Re: [PATCH v2 0/7] kexec_file, x86, powerpc: refactoring for other architecutres

[Dave Young]

On 03/09/18 at 03:46pm, Dave Young wrote:
> Hi AKASHI,
> 
> On 03/09/18 at 04:18pm, AKASHI Takahiro wrote:
> > Dave,
> > 
> > On Fri, Mar 09, 2018 at 02:44:12PM +0800, Dave Young wrote:
> > > On 03/09/18 at 02:02pm, Dave Young wrote:
> > > > On 03/08/18 at 09:05am, Dave Young wrote:
> > > > > [Cc Andrew]
> > > > > 
> > > > > On 03/06/18 at 07:22pm, AKASHI Takahiro wrote:
> > > > > > This is a preparatory patch set for adding kexec_file support on arm64.
> > > > > > 
> > > > > > It was originally included in a arm64 patch set[1], but Philipp is also
> > > > > > working on their kexec_file support on s390[2] and some changes are now
> > > > > > conflicting.
> > > > > > 
> > > > > > So these common parts were extracted and put into a separate patch set for
> > > > > > better integration. What's more, my original patch#4 was split into a few
> > > > > > small chunks for easier review after Dave's comment.
> > > > > > 
> > > > > > As such, the resulting code is basically identical with my original, and
> > > > > > the only *visible* differences are:
> > > > > >   * renamings of _kexec_kernel_image_probe() and
> > > > > >     _kimage_file_post_load_cleanup()
> > > > > >   * change one of types of arguments at prepare_elf64_headers()
> > > > > > Those, unfortunately, require a couple of trivial changes on the rest
> > > > > > (#1, #6 to #13) of my arm64 kexec_file patch set[1].
> > > > > > 
> > > > > > Patch#1 allows making a use of purgatory optional, particularly useful
> > > > > > for arm64.
> > > > > > Patch#2 commonalizes arch_kexec_kernel_{image_probe, image_load,
> > > > > > verify_sig}() and arch_kimage_file_post_load_cleanup() across architectures.
> > > > > > Patch#3-#7 is also intended to generalize parse_elf64_headers(), along with
> > > > > > exclude_mem_range(), to be made best re-use of.
> > > > > > 
> > > > > > [1] http://lists.infradead.org/pipermail/linux-arm-kernel/2018-February/561182.html
> > > > > > [2] http://lkml.iu.edu//hypermail/linux/kernel/1802.1/02596.html
> > > > > > 
> > > > > > 
> > > > > > Changes in v2 (Mar 6, 2018)
> > > > > > * use IS_ENABLED() instead of #ifdef's in kexec_calculate_store_digests() 
> > > > > > * rename _kexec_kernel_*() and make them static for now except for
> > > > > >   kexec_image_probe_default() which is actually used by arch(powerpc) code
> > > > > >   (Note: kexec_image_post_load_cleanup_default() will also be exported by
> > > > > >   arm64 patches.)
> > > > > > * change a type of kernel_map in prepare_elf64_headers() to "int"
> > > > > > * check for a total number of system ram regions in full_up_crash_elf_data()
> > > > > >   and use vzalloc() instead of vmalloc()
> > > > > > 
> > > > > > AKASHI Takahiro (7):
> > > > > >   kexec_file: make an use of purgatory optional
> > > > > >   kexec_file,x86,powerpc: factor out kexec_file_ops functions
> > > > > >   x86: kexec_file: purge system-ram walking from prepare_elf64_headers()
> > > > > >   x86: kexec_file: remove X86_64 dependency from prepare_elf64_headers()
> > > > > >   x86: kexec_file: lift CRASH_MAX_RANGES limit on crash_mem buffer
> > > > > >   x86: kexec_file: clean up prepare_elf64_headers()
> > > > > >   kexec_file, x86: move re-factored code to generic side
> > > > > > 
> > > > > >  arch/powerpc/Kconfig                        |   3 +
> > > > > >  arch/powerpc/include/asm/kexec.h            |   2 +-
> > > > > >  arch/powerpc/kernel/kexec_elf_64.c          |   2 +-
> > > > > >  arch/powerpc/kernel/machine_kexec_file_64.c |  39 +---
> > > > > >  arch/x86/Kconfig                            |   3 +
> > > > > >  arch/x86/include/asm/kexec-bzimage64.h      |   2 +-
> > > > > >  arch/x86/kernel/crash.c                     | 334 ++++++----------------------
> > > > > >  arch/x86/kernel/kexec-bzimage64.c           |   2 +-
> > > > > >  arch/x86/kernel/machine_kexec_64.c          |  45 +---
> > > > > >  include/linux/kexec.h                       |  32 ++-
> > > > > >  kernel/kexec_file.c                         | 240 +++++++++++++++++++-
> > > > > >  11 files changed, 338 insertions(+), 366 deletions(-)
> > > > > > 
> > > > > > -- 
> > > > > > 2.16.2
> > > > > > 
> > > > > 
> > > > > AKASHI, thanks for another update, looks good to me:
> > > > > 
> > > > > Acked-by: Dave Young <dyoung@xxxxxxxxxx>
> > > > 
> > > > Hold on the ack, during today's tests kexec kernel can not boot, it
> > > > hangs early.  Will see if it is caused by this series and come back
> > > > later.
> > > 
> > > AKASHI, it seems the first patch causes the hang, can you take a look?
> > 
> > Sure, but can you first tell me specifics about what you have observed?
> 
> It is a kexec load test like below:
> kexec -s -l /boot/vmlinuz-`uname -r` --reuse-cmdline
> sync
> sync
> kexec -e
> 
> And it hangs after the message "Starting new kernel".   Now I'm not sure
> how I tested the vmcore parsing before, maybe I forgot to add "-s"
> option in kexec-tools..

BTW kdump kernel also failed to boot.
I use linus tree, the git HEAD is
86f84779d8e9 (origin/master, origin/HEAD, master) Merge branch 'siginfo-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace

It is a little odd for my testing, for kdump testing I did before, I do
not need add "-s" explictly since Fedora will take it automatically in
case Secure Boot and I always enabled Secure Boot.

> 
> Thanks
> Dave
> 
> > 
> > Thanks,
> > -Takahiro AKASHI
> > 
> > 
> > > > 
> > > > > 
> > > > > Thanks
> > > > > Dave
> > > > 
> > > > Thanks
> > > > Dave