Bug 213991

Summary:

Spams system logs with "Current memory footprint: 10 MB"

Product:

WebKit

Reporter:

Michael Catanzaro <mcatanzaro>

Component:

WebKitGTK

Assignee:

Michael Catanzaro <mcatanzaro>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

benjamin, bugs-noreply, cdumez, cmarcelo, darin, ews-watchlist, mcatanzaro

Priority:

Version:

WebKit Nightly Build

Hardware:

OS:

Linux

Attachments:

Description	Flags
Patch	none

Description Michael Catanzaro 2020-07-06 07:46:09 PDT

I see new logs in my journal:

Jul 06 09:44:50 chargestone-cave WebKitWebProcess[998142]: Current memory footprint: 25 MB
Jul 06 09:44:52 chargestone-cave WebKitWebProcess[998245]: Current memory footprint: 28 MB
Jul 06 09:44:53 chargestone-cave WebKitWebProcess[998350]: Current memory footprint: 8 MB
Jul 06 09:45:15 chargestone-cave WebKitWebProcess[996536]: Current memory footprint: 52 MB
Jul 06 09:45:15 chargestone-cave WebKitWebProcess[996532]: Current memory footprint: 64 MB
Jul 06 09:45:15 chargestone-cave WebKitWebProcess[996538]: Current memory footprint: 10 MB
Jul 06 09:45:20 chargestone-cave WebKitWebProcess[998142]: Current memory footprint: 25 MB
Jul 06 09:45:22 chargestone-cave WebKitWebProcess[998245]: Current memory footprint: 27 MB
Jul 06 09:45:23 chargestone-cave WebKitWebProcess[998350]: Current memory footprint: 8 MB
Jul 06 09:45:45 chargestone-cave WebKitWebProcess[996536]: Current memory footprint: 52 MB
Jul 06 09:45:45 chargestone-cave WebKitWebProcess[996532]: Current memory footprint: 64 MB
Jul 06 09:45:45 chargestone-cave WebKitWebProcess[996538]: Current memory footprint: 10 MB
Jul 06 09:45:50 chargestone-cave WebKitWebProcess[998142]: Current memory footprint: 13 MB
Jul 06 09:45:52 chargestone-cave WebKitWebProcess[998245]: Current memory footprint: 28 MB
Jul 06 09:45:53 chargestone-cave WebKitWebProcess[998350]: Current memory footprint: 8 MB

It's too much, they are printing every few seconds. This will add up.

Comment 1 Michael Catanzaro 2020-07-06 07:56:37 PDT

Let's do this only on Apple platforms (I understand expectations for log output are different there).

Comment 2 Michael Catanzaro 2020-07-06 08:01:39 PDT

Created attachment 403596 [details]
Patch

Comment 3 Darin Adler 2020-07-06 11:49:36 PDT

Comment on attachment 403596 [details]
Patch

This is OK, but I wonder what the disconnect between macOS logging and other platforms is that makes it good to log on macOS but too noisy elsewhere. That might affect other logging too.

Comment 4 EWS 2020-07-06 12:03:50 PDT

Committed r263978: <https://trac.webkit.org/changeset/263978>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 403596 [details].

Comment 5 Michael Catanzaro 2020-07-06 14:35:02 PDT

(In reply to Darin Adler from comment #3) 
> This is OK, but I wonder what the disconnect between macOS logging and other
> platforms is that makes it good to log on macOS but too noisy elsewhere.
> That might affect other logging too.

FWIW, I don't see any other problematic logs in my journal. I think most release logging is not enabled by default. Maybe it's only memory pressure info that's being logged by default? If Apple is OK with getting rid of this particular log entirely, I guess we wouldn't have any disconnect.

Comment 6 Darin Adler 2020-07-06 14:36:08 PDT

Maybe Chris Dumez knows more?