Re: [PATCH 4.14 023/159] mm/sparsemem: Allocate mem_section at runtime for CONFIG_SPARSEMEM_EXTREME=y

From: Dou Liyang
Date: Wed Feb 07 2018 - 20:44:42 EST


Hi Baoquan,

At 02/08/2018 09:23 AM, Baoquan He wrote:
On 02/08/18 at 09:14am, Dou Liyang wrote:
Hi Baoquan,

At 02/07/2018 08:45 PM, Baoquan He wrote:
On 02/07/18 at 08:34pm, Dou Liyang wrote:


At 02/07/2018 08:27 PM, Baoquan He wrote:
On 02/07/18 at 08:17pm, Dou Liyang wrote:
Hi Baoquan,

At 02/07/2018 08:08 PM, Baoquan He wrote:
On 02/07/18 at 08:00pm, Dou Liyang wrote:
Hi Kirill,Mike

At 02/07/2018 06:45 PM, Mike Galbraith wrote:
On Wed, 2018-02-07 at 13:41 +0300, Kirill A. Shutemov wrote:
On Wed, Feb 07, 2018 at 05:25:05PM +0800, Dou Liyang wrote:
Hi All,

I met the makedumpfile failed in the upstream kernel which contained
this patch. Did I missed something else?

None I'm aware of.

Is there a reason to suspect that the issue is related to the bug this patch
fixed?


I did a contrastive test by my colleagues Indoh's suggestion.

OK, I may get the reason. kaslr is enabled, right? You can try to

I add 'nokaslr' to disable the KASLR feature.
~~~added??

oops! yes, the kaslr had already disabled by this option when I tested.


# cat /proc/cmdline
BOOT_IMAGE=/vmlinuz-4.15.0+ root=UUID=10f10326-c923-4098-86aa-afed5c54ee0b
ro crashkernel=512M rhgb console=tty0 console=ttyS0 nokaslr LANG=en_US.UTF-8

disable kaslr and try them again. Because phys_base and kaslr_offset are
got from vmlinux, while these are generated at compiling time. Just a
guess.


Oh, I will recompile the kernel with KASLR disabled in .config.

Then it's not what I guessed. Need debug makedumpfile since using
vmlinux is another code path, few people use it usually.


Understood, I will try to look into it.

Thanks,
dou



Thanks,
dou.

Revert your two commits:

commit 83e3c48729d9ebb7af5a31a504f3fd6aff0348c4
Author: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Date: Fri Sep 29 17:08:16 2017 +0300

commit 629a359bdb0e0652a8227b4ff3125431995fec6e
Author: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Date: Tue Nov 7 11:33:37 2017 +0300

...and keep others unchanged, the makedumpfile works well.

Still works fine for me with .today. Box is only 16GB desktop box though.

Btw, In the upstream kernel which contained this patch, I did two tests:

1) use the makedumpfile as core_collector in /etc/kdump.conf, then
trigger the process of kdump by echo 1 >/proc/sysrq-trigger, the
makedumpfile works well and I can get the vmcore file.

......It is OK

2) use cp as core_collector, do the same operation to get the vmcore file.
then use makedumpfile to do like above:

[douly@localhost code]$ ./makedumpfile -d 31 --message-level 31 -x
vmlinux_4.15+ vmcore_4.15+_from_cp_command vmcore_4.15+

Oh, then please ignore my previous comment. Adding '-D' can give more
debugging message.

I added '-D', Just like before, no more debugging message:

BTW, I use crash to analyze the vmcore file created by 'cp' command.

./crash ../makedumpfile/code/vmcore_4.15+_from_cp_command
../makedumpfile/code/vmlinux_4.15+

the crash works well, It's so interesting.

Thanks,
dou.

The debugging message with '-D':

And what's the debugging printing when trigger crash by sysrq?


kdump: dump target is /dev/vda2
kdump: saving to /sysroot//var/crash/127.0.0.1-2018-02-07-07:31:56/
[ 2.751352] EXT4-fs (vda2): re-mounted. Opts: data=ordered
kdump: saving vmcore-dmesg.txt
kdump: saving vmcore-dmesg.txt complete
kdump: saving vmcore
sadump: does not have partition header
sadump: read dump device as unknown format
sadump: unknown format
LOAD (0)
phys_start : 1000000
phys_end : 2a86000
virt_start : ffffffff81000000
virt_end : ffffffff82a86000
LOAD (1)
phys_start : 1000
phys_end : 9fc00
virt_start : ffff880000001000
virt_end : ffff88000009fc00
LOAD (2)
phys_start : 100000
phys_end : 13000000
virt_start : ffff880000100000
virt_end : ffff880013000000
LOAD (3)
phys_start : 33000000
phys_end : 7ffd7000
virt_start : ffff880033000000
virt_end : ffff88007ffd7000
Linux kdump
page_size : 4096

max_mapnr : 7ffd7

Buffer size for the cyclic mode: 131061

num of NODEs : 1


Memory type : SPARSEMEM_EX

mem_map (0)
mem_map : ffffea0000000000
pfn_start : 0
pfn_end : 8000
mem_map (1)
mem_map : ffffea0000200000
pfn_start : 8000
pfn_end : 10000
mem_map (2)
mem_map : ffffea0000400000
pfn_start : 10000
pfn_end : 18000
mem_map (3)
mem_map : ffffea0000600000
pfn_start : 18000
pfn_end : 20000
mem_map (4)
mem_map : ffffea0000800000
pfn_start : 20000
pfn_end : 28000
mem_map (5)
mem_map : ffffea0000a00000
pfn_start : 28000
pfn_end : 30000
mem_map (6)
mem_map : ffffea0000c00000
pfn_start : 30000
pfn_end : 38000
mem_map (7)
mem_map : ffffea0000e00000
pfn_start : 38000
pfn_end : 40000
mem_map (8)
mem_map : ffffea0001000000
pfn_start : 40000
pfn_end : 48000
mem_map (9)
mem_map : ffffea0001200000
pfn_start : 48000
pfn_end : 50000
mem_map (10)
mem_map : ffffea0001400000
pfn_start : 50000
pfn_end : 58000
mem_map (11)
mem_map : ffffea0001600000
pfn_start : 58000
pfn_end : 60000
mem_map (12)
mem_map : ffffea0001800000
pfn_start : 60000
pfn_end : 68000
mem_map (13)
mem_map : ffffea0001a00000
pfn_start : 68000
pfn_end : 70000
mem_map (14)
mem_map : ffffea0001c00000
pfn_start : 70000
pfn_end : 78000
mem_map (15)
mem_map : ffffea0001e00000
pfn_start : 78000
pfn_end : 7ffd7
mmap() is available on the kernel.
Copying data : [100.0 %] - eta: 0s
Writing erase info...
offset_eraseinfo: 9567fb0, size_eraseinfo: 0
kdump: saving vmcore complete

Thanks,
dou


[douly@localhost code]$ ./makedumpfile -D -d 31 --message-level 31 -x
vmlinux_4.15+ vmcore_4.15+_from_cp_command vmcore_4.15+
sadump: does not have partition header
sadump: read dump device as unknown format
sadump: unknown format
LOAD (0)
phys_start : 1000000
phys_end : 2a86000
virt_start : ffffffff81000000
virt_end : ffffffff82a86000
LOAD (1)
phys_start : 1000
phys_end : 9fc00
virt_start : ffff880000001000
virt_end : ffff88000009fc00
LOAD (2)
phys_start : 100000
phys_end : 13000000
virt_start : ffff880000100000
virt_end : ffff880013000000
LOAD (3)
phys_start : 33000000
phys_end : 7ffd7000
virt_start : ffff880033000000
virt_end : ffff88007ffd7000
Linux kdump
page_size : 4096

max_mapnr : 7ffd7

Buffer size for the cyclic mode: 131061
The kernel version is not supported.
The makedumpfile operation may be incomplete.

num of NODEs : 1


Memory type : SPARSEMEM_EX

mem_map (0)
mem_map : ffff88007ff26000
pfn_start : 0
pfn_end : 8000
mem_map (1)
mem_map : 0
pfn_start : 8000
pfn_end : 10000
mem_map (2)
mem_map : 0
pfn_start : 10000
pfn_end : 18000
mem_map (3)
mem_map : 0
pfn_start : 18000
pfn_end : 20000
mem_map (4)
mem_map : 0
pfn_start : 20000
pfn_end : 28000
mem_map (5)
mem_map : 0
pfn_start : 28000
pfn_end : 30000
mem_map (6)
mem_map : 0
pfn_start : 30000
pfn_end : 38000
mem_map (7)
mem_map : 0
pfn_start : 38000
pfn_end : 40000
mem_map (8)
mem_map : 0
pfn_start : 40000
pfn_end : 48000
mem_map (9)
mem_map : 0
pfn_start : 48000
pfn_end : 50000
mem_map (10)
mem_map : 0
pfn_start : 50000
pfn_end : 58000
mem_map (11)
mem_map : 0
pfn_start : 58000
pfn_end : 60000
mem_map (12)
mem_map : 0
pfn_start : 60000
pfn_end : 68000
mem_map (13)
mem_map : 0
pfn_start : 68000
pfn_end : 70000
mem_map (14)
mem_map : 0
pfn_start : 70000
pfn_end : 78000
mem_map (15)
mem_map : 0
pfn_start : 78000
pfn_end : 7ffd7
mmap() is available on the kernel.
Checking for memory holes : [100.0 %] | STEP
[Checking for memory holes ] : 0.000014 seconds
__vtop4_x86_64: Can't get a valid pte.
readmem: Can't convert a virtual address(ffff88007ffd7000) to physical
address.
readmem: type_addr: 0, addr:ffff88007ffd7000, size:32768
__exclude_unnecessary_pages: Can't read the buffer of struct page.
create_2nd_bitmap: Can't exclude unnecessary pages.
Checking for memory holes : [100.0 %] \ STEP
[Checking for memory holes ] : 0.000006 seconds
Checking for memory holes : [100.0 %] - STEP
[Checking for memory holes ] : 0.000004 seconds
__vtop4_x86_64: Can't get a valid pte.
readmem: Can't convert a virtual address(ffff88007ffd7000) to physical
address.
readmem: type_addr: 0, addr:ffff88007ffd7000, size:32768
__exclude_unnecessary_pages: Can't read the buffer of struct page.
create_2nd_bitmap: Can't exclude unnecessary pages.

makedumpfile Failed.



......It causes makedumpfile failed.


Thanks,
dou.

-Mike





















_______________________________________________
kexec mailing list
kexec@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/kexec