1312266 - Missing Sandboxing prefs in SeaMonkey

Status: RESOLVED FIXED

(SeaMonkey :: Preferences, defect)

Description

[Frank-Rainer Grahl (:frg)]

Attachment: Errors in console and missing content in about:support

SeaMonkey is missing a ton of Sandboxing preferences compared to Firefox. This is now showing in 2.47 and up because bug 1259601 added code to show the sandboxing status in about:support.

Comment 1

[Frank-Rainer Grahl (:frg)]

Attachment: 1312266-sandboxing-prefs.patch

This adds the missing preferences from Firefox but not sure if they are suitable for SeaMonkey (not using any plugins). I see no errors in the console with them and about:support display fine.

Comment 2

[rsx11m]

I don't know how this sandboxing works, but it should be a security feature to avoid that a plugin affects its hosting browser. Thus, mirroring what Firefox does appears to be a prudent approach unless we are missing anything that's needed for it to work properly.

Comment 3

[Ian Neal]

Comment on attachment 8803697 [details] [diff] [review]
1312266-sandboxing-prefs.patch

r/a=me

Comment 4

[Frank-Rainer Grahl (:frg)]

https://hg.mozilla.org/comm-central/rev/ab13990e4cbccdad4acae59e3552fc4213e8fe87

Should we put this into the branches too? If yes I would suggest c-a only for now and wait a week or two if this causes problems (not expecting them).

Comment 5

[Ian Neal]

Well they have been around for Firefox since 2014, so happy for your suggested route.

Comment 6

[Frank-Rainer Grahl (:frg)]

Comment on attachment 8803697 [details] [diff] [review]
1312266-sandboxing-prefs.patch

[Approval Request Comment]
Regression caused by (bug #): 
User impact if declined: about:support is missing a line and sandboxing settings may be wrong when using plugins.
Testing completed (on m-c, etc.): c-r to c-c
Risk to taking this patch (and alternatives if risky): None. Can be backed out fast if users report problems.
String changes made by this patch: --

Comment 7

[Ian Neal]

Comment on attachment 8803697 [details] [diff] [review]
1312266-sandboxing-prefs.patch

a=me for c-a re-request c-b when it has baked for a week or so

Comment 8

[Frank-Rainer Grahl (:frg)]

Will do but you wouldn't have need to worry. Didn't plan to push to beta right away:) 

https://hg.mozilla.org/releases/comm-aurora/rev/6cc9a5013cea95365d71a314f30c337ce69e5a32

Comment 9

[Nomis101]

Has SeaMonkey been ported to e10s? I don't know for Windows, but at least on macOS sandboxing depends on e10s.
https://bugzilla.mozilla.org/show_bug.cgi?id=1135388#c1

Comment 10

[Frank-Rainer Grahl (:frg)]

>> Has SeaMonkey been ported to e10s?

No, but the sandboxing code is imho older than e10s and Nightly FF might not enable e10s too if the wrong addons are installed. SeaMonkey now shows 2 in 2.49a1 but that doesn't mean much. If it causes problems with plugins we might need to adjust this globally to 0 to just keep the diags and about:support happy. Thansk for the pointer to the bugs.

Comment 11

[Philip Chee]

Comment on attachment 8803697 [details] [diff] [review]
1312266-sandboxing-prefs.patch

I compared this with the changes in firefox.js. Everything looks reasonable.

Comment 12

[Frank-Rainer Grahl (:frg)]

Comment on attachment 8803697 [details] [diff] [review]
1312266-sandboxing-prefs.patch

[Approval Request Comment]
Seems to not cause any problems in c-b to c-c so I think itc an be taken to c-r.

Comment 13

[Ian Neal]

Comment on attachment 8803697 [details] [diff] [review]
1312266-sandboxing-prefs.patch

a=me for 2.47

Comment 14

[Frank-Rainer Grahl (:frg)]

 https://hg.mozilla.org/releases/comm-release/rev/56522f871e48bce916715f7951e6e9385ae08408