On Thu, May 21, 2020 at 11:22 AM Naresh Kamboju naresh.kamboju@linaro.org wrote:
On Thu, 21 May 2020 at 00:39, Chris Down chris@chrisdown.name wrote:
Since you have i386 hardware available, and I don't, could you please apply only "avoid stale protection" again and check if it only happens with that commit, or requires both? That would help narrow down the suspects.
Note that Naresh is running an i386 kernel on regular 64-bit hardware that most people have access to.
kernel config link, https://builds.tuxbuild.com/8lg6WQibcwtQRRtIa0bcFA/kernel.config
Do you know if the same bug shows up running a kernel with that configuration in qemu? I would expect it to, and that would make it much easier to reproduce.
I would also not be surprised if it happens on all architectures but only shows up on the 32-bit arm and x86 machines first because they have a rather limited amount of lowmem. Maybe booting a 64-bit kernel with "mem=512M" and then running "dd if=/dev/sda of=/dev/null bs=1M" will also trigger it. I did not attempt to run this myself.
Arnd