Migration to FPR8

[OlgaTPark]

Additional patches from Firefox ESR52 for FPR8 on Intel

This bug documents things took from https://github.com/classilla/tenfourfox/issues/500.

Not taking:

• [ ] M1459206 → Not in code (we don't have M1186932);
• [ ] M1465108 → Apparently caused by M1257186 that we don't have;
• [ ] M1392739 → Not-in-code since we don't have M1433609;
• [ ] M1453127 → Cannot trigger the testcase and we don't have M1266646 which introduced the bug;
• [ ] M1456189 → Not-in-code and we don't have M1264642 which introduced the bug;
• [ ] M1462912 → Not-in-code and we don't have M1451991 which introduced the bug;
• [ ] M1413868 → Windows only and unaffected;
• [ ] M1465898 → Affects Firefox 48+, according to the bug report;
• [ ] M1462682 → TenSixFox uses Skia but this is not-in-code;
• [ ] M1414829 → ESR52 is marked as affected but Mozilla wontfixed it because they cannot backport the patch;

• [ ] M1452375 → Related to the SSE3 video scaler which isn't in this fork;
• [ ] M1451297 → Related to M1388020 which is part of this fork BUT that fix was really different so, I think we only need the if (reqSize == 0 || bufSize < reqSize) fix → Nope, this is also related to the SSE3 video scaler, so ditto;

Taking:

• [x] M1458264 → WebGL is enabled in TenSixFox and TenFiveFox;
• [x] M1436241 → TenSixFox got plugins;
• [x] M1464079 → In code, we have WebRTC, but I cannot trigger the testcase;

Others:

• M1464039 (Mozilla patch https://hg.mozilla.org/releases/mozilla-esr52/rev/528d4d997bb306aee3a5e9c65ddfad0c3787e8cf) is partially fixed. The omitted part prevents a failing code from running again but this shouldn't be an issue.

Note for those who will be ever interested in what I do:
Because my fork is kinda old and I need to keep up, I won't be releasing a binary version of Ten*Fox FPR8 Intel. Don't worry, that's not meaning I'm giving up on this!