Security/Sandbox/2017-05-11

« previous week | index | next week »

bobowen

• bug 1351358 - Can't submit form to http(s) URL using POST method from a file:// page
  • Nearly there, just waiting on a couple of reviews.
• bug 1175267 - [e10s] about:addons page turns blank when opening XPI file
  • Patches up for review.
• bug 1336657 - Firefox 51.0.1 prints only blank pages
  • Patch reviewed, will land tomorrow and request uplift to Fx54.

haik

• bug 1334550 - Proxy moz-extension protocol requests to the parent process
  • Got some review comments
• bug 1361304 - [Mac] Remove /private/var read access from level 3 Content Sandbox
  • Landed
• bug 1350642 - Remove the PBrowser::Msg_GetTabCount sync IPC
  • Testing a fix, need to root cause failure
• bug 1358090 - Cleanup Mac sandbox policies considering the file content process

gcp

• bug 1308400 - Construct a file broker policy for default-deny read access on the Linux Desktop
• Progress on try orange
• Investigations regarding --appname, test js layout, SpecialPowers packaging/install, etc

Alex_Gaynor

• bug 1361733 - Disallow writing to all of /private/var in DEBUG
  • Landed
• bug 1358223 - Hardcode the lowest sandbox level for macOS and Windows
  • Implemented the changes we talked about last week
• bug 1357758 - Replace blacklisting macOS sandbox with whitelisting
  • Debugged the root cause of many of the failures
• bug 1363760 - Remote file reads of specialpowers JS to parent process
  • Working with the addons/testing folks to decide on the right course, but it looks like some work the addons folks were already doing makes fixing this easy

jld

• DBus: a11y and wakelocks should be the last holdouts
  • bug 1361338 (don't a11y in content procs) landed, so bug 1362537 (re-disallow accept4) can happen
  • bug 1360069 (wakelocks) I more or less have the patch; finding a reviewer might be interesting
• ESET AV
  • LD_AUDIT will let us block their library if we need to
• SysV
  • We might be able to stop using SysV IPC after all.
  • msg -> ESET; get rid of it
  • sem -> ALSA; off by default
  • shm -> graphics; complicated but I have a plan
    • MIT-SHM 1.2 with fd-passing
• madvise
  • It's MADV_NO_HUGEPAGE; I'll file a bug
  • Not sure if we should allow or ignore

roundtable

• No minimum level on linux
• SandboxStatus/Settings/Config class name - decided on SandboxSettings