLE=y@@ -241,10 +254,6 @@ Dump-capture kernel config options (Arch Dependent, ia64) crashkernel=256M - If the start address is specified, note that the start address of the - kernel will be aligned to 64Mb, so if the start address is not then - any space below the alignment point will be wasted. - Dump-capture kernel config options (Arch Dependent, arm) ----------------------------------------------------------@@ -260,46 +269,81 @@ Dump-capture kernel config options (Arch Dependent, arm64) on non-VHE systems even if it is configured. This is because the CPU will not be reset to EL2 on panic. -Extended crashkernel syntax +crashkernel syntax =========================== +1) crashkernel=size@offset -While the "crashkernel=size[@offset]" syntax is sufficient for most -configurations, sometimes it's handy to have the reserved memory dependent -on the value of System RAM -- that's mostly for distributors that pre-setup
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, arm64) · line 282
Quoted 10 lines
Documentation/admin-guide/kdump/kdump.rst +6
-3-the kernel command line to avoid a unbootable system after some memory has -been removed from the machine. + Here 'size' specifies how much memory to reserve for the dump-capture kernel + and 'offset' specifies the beginning of this reserved memory. For example, + "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory + starting at physical address 0x01000000 (16MB) for the dump-capture kernel. -The syntax is:: + The crashkernel region can be automatically placed by the system + kernel at run time. This is done by specifying the base address as 0,
s/run time/runtime
Quoted 41 lines
Documentation/admin-guide/kdump/kdump.rst +22
-11+ or omitting it all together:: - crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] - range=start-[end] + crashkernel=256M@0 + or:: -For example:: + crashkernel=256M - crashkernel=512M-2G:64M,2G-:128M + If the start address is specified, note that the start address of the + kernel will be aligned to a value (which is Arch dependent), so if the + start address is not then any space below the alignment point will be + wasted. -This would mean: +2) range1:size1[,range2:size2,...][@offset] - 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 + While the "crashkernel=size[@offset]" syntax is sufficient for most + configurations, sometimes it's handy to have the reserved memory dependent + on the value of System RAM -- that's mostly for distributors that pre-setup + the kernel command line to avoid a unbootable system after some memory has + been removed from the machine. + The syntax is:: + crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] + range=start-[end] -Boot into System Kernel -======================= + For example:: + + crashkernel=512M-2G:64M,2G-:128M + + This would mean:
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, arm64) · line 322
Quoted 10 lines
Documentation/admin-guide/kdump/kdump.rst +9
-0+ 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 to + fulfill that. With it, physical memory is allowed to allocate from top,
s/allocate/be allocated
Documentation/admin-guide/kdump/kdump.rst Dump-capture kernel config options (Arch Dependent, arm64) · line 332
Quoted 10 lines
Documentation/admin-guide/kdump/kdump.rst +10
-0+ 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 memory + region above 4G, low memory under 4G is needed in this case. There are + three ways to get low memory: + + 1) Kernel will allocate at least 256M memory below 4G automatically + if crashkernel=Y,low is not specified. + 2) Let user specify low memory size instread.
s/instread/instead
Quoted 84 lines
Documentation/admin-guide/kdump/kdump.rst +19
-12+ 3) Specified value 0 will disable low memory allocation:: + + crashkernel=0,low + +Boot into System Kernel +----------------------- 1) Update the boot loader (such as grub, yaboot, or lilo) configuration files as necessary. -2) Boot the system kernel with the boot parameter "crashkernel=Y@X", - where Y specifies how much memory to reserve for the dump-capture kernel - and X specifies the beginning of this reserved memory. For example, - "crashkernel=64M@16M" tells the system kernel to reserve 64 MB of memory - starting at physical address 0x01000000 (16MB) for the dump-capture kernel. +2) Boot the system kernel with the boot parameter "crashkernel=Y@X". - On x86 and x86_64, use "crashkernel=64M@16M". + On x86 and x86_64, use "crashkernel=Y[@X]". Most of the time, the + start address 'X' is not necessary, kernel will search a suitable + area. Unless an explicit start address is expected. On ppc64, use "crashkernel=128M@32M".@@ -331,8 +375,8 @@ of dump-capture kernel. Following is the summary. For i386 and x86_64: - - Use vmlinux if kernel is not relocatable. - Use bzImage/vmlinuz if kernel is relocatable. + - Use vmlinux if kernel is not relocatable. For ppc64:@@ -392,7 +436,7 @@ loading dump-capture kernel. For i386, x86_64 and ia64: - "1 irqpoll maxcpus=1 reset_devices" + "1 irqpoll nr_cpus=1 reset_devices" For ppc64:@@ -400,7 +444,7 @@ For ppc64: For s390x: - "1 maxcpus=1 cgroup_disable=memory" + "1 nr_cpus=1 cgroup_disable=memory" For arm:@@ -408,7 +452,7 @@ For arm: For arm64: - "1 maxcpus=1 reset_devices" + "1 nr_cpus=1 reset_devices" Notes on loading the dump-capture kernel:@@ -488,6 +532,10 @@ the following command:: cp /proc/vmcore <dump-file> +You can also use makedumpfile utility to write out the dump file +with specified options to filter out unwanted contents, e.g:: + + makedumpfile -l --message-level 1 -d 31 /proc/vmcore <dump-file> Analysis ========@@ -535,8 +583,7 @@ This will cause a kdump to occur at the add_taint()->panic() call. Contact ======= -- Vivek Goyal ([email protected]) -- Maneesh Soni ([email protected]) +- [email protected] GDB macros ==========Signature · 1 line
2.17.2
Thanks
Dave