Security/Sandbox/2016-12-22

2016-12-22

bug 1309394 - Introduce automated tests to validate content process sandboxing works as intended
- Got review feedback from Bob to work on
bug 1322370 - Disable camera access in the Mac content sandbox
- Got a test semi-working to validate camera access no longer works
- Out for review
bug 1322716 - GMP Security bug
- Out for review

bug 1273372 - [EME] Crash in mozilla::gmp::GMPChild::ProcessingError - (Applocker)
- on inbound
bug 1321020 - When you open a new file content tab from the file content process the wrong remote type gets set.
- patches r+ - just need to addresses test review comments before landing
bug 1324908 - [e10s] OS X printing related crashes in CoreGraphics@0x regressing in Firefox 51
- looked like it might be down to bug 1308259, but it's not so clear now
- I can reproduce a similar crash on beta, but not when built locally or on try
bug 1321522 - Crash in mozilla::gfx::RecordedScaledFontCreation::PlayEvent
- uplifted to beta
bug 1321256 - Run firefox 50 from a network drive (not working anymore)
- uplifted to beta

bug 1323750 - Flash Stage3D fails on youngjump.jp in 64-bit Firefox
- wmode=direct doesn't work with async painting
- 64-bit blocker
bug 1185472 - Only allow NPAPI HWNDs to be adopted by an HWND in the chrome process
- Ready to land
bug 1284897 - 64 bit Flash Player has storage permissions issues
- addressing reviews
bug 1321493 - NPAPI sandbox is blocking Flash SecureSocket from using Windows certificate APIs on Win64
- Scope of API usage in Flash is not fully known. cpeterson proposes we just break it.
- 64-bit blocker
- Jimm to ask adobe for information on how they broker these calls in 32-bit
bug 1306239 - Add pref to toggle OS X sandbox violation debugging, default off
- Just started
bug 1312788 - Add console warning and telemetry if service workers are used in the file content process
- Ready for review

bug 1322506 - the WebRTC network thing; investigated and commented; now we have needinfo
bug 1325242 - the nsProtocolProxyService bug; investigated and filed
- Should we try to work around the gconf/utime crashes somehow?

Remoting audio
- talked to Anthony Jones and Paul in #aukland (chatlog: https://pastebin.mozilla.org/8955812 )
Working on a PoC for the automated range checking with IPDL integers

Move meeting forward an hour, so it's not at 8 AM in US/Pacific?
- No, that would move it from 5 PM to 6 PM in Central Europe.
Linux and existing connections to local services (dbus, etc.) besides X11 — do we have a bug for auditing this?
- Some of this might vary by distribution (gconf vs. dconf, ???)
- Resolved: don't have a bug; jld will file
fonts in printing

Contents