atlas-probe: version 2.4.1-2 "do_page_fault(): sending SIGSEGV to perd for invalid read access from XXXXXXXX"

[KsmoinO]

Maintainer: @ja-pa
Environment: ath79, dlink_dir-842-c3, 22.03.2

Description:
About once a day and up to 5 times a day, perd is killed with following kernel message "do_page_fault(): sending SIGSEGV to perd for invalid read access from XXXXXXXX"

As workaround I schedule this script every minutes to restart atlas each time perd is killed :

#!/bin/sh
pgrep /usr/libexec/atlas-probe/bin/perd
if [ $? -ne 0 ] ; then
  /etc/init.d/atlas restart
fi

[Ansuel]

This should be fixed in later release... pls upgrade

[BKPepe]

Wasn't this related to #19739 or it happens one more time?

[KsmoinO]

This should be fixed in later release... pls upgrade

I'm already at latest atlas-probe version (2.4.1-2) available on latest openwrt stable release (22.03.2)

Wasn't this related to #19739 or it happens one more time?

No, this time it's not 'eperd' but 'perd' which is killed

[Ansuel]

@BKPepe both were fixed so this sounds strange to me