--- Comment From kowshik.j...@in.ibm.com 2024-11-26 01:36 EDT---
I had verified the patch earlier on 6.8.0-48-generic kernel(comment #26) and
the issue was still reproducible. The reason turned out to be the
"CONFIG_KFENCE" which was set to 'y' for all the architecture on Ubuntu.
Is that
--- Comment From kowshik.j...@in.ibm.com 2024-10-23 11:25 EDT---
(In reply to comment #34)
> Kowshik, please try increasing the memory reserved for fadump from
> 512M to 768M or 1024M and see if that results in a successful dump
> capture..
I tried FADump with both 768M and 1024M reserved
--- Comment From hbath...@in.ibm.com 2024-10-22 14:01 EDT---
Note that kfence fix brings down the memory requirement for fadump
capture kernel but does not guarantee a success dump capture. So,
fadump capture kernel hitting OOM in this case does not have to be
considered a test fail. Also,
--- Comment From kowshik.j...@in.ibm.com 2024-10-22 04:51 EDT---
I think it is better to leave the commit as is and we can think of pros and
cons with setting CONFIG_KFENCE to 'n' for ppc64el.
What do you say @Hari Kirshna, @Hariharan ?
--
You received this bug notification because you
--- Comment From kowshik.j...@in.ibm.com 2024-10-16 07:56 EDT---
(In reply to comment #26)
> root@ubuntu2404:~# cat /proc/cmdline
> BOOT_IMAGE=/vmlinux-6.8.0-48-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv
> ro fadump=on fadump_reserve_mem=512M
>
>
> root@ubuntu2404:~# dmesg | grep -i re
--- Comment From kowshik.j...@in.ibm.com 2024-10-16 07:47 EDT---
I have verified bug on 6.8.0-48-generic kernel and the issue is still getting
recreated.
My system had 15G Memory and according to the recommendation here:
https://wiki.ubuntu.com/ppc64el/Recommendations I had reserved 512M
--- Comment From sthou...@in.ibm.com 2024-10-16 02:17 EDT---
Test team will be able to conclude and validate the bug soon as possible. will
update the bug.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad
--- Comment From sthou...@in.ibm.com 2024-07-23 08:55 EDT---
(In reply to comment #19)
> Fix accepted upstream (powerpc/next):
>
> https://git.kernel.org/powerpc/c/353d7a84c214f
> ("powerpc/64s/radix/kfence: map __kfence_pool at page granularity")
Ubuntu team, let us know if these patc
--- Comment From hbath...@in.ibm.com 2024-07-08 04:19 EDT---
Fix accepted upstream (powerpc/next):
https://git.kernel.org/powerpc/c/353d7a84c214f
("powerpc/64s/radix/kfence: map __kfence_pool at page granularity")
--
You received this bug notification because you are a member of Ubuntu
B
--- Comment From hbath...@in.ibm.com 2024-06-19 06:54 EDT---
Posted v2 upstream:
https://lore.kernel.org/linuxppc-dev/20240619104919.20772-1-hbath...@linux.ibm.com/
("[PATCH v2] radix/kfence: map __kfence_pool at page granularity")
--
You received this bug notification because you are a
--- Comment From hbath...@in.ibm.com 2024-04-24 12:23 EDT---
Posted the patches to reduce memory consumption for KFENCE upstream:
https://lore.kernel.org/all/20240424110926.184077-1-hbath...@linux.ibm.com/
("[PATCH 1/2] radix/kfence: map __kfence_pool at page granularity")
https://lore.ke
--- Comment From hbath...@in.ibm.com 2024-04-08 07:09 EDT---
CONFIG_KFENCE is the config option that is increasing the memory requirement
significantly for radix MMU.
For radix MMU case, memory is direct mapped with 2MB size.
But when KFENCE is used, direct mapping is done at pagesize gran
12 matches
Mail list logo
