Just want to mention, that Firefox writes (just) a timestamp to "recovery.js",
whenever a page is shifted during browsing. (Date+Time for the last used Url-history.)
Few pixels moved via the vertical slider up/down are sufficient, to get Clamscan working - again and again.
The actual scan of few MB is fast, but db-loading takes time.
(Few MB because I like to keep tabs around and recovery.js holds all the info about all open tab's.)

Suggestion for Clam Sentinel Users: copy the reported line (in the Clam Sentinel log)
and put it in -Advanced settings- -Paths or files not scanned- .
To have it scanned (shutdown/boot), it could be done with a batch file:
either dummy-copy to "x:\yz\*.scanme" /Y
or by calling Clamscan in a batch file.

I have not used the mem scan in Clam Sentinel for a long time. The ClamWin mem scan is more efficient.

Regards,

with the Task-Manager (Option: Always on top)
it can be seen that Clamscan starts, whenever a browsed Firefox-page is shifted.