[PATCH V6 0/6] RISC-V fixups to work with crash tool

Xianting Tian xianting.tian at linux.alibaba.com
Wed Oct 12 19:28:09 PDT 2022


在 2022/10/12 下午5:52, Conor Dooley 写道:
> On Wed, Oct 12, 2022 at 05:42:37PM +0800, Xianting Tian wrote:
>> 在 2022/8/12 上午12:17, Palmer Dabbelt 写道:
>>> On Thu, 11 Aug 2022 00:41:44 PDT (-0700),
>>> xianting.tian at linux.alibaba.com wrote:
>>>> I ever sent the patch 1 in the link:
>>>> https://patchwork.kernel.org/project/linux-riscv/patch/20220708073150.352830-3-xianting.tian@linux.alibaba.com/
>>>>
>>>> And patch 2,3 in the link:
>>>> https://patchwork.kernel.org/project/linux-riscv/patch/20220714113300.367854-2-xianting.tian@linux.alibaba.com/
>>>>
>>>> https://patchwork.kernel.org/project/linux-riscv/patch/20220714113300.367854-3-xianting.tian@linux.alibaba.com/
>>>>
>>>>
>>>> This patch set just put these patches together, and with three new
>>>> patch 4, 5, 6.
>>>> these six patches are the fixups for machine_kexec, kernel mode PC
>>>> for vmcore
>>>> and improvements for vmcoreinfo, memory layout dump and fixup
>>>> schedule out issue
>>>> in machine_crash_shutdown().
>>>>
>>>> The main changes in the six patchs as below,
>>>> Patch 1: Fixup use of smp_processor_id() in preemptible context, to
>>>> cleanup
>>>>           the console prints.
>>>> Patch 2: Fixup to get correct kernel mode PC for kernel mode regs
>>>> for vmcore.
>>>> Patch 3: Fixup schedule out issue in machine_crash_shutdown()
>>>> Patch 4: Add modules to virtual kernel memory layout dump.
>>>> Patch 5: Add VM layout, va bits, ram base to vmcoreinfo, which can
>>>> simplify
>>>>           the development of crash tool as ARM64 already did
>>>>           (arch/arm64/kernel/crash_core.c).
>>>> Patch 6: Updates vmcoreinfo.rst for vmcoreinfo export for RISCV64.
>>>>
>>>> With these six patches(patch 2 is must), crash tool can work well to
>>>> analyze
>>>> a vmcore. The patches for crash tool for RISCV64 is in the link:
>>>> https://lore.kernel.org/linux-riscv/20220801043040.2003264-1-xianting.tian@linux.alibaba.com/
>>>>
>>>>
>>>> ------
>>>> Changes v1 -> v2:
>>>>    1, remove the patch "Add a fast call path of crash_kexec()" from
>>>> this series
>>>>       of patches, as it already applied to riscv git.
>>>> https://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux.git/commit/?h=for-next&id=3f1901110a89b0e2e13adb2ac8d1a7102879ea98
>>>>    2, add 'Reviewed-by' based on the comments of v1.
>>>> Changes v2 -> v3:
>>>>    use "riscv" instead of "riscv64" in patch 5 subject line.
>>>> Changes v3 -> v4:
>>>>    use "riscv" instead of "riscv64" in the summary of patch 5 subject
>>>> line.
>>>> Changes v4 -> v5:
>>>>    add a new patch "RISC-V: Fixup schedule out issue in
>>>> machine_crash_shutdown()"
>>>> Changes v5 -> v6:
>>>>    1, move "fixup" patches to the start of the patch set.
>>>>    2, change patch 1, 2, 6's subject to make it tell more what it's
>>>> about.
>>>>    3, add Fixes for patch 3.
>>>>    4, adjuest the changes format for patch 6.
>>>>
>>>>
>>>> Xianting Tian (6):
>>>>    RISC-V: kexec: Fixup use of smp_processor_id() in preemptible context
>>>>    RISC-V: Fixup get incorrect user mode PC for kernel mode regs
>>>>    RISC-V: Fixup schedule out issue in machine_crash_shutdown()
>>>>    RISC-V: Add modules to virtual kernel memory layout dump
>>>>    RISC-V: Add arch_crash_save_vmcoreinfo support
>>>>    Documentation: kdump: describe VMCOREINFO export for RISCV64
>>>>
>>>>   .../admin-guide/kdump/vmcoreinfo.rst          | 31 +++++++++++++++++++
>>>>   arch/riscv/kernel/Makefile                    |  1 +
>>>>   arch/riscv/kernel/crash_core.c                | 29 +++++++++++++++++
>>>>   arch/riscv/kernel/crash_save_regs.S           |  2 +-
>>>>   arch/riscv/kernel/machine_kexec.c             | 28 ++++++++++++++---
>>>>   arch/riscv/mm/init.c                          |  4 +++
>>>>   6 files changed, 89 insertions(+), 6 deletions(-)
>>>>   create mode 100644 arch/riscv/kernel/crash_core.c
>>> Thank.  I've taken the first 4 onto for-next, which is still targeted
>>> for 5.20, as they're fixes.  I'm not opposed to taking the documentation
>>> patch for this cycle as well, it just needs some going-over as the
>>> wording looks very odd (or at least it does to me right now, maybe I'm
>>> just still half asleep).  Patch 5 is a new feature, and given that it's
>>> being spun during the merge window it's too late.
>> Hi Palmer
>>
>> Do you plan to merge the two patch to Linux 6.1 to support crash tool work?
>> thanks
>>
>>    RISC-V: kexec: Fixup use of smp_processor_id() in preemptible context
> 357628e68f5c ("RISC-V: kexec: Fixup use of smp_processor_id() in preemptible context"
>
>>    RISC-V: Fixup get incorrect user mode PC for kernel mode regs
> 59c026c359c3 ("RISC-V: Fixup get incorrect user mode PC for kernel mode regs")
>
> Hey Xianting, those two commits already seem to have been applied, do
> you perhaps instead mean the documentation patch and the addition of
> support for arch_crash_save_vmcoreinfo? I recalled asking if you needed
> to respin at the time, but do not see a response:
> https://lore.kernel.org/linux-riscv/39fdc85e-b2d5-863c-4878-4b3380d76bc4@microchip.com/
> IIRC Bagas had some outstanding comments on the documentation change
> too. Was I incorrect?

Sorry, It is the two patches:

   RISC-V: Add arch_crash_save_vmcoreinfo support
   Documentation: kdump: describe VMCOREINFO export for RISCV64

I saw Palmer already merged the two patches to his riscv-crash branch, 
it means they are OK?

https://git.kernel.org/pub/scm/linux/kernel/git/palmer/linux.git/log/?h=riscv-crash

I saw Bagas's comments, I will send new version patch for the two 
patches, thanks for the reminder.

>
> Thanks,
> Conor.



More information about the kexec mailing list