| Message ID | 20230312181154.278900-3-sourabhjain@linux.ibm.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | PowerPC: in kernel handling of CPU hotplug events for crash kernel | expand |
On 3/12/23 13:11, Sourabh Jain wrote: > Due to CPU/Memory hotplug events the system resources changes. A similar > change should reflect in the loaded kdump kernel image that describes > the state of the CPU and memory of the running kernel. > > If the kdump kernel image is not updated after the CPU or Memory hotplug > events and it tries to collect the dump with the stale system resource > data this might lead to dump collection failure or an inaccurate dump > collection. > > The current method to keep the kdump kernel up to date is by triggering > reload (i.e unload and load) the entire kdump kernel image whenever a > CPU or Memory hotplug event is observed by udev in the userspace. > Reloading the complete kdump kernel image is an expensive task. It can > be easily avoided by doing the in-kernel updates to specific kdump > kernel image components which are responsible for describing CPU and > Memory resources of the running kernel to the kdump kernel. > > The kernel changes related to in-kernel update to the kdump kernel image > on CPU/Memory hotplug events are kept under the CRASH_HOTPLUG config > option. > > Later in the series, a powerpc crash hotplug handler is introduced to > update the kdump kernel image on CPU/Memory hotplug events. This arch > specific handler is trigger from a generic crash handler that registers > with the CPU and memory notifiers. > > The CRASH_HOTPLUG config option is enabled by default. > > Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com> > --- > arch/powerpc/Kconfig | 12 ++++++++++++ > 1 file changed, 12 insertions(+) > > diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig > index a6c4407d3ec83..2f45b3f5175cb 100644 > --- a/arch/powerpc/Kconfig > +++ b/arch/powerpc/Kconfig > @@ -681,6 +681,18 @@ config CRASH_DUMP > The same kernel binary can be used as production kernel and dump > capture kernel. > > +config CRASH_HOTPLUG > + bool "Update crash capture system on CPU/Memory hotplug event" Fwiw, online/offline changes also flow through this infrastructure... eric > + default y > + depends on CRASH_DUMP && (HOTPLUG_CPU || MEMORY_HOTPLUG) > + help > + In kernel update to relevant kexec segments due to change > + in the system configuration, rather reloading all the kexec > + segments again from userspace by monitoring CPU/Memory > + hotplug events in the userspace using udev. > + > + If unsure, say N. > + > config FA_DUMP > bool "Firmware-assisted dump" > depends on PPC64 && (PPC_RTAS || PPC_POWERNV)
On 13/03/23 21:16, Eric DeVolder wrote: > > > On 3/12/23 13:11, Sourabh Jain wrote: >> Due to CPU/Memory hotplug events the system resources changes. A similar >> change should reflect in the loaded kdump kernel image that describes >> the state of the CPU and memory of the running kernel. >> >> If the kdump kernel image is not updated after the CPU or Memory hotplug >> events and it tries to collect the dump with the stale system resource >> data this might lead to dump collection failure or an inaccurate dump >> collection. >> >> The current method to keep the kdump kernel up to date is by triggering >> reload (i.e unload and load) the entire kdump kernel image whenever a >> CPU or Memory hotplug event is observed by udev in the userspace. >> Reloading the complete kdump kernel image is an expensive task. It can >> be easily avoided by doing the in-kernel updates to specific kdump >> kernel image components which are responsible for describing CPU and >> Memory resources of the running kernel to the kdump kernel. >> >> The kernel changes related to in-kernel update to the kdump kernel image >> on CPU/Memory hotplug events are kept under the CRASH_HOTPLUG config >> option. >> >> Later in the series, a powerpc crash hotplug handler is introduced to >> update the kdump kernel image on CPU/Memory hotplug events. This arch >> specific handler is trigger from a generic crash handler that registers >> with the CPU and memory notifiers. >> >> The CRASH_HOTPLUG config option is enabled by default. >> >> Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com> >> --- >> arch/powerpc/Kconfig | 12 ++++++++++++ >> 1 file changed, 12 insertions(+) >> >> diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig >> index a6c4407d3ec83..2f45b3f5175cb 100644 >> --- a/arch/powerpc/Kconfig >> +++ b/arch/powerpc/Kconfig >> @@ -681,6 +681,18 @@ config CRASH_DUMP >> The same kernel binary can be used as production kernel and dump >> capture kernel. >> +config CRASH_HOTPLUG >> + bool "Update crash capture system on CPU/Memory hotplug event" > Fwiw, online/offline changes also flow through this infrastructure... > eric Yes I will update the CONFIG summary and commit message to convey the same. Thanks for the review. Thanks, Sourabh Jain
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index a6c4407d3ec83..2f45b3f5175cb 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -681,6 +681,18 @@ config CRASH_DUMP The same kernel binary can be used as production kernel and dump capture kernel. +config CRASH_HOTPLUG + bool "Update crash capture system on CPU/Memory hotplug event" + default y + depends on CRASH_DUMP && (HOTPLUG_CPU || MEMORY_HOTPLUG) + help + In kernel update to relevant kexec segments due to change + in the system configuration, rather reloading all the kexec + segments again from userspace by monitoring CPU/Memory + hotplug events in the userspace using udev. + + If unsure, say N. + config FA_DUMP bool "Firmware-assisted dump" depends on PPC64 && (PPC_RTAS || PPC_POWERNV)
Due to CPU/Memory hotplug events the system resources changes. A similar change should reflect in the loaded kdump kernel image that describes the state of the CPU and memory of the running kernel. If the kdump kernel image is not updated after the CPU or Memory hotplug events and it tries to collect the dump with the stale system resource data this might lead to dump collection failure or an inaccurate dump collection. The current method to keep the kdump kernel up to date is by triggering reload (i.e unload and load) the entire kdump kernel image whenever a CPU or Memory hotplug event is observed by udev in the userspace. Reloading the complete kdump kernel image is an expensive task. It can be easily avoided by doing the in-kernel updates to specific kdump kernel image components which are responsible for describing CPU and Memory resources of the running kernel to the kdump kernel. The kernel changes related to in-kernel update to the kdump kernel image on CPU/Memory hotplug events are kept under the CRASH_HOTPLUG config option. Later in the series, a powerpc crash hotplug handler is introduced to update the kdump kernel image on CPU/Memory hotplug events. This arch specific handler is trigger from a generic crash handler that registers with the CPU and memory notifiers. The CRASH_HOTPLUG config option is enabled by default. Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com> --- arch/powerpc/Kconfig | 12 ++++++++++++ 1 file changed, 12 insertions(+)