MozillaZine

The Official 20170730 builds are out

Discussion about official Mozilla Firefox builds
Josa
 
Posts: 2657
Joined: July 28th, 2009, 4:52 pm

Post Posted July 29th, 2017, 8:25 am

Previous Nightly Builds thread 20170729

The Official 20170730 NIGHTLY builds are out
.The Official Firefox 55 RC1 builds are not yet out

Previous NIGHTLY: 20170729 (Fx 56.0a1)
Previous BETA: 20170728 (Fx 55.0b13)

Tinderbox Builds
Treeherder
Release Notes: NIGHTLY | BETA
Changelogs: NIGHTLY | BETA


Fixed:
    NIGHTLY +49
  1. #1378374
  2. #1385481 [Core:Audio/Video: cubeb]-Update cubeb from upstream to 42da8d2 [Uns][]
  3. #1383653 [Core:Audio/Video: Playback]-[Shutdown Decoder] video can only be suspended once. [All][]
  4. #1324619 [Core:CSS Parsing and Computation]-stylo: support ::first-line [Uns][]
  5. #1381690 [Core:CSS Parsing and Computation]-stylo: Have check_allowed_in read the Gecko bool table directly [Uns][]
  6. #1382568 [Core:CSS Parsing and Computation]-stylo: crash in Servo_ResolveStyle loading Twitter [Uns][]
  7. #1384275 [Core:CSS Parsing and Computation]-stylo: icon fonts are not loaded on fluentcpp.com [Uns][]
  8. #1384410 [Core:CSS Parsing and Computation]-Stylo: wrong rotate angle in animation [Uns][]
  9. #1385529 [Core:CSS Parsing and Computation]-stylo: Remove Gecko_GetMediaFeatures and access nsMediaFeatures::features directly instead [Uns][]
  10. #1380012 [Core:Document Navigation]-Nightly: URL not updated in URL bar after a cross-domain redirect, document.URL neither (session store doesn't contain ResultPrincipalURI) [Win][]
  11. #1376695 [Core:DOM]-nsHTMLDocument::MatchNameAttribute() shows up in profiles due to the search on the name attribute [Uns][]
  12. #1385369 [Core:DOM]-Reserve 1 space in Selection::mRanges [All][]
  13. #1385324 [Core:DOM]-HTMLFormControlsCollection::GetSortedControls() is very malloc heavy [Uns][]
  14. #1383330 [Core:DOM: Content Processes]-Use current thread instead of main thread for ContentChild recursion check [Uns][]
  15. #1384929 [Core:Graphics]-Crash in mozilla::gfx::DrawTarget::IntoLuminanceSource [Win][]
  16. #1385368 [Core:Graphics: Layers]-Error building DebugGLDrawData::DebugGLDrawData() with static analysis [Uns][]
  17. #1377118 [Core:ImageLib]-Update in-tree libpng to version 1.6.31 [All][]
  18. #1367791 [Core:JavaScript Engine]-Constrain toFixed precision to 0 <= n <= 100 [Uns][]
  19. #1379717 [Core:JavaScript Engine]-Throw TypeError when accessing "caller/arguments" properties on accessor methods [Uns][]
  20. #1382306 [Core:JavaScript Engine]-async function not working in the module type script [Uns][]
  21. #1384244 [Core:JavaScript Engine]-MaybeHasInterestingSymbolProperty has an unnecessary check for hasDynamicPrototype() [Uns][]
  22. #1384397 [Core:JavaScript Engine]-Allow scheduling CanWait interrupt callbacks [Uns][]
  23. #1384399 [Core:JavaScript Engine]-Change meaning of js::SystemZoneAvailable [Uns][]
  24. #1383645 [Core:JavaScript Engine: JIT]-UnsafeGetReservedSlot not inlined for regexp methods when inlined from string methods [Uns][]
  25. #1382656 [Core:JavaScript: Internationalization API]-Don't search for multiple Unicode extension sequences in a single language tag [Uns][]
  26. #1384395 [Core:MFBT]-Support cooperative threading in mfbt/WeakPtr.h thread safety assertions [Uns][]
  27. #1367551 [Core:Networking: HTTP]-Cancel HTTP2 push when the resource is already in the disk cache [All][[necko-active][qf:p1][platform-rel-Facebook], parity-chrome]
  28. #1264811 [Core:Security: Process Sandboxing]-Clean up OS X sandbox setup code related to MacSandboxInfo [Uns][sb+]
  29. #1366697 [Core:Security: Process Sandboxing]-Let Windows level 3 content process sandbox ride the trains. [Win][sbwc2]
  30. #1369669 [Core:Security: Process Sandboxing]-Unable to preview files from google drive [Win][sbwc2]
  31. #1384718 [Core:Security: Process Sandboxing]-WebGL doesn't initialize with Mesa 17.1 on Intel HD4600 [Lin][sb+]
  32. #1306156 [Core:WebVR]-[webvr] Use telemetry for tracking usage of WebVR API [Win][[gfx-noted][webvr]]
  33. #1383328 [Core:XPCOM]-Label nsBrowserStatusFilter timer with TabGroup of tab that owns the filter [Uns][]
  34. #1385413 [Core:XPCOM]-Precursor patches to bug 1382922 (event loop refactoring) [Uns][]
  35. #1385133 [Firefox:Activity Streams: Newtab]-Add Snippets v5 support, responsive layout and bug fixes to Activity Stream [Uns][]
  36. #1384976 [Firefox:Developer Tools: CSS Rules Inspector]-Remove the height transition from the pseudo class panel [Uns][]
  37. #1385421 [Firefox:Developer Tools: Debugger]-Update Debugger frontend (7/28/2017). [Uns][]
  38. #1377009 [Firefox:Preferences]-Search field should always float on the top [Uns][[photon-preference]]
  39. #1384566 [Firefox:Preferences]-Wrong Tooltip position after searching for certain keywords [All][[photon-preference][triage]]
  40. #1384813 [Firefox:Preferences]-Going to about:preferences#privacy-reports, there is no header for the content. [Uns][[photon-preference][triage]]
  41. #1379270 [Firefox:Tabbed Browser]-Urlbar is focused and selected too early on switch to new tab. [Uns][]
  42. #1377011 [Firefox:Theme]-[Photon] Update sidebar tree icons on all platforms [Uns][[photon-visual][p1]]
  43. #1383367 [Firefox:Toolbars and Customization]-Fix uninterruptable reflow in ToolbarIconColor::inferFromText [Uns][[qf:p1]]
  44. #1383338 [Shield:Add-on]-fetch+run shield studies soon after first startup [Uns][]
  45. #1385177 [Shield:Add-on]-remove check for built-in self-support [Uns][]
  46. #1381560 [Toolkit:Crash Reporting]-Panic reason string length is stored twice. [Uns][]
  47. #1385201 [Toolkit:Form Manager]-[Form Autofill] Rename "extensions.formautofill.experimental" [Uns][[form autofill:MVP]]
  48. #1357902 [Toolkit:WebExtensions: General]-WebExtensions I18n API should use the correct LocaleService data [Uns][[l10n], triaged]
  49. #1385528 [Toolkit:WebExtensions: General]-regression: WebExt's tab context menu is not added correctly [Uns][]

    Partial Landings/Diagnostic Patches:
  1. #1272697 [Core:JavaScript Engine]-Implement ReadableStream in the JavaScript Engine [Uns][]
  2. #1357517 [Firefox:General]-Consider removing Preferences.jsm [Uns][]

    Beta 13 -> RC 1 Changelog +1
  1. #1371484

Nightly 56 fixes since 20170613 (Gecko 55) ~2491
Beta 55 fixes since 20170306 (Gecko 54) ~5209

Nightly Blog
Release Calendar
Release Tracking
MozRegression Guide
Last edited by Josa on July 30th, 2017, 7:38 am, edited 1 time in total.

TheVisitor
 
Posts: 4412
Joined: May 13th, 2012, 10:43 am

Post Posted July 29th, 2017, 8:41 am

For grins I ran http://browserbench.org/Speedometer/ on chrome Version 59.0.3071.115 64bit result: 107

Edge scored 77.5

Poor IE 11 would not even start the test run :mrgreen:

streetwolf

User avatar
 
Posts: 2119
Joined: August 21st, 2011, 8:07 am
Location: NJ (USA)

Post Posted July 29th, 2017, 9:19 am

TheVisitor wrote:For grins I ran http://browserbench.org/Speedometer/ on chrome Version 59.0.3071.115 64bit result: 107

Edge scored 77.5

Poor IE 11 would not even start the test run :mrgreen:


Why the heck am I getting a 34 on Nightly x64? Do you guys who get higher scores run with e10s enabled? Is your compositing Direct3D 11 (Advanced Layers)? Running with Stylo enabled? Trying to ascertain what might be causing my low scores. My specs are high enough to score much higher.
Intel Core i7-8700K CPU @ 5GHz | Gigabyte Z370 AORUS Gaming 7 | Corsair 1000W PSU | Corsair H115i v2 CPU Cooler | Corsair 64GB RAM | EVGA GeForce GTX 1080 Ti FTW3 GAMING 11GB | DUAL ASUS PA249Q IPS 24" LCDs | 2-512GB Samsung 960 PRO M.2 | 2TB Seagate | Windows 10 Pro x64 | FIOS 1Gb Internet

TheVisitor
 
Posts: 4412
Joined: May 13th, 2012, 10:43 am

Post Posted July 29th, 2017, 9:43 am

streetwolf wrote:
TheVisitor wrote:For grins I ran http://browserbench.org/Speedometer/ on chrome Version 59.0.3071.115 64bit result: 107

Edge scored 77.5

Poor IE 11 would not even start the test run :mrgreen:


Why the heck am I getting a 34 on Nightly x64? Do you guys who get higher scores run with e10s enabled? Is your compositing Direct3D 11 (Advanced Layers)? Running with Stylo enabled? Trying to ascertain what might be causing my low scores. My specs are high enough to score much higher.


1. Yes e10s is enabled
2. Yes - advanced layers
3. Stylo is not enabled

Intel(R) HD Graphics 530 chipset - nothing fancy is the GPU category here.

geeknik

User avatar
 
Posts: 955
Joined: February 25th, 2003, 10:11 pm
Location: Oklahoma, USA

Post Posted July 29th, 2017, 9:49 am

115 with Nightly on Fedora 26 x64.
92.3 with webrender and stylo enabled.

SLK350

User avatar
 
Posts: 109
Joined: July 21st, 2011, 3:19 am

Post Posted July 29th, 2017, 9:53 am

With e10s, advanced layers, servo and webrender enabled i'm getting 88.8. :)

Edit: With all addons disabled i'm getting 120.

Edit 2: With only ublock enabled it drops to 90.

Edit 3: After whitelisting "browserbench.org/Speedometer/" in ublock options performance goes up again, even with all addons enabled.
Last edited by SLK350 on July 29th, 2017, 11:36 am, edited 3 times in total.
Fractal Design Define R5 Blackout * Intel Core i7 7700K @ 4,8 GHz * Asus Maximus IX Formula * G.SKILL DDR4 4x8GB PC4000 CL18 Trident Z * Samsung 960 Pro 512GB * Benq BL3201PT * ASUS GeForce GTX 1080 TI ROG Strix * Creative Soundblaster X7 * Windows 10 Pro x64 build 16251

streetwolf

User avatar
 
Posts: 2119
Joined: August 21st, 2011, 8:07 am
Location: NJ (USA)

Post Posted July 29th, 2017, 11:06 am

Disabled all my add-ons and got 97.1. Now to find the culprit.
Intel Core i7-8700K CPU @ 5GHz | Gigabyte Z370 AORUS Gaming 7 | Corsair 1000W PSU | Corsair H115i v2 CPU Cooler | Corsair 64GB RAM | EVGA GeForce GTX 1080 Ti FTW3 GAMING 11GB | DUAL ASUS PA249Q IPS 24" LCDs | 2-512GB Samsung 960 PRO M.2 | 2TB Seagate | Windows 10 Pro x64 | FIOS 1Gb Internet

GHM113

User avatar
 
Posts: 664
Joined: December 16th, 2015, 3:59 am
Location: Moscow, Russia

Post Posted July 29th, 2017, 11:09 am

NoScript uses synchronous content blocking API so it might be the reason.
I am an idiot. Sorry for my poor English. Current list of config tweaks: https://pastebin.com/1dCT4Cvp

Lurtz
 
Posts: 194
Joined: June 12th, 2016, 12:25 pm

Post Posted July 29th, 2017, 12:05 pm

TheVisitor wrote:For grins I ran http://browserbench.org/Speedometer/ on chrome Version 59.0.3071.115 64bit result: 107

Edge scored 77.5

Poor IE 11 would not even start the test run :mrgreen:

Mozilla is specifically optimizing for Speedometer v2 though:
https://bugzilla.mozilla.org/show_bug.cgi?id=1339557

streetwolf

User avatar
 
Posts: 2119
Joined: August 21st, 2011, 8:07 am
Location: NJ (USA)

Post Posted July 29th, 2017, 12:13 pm

After enabling my add-ons one by one and running the benchmark I cam up with these number. No add-ons resulted in a score of 97.1.

1. AutoplayStopper 0.9.8. This is the Webextensions version of FlashStopper Lost ~ 12%
2. Zoom Page WE 5.1. Lost ~ 50%
3. UBlock Origin/webext 1.13.9b3. Lost ~ 28%. If I disable uBO for the benchmark site I don't lose any points, I get a 97.1.

I'll contact the developers of these add-ons to report my findings. If anyone else uses any of these three add-ons disable them and see if you score higher.

There is a bug report about Speedometer located at https://bugzilla.mozilla.org/show_bug.cgi?id=1245279
Intel Core i7-8700K CPU @ 5GHz | Gigabyte Z370 AORUS Gaming 7 | Corsair 1000W PSU | Corsair H115i v2 CPU Cooler | Corsair 64GB RAM | EVGA GeForce GTX 1080 Ti FTW3 GAMING 11GB | DUAL ASUS PA249Q IPS 24" LCDs | 2-512GB Samsung 960 PRO M.2 | 2TB Seagate | Windows 10 Pro x64 | FIOS 1Gb Internet

Lurtz
 
Posts: 194
Joined: June 12th, 2016, 12:25 pm

Post Posted July 29th, 2017, 12:40 pm

Speedometer loses about 25% with uBlock Origin enabled in Chrome also. I don't think this is an issue with Firefox, it's more of an issue how the addon works. WebExtensions can be very expensive in terms of performance: https://twitter.com/_developit/status/8 ... 8919099392

SLK350

User avatar
 
Posts: 109
Joined: July 21st, 2011, 3:19 am

Post Posted July 29th, 2017, 12:44 pm

streetwolf wrote:After enabling my add-ons one by one and running the benchmark I cam up with these number. No add-ons resulted in a score of 97.1.

1. AutoplayStopper 0.9.8. This is the Webextensions version of FlashStopper Lost ~ 12%
2. Zoom Page WE 5.1. Lost ~ 50%
3. UBlock Origin/webext 1.13.9b3. Lost ~ 28%. If I disable uBO for the benchmark site I don't lose any points, I get a 97.1.

I'll contact the developers of these add-ons to report my findings. If anyone else uses any of these three add-ons disable them and see if you score higher.

There is a bug report about Speedometer located at https://bugzilla.mozilla.org/show_bug.cgi?id=1245279

If you whitelist the site in ublock it will perform better.
Fractal Design Define R5 Blackout * Intel Core i7 7700K @ 4,8 GHz * Asus Maximus IX Formula * G.SKILL DDR4 4x8GB PC4000 CL18 Trident Z * Samsung 960 Pro 512GB * Benq BL3201PT * ASUS GeForce GTX 1080 TI ROG Strix * Creative Soundblaster X7 * Windows 10 Pro x64 build 16251

GHM113

User avatar
 
Posts: 664
Joined: December 16th, 2015, 3:59 am
Location: Moscow, Russia

Post Posted July 29th, 2017, 1:10 pm

streetwolf wrote:After enabling my add-ons one by one and running the benchmark I cam up with these number. No add-ons resulted in a score of 97.1.

I don't like this at all. -28% and -50% is a lot.
I am an idiot. Sorry for my poor English. Current list of config tweaks: https://pastebin.com/1dCT4Cvp

streetwolf

User avatar
 
Posts: 2119
Joined: August 21st, 2011, 8:07 am
Location: NJ (USA)

Post Posted July 29th, 2017, 1:22 pm

SLK350 wrote:
streetwolf wrote:After enabling my add-ons one by one and running the benchmark I cam up with these number. No add-ons resulted in a score of 97.1.

1. AutoplayStopper 0.9.8. This is the Webextensions version of FlashStopper Lost ~ 12%
2. Zoom Page WE 5.1. Lost ~ 50%
3. UBlock Origin/webext 1.13.9b3. Lost ~ 28%. If I disable uBO for the benchmark site I don't lose any points, I get a 97.1.

I'll contact the developers of these add-ons to report my findings. If anyone else uses any of these three add-ons disable them and see if you score higher.

There is a bug report about Speedometer located at https://bugzilla.mozilla.org/show_bug.cgi?id=1245279

If you whitelist the site in ublock it will perform better.


I already mentioned this. My concern is whether the benchmark is a good predictor of how well 'normal' sites perform. I can't whitelist everything in uBO.
Intel Core i7-8700K CPU @ 5GHz | Gigabyte Z370 AORUS Gaming 7 | Corsair 1000W PSU | Corsair H115i v2 CPU Cooler | Corsair 64GB RAM | EVGA GeForce GTX 1080 Ti FTW3 GAMING 11GB | DUAL ASUS PA249Q IPS 24" LCDs | 2-512GB Samsung 960 PRO M.2 | 2TB Seagate | Windows 10 Pro x64 | FIOS 1Gb Internet

gorhill
 
Posts: 146
Joined: February 10th, 2015, 8:00 am

Post Posted July 29th, 2017, 3:37 pm

SLK350 wrote:If you whitelist the site in ublock it will perform better.

Whitelisting is extreme. There are various steps one can take before this:
- Disable cosmetic filtering on a per-site basis.
- Ignore all generic cosmetic filters from 3rd-party filter lists (default on Android).
- Disable generic cosmetic filtering on a per-site basis (using "@@||example.com^$generichide").

Github: https://github.com/gorhill/uBlock/issues/2552
Bugzilla: https://bugzilla.mozilla.org/show_bug.cgi?id=1339572

Return to Firefox Builds


Who is online

Users browsing this forum: No registered users and 2 guests