On 24/04/23 19:35, Eric DeVolder wrote:
On 4/23/23 05:52, Sourabh Jain wrote:
The Problem:
============
Post CPU/Memory hot plug/unplug and online/offline events the kernel
holds stale information about the system. Dump collection with stale
kdump kernel might end up in dump capture failure or an inaccurate dump
collection.
Existing solution:
==================
The existing solution to keep the kdump kernel up-to-date by monitoring
CPU/Memory hotplug/online/offline events via udev rule and trigger a
full
kdump kernel reload for every hotplug event.
Shortcomings:
------------------------------------------------
- Leaves a window where kernel crash might not lead to a successful dump
collection.
- Reloading all kexec components for each hotplug is inefficient.
- udev rules are prone to races if hotplug events are frequent.
More about issues with an existing solution is posted here:
- https://lkml.org/lkml/2020/12/14/532
-
https://lists.ozlabs.org/pipermail/linuxppc-dev/2022-February/240254.html
Proposed Solution:
==================
Instead of reloading all kexec segments on CPU/Memory
hotplug/online/offline
event, this patch series focuses on updating only the relevant kexec
segment.
Once the kexec segments are loaded in the kernel reserved area then an
arch-specific hotplug handler will update the relevant kexec segment
based on
hotplug event type.
Series Dependencies
====================
This patch series implements the crash hotplug handler on PowerPC.
The generic
crash hotplug handler is introduced by
https://lkml.org/lkml/2023/4/4/1136 patch
series.
Git tree for testing:
=====================
The below git tree has this patch series applied on top of dependent
patch
series.
https://github.com/sourabhjains/linux/tree/e21-s10
To realise the feature the kdump udev rule must updated to avoid
reloading of kdump reload on CPU/Memory hotplug/online/offline events.
RHEL: /usr/lib/udev/rules.d/98-kexec.rules
-SUBSYSTEM=="cpu", ACTION=="online", GOTO="kdump_reload_cpu"
-SUBSYSTEM=="memory", ACTION=="online", GOTO="kdump_reload_mem"
-SUBSYSTEM=="memory", ACTION=="offline", GOTO="kdump_reload_mem"
+SUBSYSTEM=="cpu", ATTRS{crash_hotplug}=="1",
GOTO="kdump_reload_end"
+SUBSYSTEM=="memory", ATTRS{crash_hotplug}=="1",
GOTO="kdump_reload_end"
I didn't see in the patch series where you would have the equivalent
to the following (needed for the sysfs crash_hotplug entries):
#ifdef CONFIG_HOTPLUG_CPU
static inline int crash_hotplug_cpu_support(void) { return 1; }
#define crash_hotplug_cpu_support crash_hotplug_cpu_support
#endif
#ifdef CONFIG_MEMORY_HOTPLUG
static inline int crash_hotplug_memory_support(void) { return 1; }
#define crash_hotplug_memory_support crash_hotplug_memory_support
#endif
I missed the above diff in my testing environment. Thanks you for
bringing it
to my attention. I will fix this next version.
- Sourabh Jain
_______________________________________________
kexec mailing list
kexec@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/kexec
