Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 813225 - sys-kernel/gentoo-sources-5.10.62 with net-fs/nfs-utils-2.5.4: kernel rip if i reboot main PC with a diskless node
Summary: sys-kernel/gentoo-sources-5.10.62 with net-fs/nfs-utils-2.5.4: kernel rip if ...
Status: RESOLVED UPSTREAM
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: Current packages (show other bugs)
Hardware: AMD64 Linux
: Normal normal
Assignee: Gentoo Kernel Bug Wranglers and Kernel Maintainers
URL: https://bugzilla.kernel.org/show_bug....
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2021-09-15 10:01 UTC by malabay
Modified: 2022-02-18 22:47 UTC (History)
2 users (show)

See Also:
Package list:
Runtime testing required: ---

Attachments
Config of pxe (pxedefault.cfg,167 bytes, text/plain)
2021-09-15 10:03 UTC, malabay 		Details
fstab on master (binding) (fstab,338 bytes, text/plain)
2021-09-15 10:04 UTC, malabay 		Details
exportfs on master for the diskless node (exports,675 bytes, text/plain)
2021-09-15 10:04 UTC, malabay 		Details
Tracelog of crash kernel (tracelog,5.53 KB, text/plain)
2021-09-15 10:05 UTC, malabay 		Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description malabay 2021-09-15 10:01:40 UTC 
I builded a net boot of a diskless node with PXE and NFS, when a part of locals directories (/bin,/sbin,/lib,/lib64) mounting to directories of a diskless node with mode --bind. The diskless node run and work succesfully , but if reboot a main host, then get a kernel crash on a main host:

Sep 15 15:38:00 lenovo shutdown[2778]: shutting down for system reboot
Sep 15 15:38:00 lenovo init: Switching to runlevel: 6
Sep 15 15:38:00 lenovo init: Trying to re-exec init
Sep 15 15:37:59 lenovo kernel: BUG: kernel NULL pointer dereference, address: 0000000000000010
Sep 15 15:37:59 lenovo kernel: #PF: supervisor read access in kernel mode
Sep 15 15:37:59 lenovo kernel: #PF: error_code(0x0000) - not-present page
Sep 15 15:37:59 lenovo kernel: PGD 0 P4D 0
Sep 15 15:37:59 lenovo kernel: Oops: 0000 [#1] SMP NOPTI
Sep 15 15:37:59 lenovo kernel: CPU: 7 PID: 2805 Comm: kworker/u16:5 Not tainted 5.10.61-gentoo-x86_64 #2
(See the full tracelog in a attachment)

Reproducible: Always

Steps to Reproduce:
Steps:

1. Installing a main node(kernel-5.10.62,nfs-utils-2.5.4,dnsmasq)
2. Configuring PXE boot of the diskless node
    (see the  pxelinux.cfg file in a attachment )
3. Preparing the file system structure for a diskless node 
4. Mount --bind of directories for the diskless node (/bin,/sbin,/lib,/lib64) on the main host
(see the fstab file in a attachment)
5. Configuring  exportfs on the main host 
(see the exports file in a attachment )
6. Booting and running of the diskless node  is sucessfully
7. If i reboot of the main host then get message on the main host about  rip kernel:
Sep 15 15:37:59 lenovo kernel: BUG: kernel NULL pointer dereference, address: 0000000000000010
Sep 15 15:37:59 lenovo kernel: #PF: supervisor read access in kernel mode
Sep 15 15:37:59 lenovo kernel: #PF: error_code(0x0000) - not-present page

(see the full tracelog in a attachment) 



On kernel 4.19.97 this problem not existed
Comment 1 malabay 2021-09-15 10:03:42 UTC 
Created attachment 739389 [details]
Config of pxe
Comment 2 malabay 2021-09-15 10:04:19 UTC 
Created attachment 739392 [details]
fstab on master (binding)
Comment 3 malabay 2021-09-15 10:04:58 UTC 
Created attachment 739395 [details]
exportfs on master for the diskless node
Comment 4 malabay 2021-09-15 10:05:53 UTC 
Created attachment 739398 [details]
Tracelog of crash kernel
Comment 5 Matt Turner gentoo-dev 2021-09-25 21:32:50 UTC 
Sounds like you should bisect the kernel and report an issue upstream.
Comment 6 Mike Pagano gentoo-dev 2022-02-18 22:47:36 UTC 
We'll follow the upstream bug and backport any fixes identified.