The first official 20171108 builds are out

Discussion about official Mozilla Firefox builds
Locked
Josa
Posts: 7360
Joined: July 28th, 2009, 4:52 pm

The first official 20171108 builds are out

Post by Josa »

Previous Nightly Builds thread 20171107

The first official 20171108 NIGHTLY builds are out
.The Official Firefox 57 RC2 builds are not yet out

Previous NIGHTLY: 20171107 (Fx 58.0a1)
Previous BETA: 20171107 (Fx 57.0 RC1)

Mozilla Inbound Builds
Treeherder
Release Notes: NIGHTLY | BETA
Changelogs: NIGHTLY | BETA


Fixed:
  1. NIGHTLY +69
  2. #1402014
  3. #1405599
  4. #1410106
  5. #1411631
  6. #1414692
  7. #1414279
  8. #1414762
  9. #1393045 [Core:Audio/Video: Playback]-Add support for AudioSampleEntry with a version 2 (QT Extension) [Uns][]
  10. #1395922 [Core:Audio/Video: Playback]-Handle HTMLMediaElement::SetMediaKeys asynchronously. [Uns][]
  11. #1412737 [Core:Audio/Video: Playback]-MediaCacheStream::Read() should read enough bytes until EOS or error is encounted [Uns][]
  12. #1413076 [Core:Audio/Video: Playback]-Return other error in H264Converter [Uns][]
  13. #1412180 [Core:Audio/Video: Playback]-[webvtt] implement region id on VTTRegion. [Uns][]
  14. #1412558 [Core:Audio/Video: Playback]-media.ffvpx.enabled=true (vp9) is too slow on 32bit Unix [Lin][]
  15. #1415086 [Core:DOM: Workers]-AddressSanitizer: stack-use-after-scope /builds/slave/moz-toolchain/src/llvm/projects/compiler-rt/lib/asan/../sanitizer_common/sanitizer_common_interceptors.inc:302:5 in __interceptor_strlen [Uns][]
  16. #1413253 [Core:DOM]-Assertion failure: !mMightHaveUnreportedJSException, at /builds/worker/workspace/build/src/obj-firefox/dist/include/mozilla/ErrorResult.h:486 [Uns][]
  17. #1415389 [Core:DOM]-Make ChromeUtils a namespace [Uns][]
  18. #1415182 [Core:DOM]-The External interface should be [NoInterfaceObject] [Uns][]
  19. #1414713 [Core:Editor]-Get rid of EditorUtils::IsDescendantOf() which returns offset [All][]
  20. #1414581 [Core:Editor]-crash near null [@ IsList] [Uns][]
  21. #1414271 [Core:Graphics: WebRender]-Update webrender to da493fad52ad28da5aa062fa18100c0fd4f30b24 [Uns][[wr-mvp] [gfx-noted]]
  22. #1411627 [Core:Graphics: WebRender]-WR API for position:sticky needs to support painting at nonzero scroll offsets [Uns][[wr-mvp] [gfx-noted]]
  23. #1414714 [Core:Graphics]-Remove some unused gfx prefs [Uns][[gfx-noted]]
  24. #1414631 [Core:IPC]-chromium: Remove the Chromium Process class [Uns][]
  25. #1415085 [Core:ImageLib]-AddressSanitizer: stack-use-after-scope /builds/worker/workspace/build/src/image/SurfaceCache.cpp:228:37 in AreaOfIntSize [Uns][[gfx-noted]]
  26. #1415161 [Core:JavaScript Engine]-JSContext::findVersion slows down CheckPatternSyntax [All][]
  27. #1339895 [Core:JavaScript Engine]-Javascript statement var x += 2 yields misleading error [Uns][]
  28. #1244280 [Core:JavaScript Engine]-Linux javascript call stack greatly limited when using baseline jit [Lin][]
  29. #1414546 [Core:JavaScript Engine]-No longer treat "yield" as keyword in js>=1.7 [Uns][]
  30. #1415124 [Core:JavaScript Engine]-Nursery allocate cloned lexical environments [All][]
  31. #1412110 [Core:Layout: Web Painting]-Assertion failure: !item->GetAbove(), at /builds/worker/workspace/build/src/layout/generic/nsFrame.cpp:3151 [Uns][]
  32. #1413833 [Core:Layout: Web Painting]-Poor performance on searchfox with retained-dl [Uns][]
  33. #1414397 [Core:Layout: Web Painting]-Retained layers cause Confluence table cell content to flicker or vanish [Uns][]
  34. #1411765 [Core:Layout]-Fix reflow status when size is exactly the available size and undo comment changes [Uns][]
  35. #1415152 [Core:Layout]-GetInsertionPrevSibling doesn't need to call AdjustAppendForAfter, since ::after is handled in FindNextSibling. [Uns][]
  36. #1409570 [Core:Networking: HTTP]-Regression: Loading site using HTTP2 server push hangs page while loading resources [Uns][[necko-triaged]]
  37. #1414188 [Core:Preferences: Backend]-Remove ValueObserver [Uns][]
  38. #1414096 [Core:Preferences: Backend]-Remove support for nsISupportsString values in nsPrefBranch::{get,set}ComplexValue() [All][]
  39. #1410546 [Core:Security: PSM]-Disable EV treatment for "Security Communication EV RootCA1" root certificate [Uns][[psm-assigned]]
  40. #1415250 [Core:Security: Process Sandboxing]-Let level 4 for the Windows content sandbox ride again. [Win][sb+]
  41. #1409622 [Core:Web Audio]-Test failures on OSX when building with clang 5 [Uns][]
  42. #1415249 [Core:WebRTC: Networking]-Remove unused variables introduced during EUI 64 (MAC-based) IPv6 candidate gathering [Uns][]
  43. #1414272 [Core:Widget: Gtk]-Implement SetNonClientMargins() for Gtk+ [Uns][]
  44. #1414150 [Core:XPCOM]-Remove some unnecessary prefs in and around XPCOM [Uns][]
  45. #1414292 [Core:XPConnect]-Update to current spec for cross-origin object property enumerability [Uns][]
  46. #1413692 [Firefox:Address Bar]-Extensions page actions are pinned to address bar again and again [Uns][[fxsearch]]
  47. #1413843 [Firefox:Bookmarks & History]-Cut/Copy/Move operations involving the left pane of the library are typically very slow due to non-batching [Uns][[fxsearch]]
  48. #1415137 [Firefox:Developer Tools: Console]-Group's vertical line is moved one pixel [Uns][]
  49. #1408790 [Firefox:Developer Tools: Inspector]-The non-selected elements with ids should be lightened to the same color as the non-selected classes [Uns][]
  50. #1400963 [Firefox:Developer Tools: JSON Viewer]-scrolling out in json viewer [Uns][]
  51. #1414609 [Firefox:Developer Tools: Netmonitor]-Filter autocomplete popup is hidden behind the floating status header [Uns][]
  52. #1413540 [Firefox:Developer Tools: Netmonitor]-Merge 'flash' filter into 'others' [Uns][]
  53. #1413832 [Firefox:Developer Tools]-Keep the green/active toolbar icon styling when the tool is selected [Uns][]
  54. #1408334 [Firefox:Developer Tools]-[dt-onboarding] Add form to subscribe to devtools newsletter [Uns][]
  55. #1414965 [Firefox:Search]-Reset search geo preferences due to bug 1413652 [Uns][]
  56. #1339329 [Firefox:Security]-Incorrect messaging and user instructions for expired SSL/TLS certificates [Uns][]
  57. #1415012 [Firefox:Sync]-Sync should not trigger immediate resync if some engines failed during the sync. [Uns][]
  58. #1414222 [Firefox:Theme]-Configure titlebar rendering on Linux/Gtk+ [Uns][]
  59. #1411309 [Firefox:Theme]-Drag space in title bar does not collapse margins [Uns][[reserve-photon-visual]]
  60. #1299286 [Firefox:Toolbars and Customization]-In touch enabled laptop drag and drop by finger is not working on toolbar and customization window [Win][[reserve-photon-visual][p4][high-priority]]
  61. #1415017 [Toolkit:Add-ons Manager]-Remove AddonRepository sqlite migration code [Uns][]
  62. #1415103 [Toolkit:Crash Reporting]-Remove dead code in crash reporting [Uns][]
  63. #1402963 [Toolkit:Form Manager]-Credit cards need deduplication [Uns][[form autofill:MVP]]
  64. #1410907 [Toolkit:Telemetry]-Add canRecordReleaseData/canRecordPrereleaseData to nsITelemetry [All][]
  65. #1409323 [Toolkit:Telemetry]-Allow to register new addon scalars for an existing category [All][]
  66. #1410143 [Toolkit:Telemetry]-It is difficult to determine which crash pings are from ShutDownKill [Uns][]
  67. #464450 [Toolkit:Themes]-remove "padded" class from type="autocomplete" textboxes without icons [All][]
  68. #1414512 [Toolkit:WebExtensions: Frontend]-browser.theme.getCurrent() always returns an empty theme object if theme doesn't originate from the extension [Uns][]
  69. #1414997 [Toolkit:WebExtensions: General]-ExtensionStorageSync might initialize Kinto multiple times [Uns][]
  70. #1291399 [Toolkit:WebExtensions: General]-Move uses of getExtensionUUID to UUIDMap [Uns][triaged]
  1. Partial Landings/Diagnostic Patches:
  2. #1415397 [Core:Audio/Video: Playback]-Use Span<> to replace low level pointer arithmetic in MediaCacheStream::ReadFromCache() [Uns][]
  3. #1412545 [Core:Graphics: WebRender]-regression: blob-images: libfreetype crashes (broken webrender_bindings) [Lin][[wr-reserve]]
  1. RC1 -> RC2 Changelog +0
    None

Nightly 58 fixes since 20170921 (Gecko 57) ~2483
Beta 57 fixes since 20170802 (Gecko 56) ~3574

Nightly Blog
Release Calendar
Release Tracking
MozRegression Guide
Last edited by Josa on November 8th, 2017, 3:40 pm, edited 1 time in total.
User avatar
Omega X
Posts: 8225
Joined: October 18th, 2007, 2:38 pm
Location: A Parallel Dimension...

Re: The first official 20171108 builds are not yet out

Post by Omega X »

The VP9 patches are in and performance is better than ever now.
gorhill
Posts: 163
Joined: February 10th, 2015, 8:00 am

Re: The first official 20171108 builds are not yet out

Post by gorhill »

I am having bout of relatively high CPU usage (~20% on quad-core) with 58.0a1 (2017-11-08) (64-bit) (linux version). This is intermittent, and unpredictable (might be happening when I open a new web page, need more data to make the case), and last for a few seconds (10s-20s very roughly estimated).

I looked into this using Gecko profiler, but it shows nothing -- all processes just show they are waiting at poll(), which is normal when the browser is idle (my understanding). The OS System Monitor shows that this is happening in the main process.

This observation is new to me -- not something I observed in previous builds. So anyway, just throwing this out there in case someone else experiences this. I became aware of this through the CPU fan noise.
User avatar
Edds
Posts: 332
Joined: July 21st, 2010, 1:13 pm
Location: SVK

Re: The first official 20171108 builds are not yet out

Post by Edds »

Hmm, maybe something is happening. For a past few builds I noticed from time to time more sluggishness e.g. loading content icon for a moment when switching from one tab to another... obviously it should be instant except some edge cases with very heavy load on tabs. I thought there was something on the background which is consuming CPU, but I'm not sure now.
Laptop: Intel Core i5 1240P | 16GB DDR4 | 500GB NVMe | 1080p | Win 11
Phone: Samsung S24 Ultra | Android 14
shellye5
Posts: 47
Joined: May 23rd, 2017, 9:57 am

Re: The first official 20171108 builds are not yet out

Post by shellye5 »

gorhill wrote:I am having bout of relatively high CPU usage (~20% on quad-core) with 58.0a1 (2017-11-08) (64-bit) (linux version). This is intermittent, and unpredictable (might be happening when I open a new web page, need more data to make the case), and last for a few seconds (10s-20s very roughly estimated).

I looked into this using Gecko profiler, but it shows nothing -- all processes just show they are waiting at poll(), which is normal when the browser is idle (my understanding). The OS System Monitor shows that this is happening in the main process.

This observation is new to me -- not something I observed in previous builds. So anyway, just throwing this out there in case someone else experiences this. I became aware of this through the CPU fan noise.

Bug 1406414 by the looks of what you are describing but not certain :-k
tauio111
Posts: 18
Joined: August 12th, 2017, 8:52 am

Re: The first official 20171108 builds are not yet out

Post by tauio111 »

gorhill wrote:I am having bout of relatively high CPU usage (~20% on quad-core) with 58.0a1 (2017-11-08) (64-bit) (linux version). This is intermittent, and unpredictable (might be happening when I open a new web page, need more data to make the case), and last for a few seconds (10s-20s very roughly estimated).

I looked into this using Gecko profiler, but it shows nothing -- all processes just show they are waiting at poll(), which is normal when the browser is idle (my understanding). The OS System Monitor shows that this is happening in the main process.

This observation is new to me -- not something I observed in previous builds. So anyway, just throwing this out there in case someone else experiences this. I became aware of this through the CPU fan noise.
I see Nigtly using 10% on my Ryzen 5 1600 while doing absolutely nothing. Something is broken
shellye5
Posts: 47
Joined: May 23rd, 2017, 9:57 am

Re: The first official 20171108 builds are not yet out

Post by shellye5 »

tauio111 wrote:
gorhill wrote:I am having bout of relatively high CPU usage (~20% on quad-core) with 58.0a1 (2017-11-08) (64-bit) (linux version). This is intermittent, and unpredictable (might be happening when I open a new web page, need more data to make the case), and last for a few seconds (10s-20s very roughly estimated).

I looked into this using Gecko profiler, but it shows nothing -- all processes just show they are waiting at poll(), which is normal when the browser is idle (my understanding). The OS System Monitor shows that this is happening in the main process.

This observation is new to me -- not something I observed in previous builds. So anyway, just throwing this out there in case someone else experiences this. I became aware of this through the CPU fan noise.
I see Nigtly using 10% on my Ryzen 5 1600 while doing absolutely nothing. Something is broken


Same but 25% ](*,)
User avatar
Mark12547
Posts: 327
Joined: May 13th, 2017, 11:36 am
Location: Oregon, United States, Earth

Re: The first official 20171108 builds are not yet out

Post by Mark12547 »

Yes, starting last night, after last night's upgrade, I noticed about 3 to 5 seconds of 25% CPU usage while Firefox is idle. (25% corresponds to a core of this 4-core CPU. I can notice this because 25% heats the CPU enough to kick on the CPU fan.)

It has nothing to do with a throbber because no throbber is running, so it isn't Bug 1406414.

The problem persists after this morning's update. This time, though, it hit about 40% while watching a YouTube video, whereas just watching the video is about 15% CPU.

Last night these were also installed through Windows Update:
  • KB2952664 Update for Windows 7 for x64-based Systems
  • KB4041686 2017-10 Preview Monthly Quality Rollup for Windows 7 for x64-based Systems
  • KB4042076 2017-10 Preview of Quality Rollup for .NET Framework 3.5.1, 4.5.2, 4.6, 4.6.1, 4.6.2, 4.7 on Windows 7 and Server 2008 R2 for x64
  • KB4011188 Update for Microsoft Office 2010 32-Bit Edition
Extensions installed and enabled:
  • uBlock Origin
  • Country Flag +
  • Hostname in title
  • IPvFoo
  • Simple Translate
  • small-clock
  • Tab Center Redux
  • Violentmonkey -> AAK-Cont Userscript For uBlock Origin
My system is:
  • Windows 7 Home Premium SP1 (64-bit)
  • 4 core CPU: AMD A6-3620 APU with Radeon(tm) HD Graphics 2.20 GHz
  • gpu: AMD Radeon HD 6530D (as reported by CPU-Z)
  • 8 GB RAM (7.49 GB usable)
  • Antivirus: Norton Security Suite (but with "community watch" turned off; it interferes with Nightly updates). I started using it 8 months ago when Avast Internet Security kept on dropping too many packets for distant websites. (Yes, I know antivirus provided by one's ISP generally has poor reputation, but it proved less troublesome than Avast Internet Security.)
Do we have any commonalities for these CPU spikes?

Do we have things we can rule out?
  • Operating system - this thread already lists Windows 10, Windows 7 and linux.
  • Processors - i7 and AMD are both mentioned
shellye5
Posts: 47
Joined: May 23rd, 2017, 9:57 am

Re: The first official 20171108 builds are not yet out

Post by shellye5 »

Mark12547 wrote:Yes, starting last night, after last night's upgrade, I noticed about 3 to 5 seconds of 25% CPU usage while Firefox is idle. (25% corresponds to a core of this 4-core CPU. I can notice this because 25% heats the CPU enough to kick on the CPU fan.)

It has nothing to do with a throbber because no throbber is running, so it isn't Bug 1406414.

The problem persists after this morning's update. This time, though, it hit about 40% while watching a YouTube video, whereas just watching the video is about 15% CPU.

Last night these were also installed through Windows Update:
  • KB2952664 Update for Windows 7 for x64-based Systems
  • KB4041686 2017-10 Preview Monthly Quality Rollup for Windows 7 for x64-based Systems
  • KB4042076 2017-10 Preview of Quality Rollup for .NET Framework 3.5.1, 4.5.2, 4.6, 4.6.1, 4.6.2, 4.7 on Windows 7 and Server 2008 R2 for x64
  • KB4011188 Update for Microsoft Office 2010 32-Bit Edition
Extensions installed and enabled:
  • uBlock Origin
  • Country Flag +
  • Hostname in title
  • IPvFoo
  • Simple Translate
  • small-clock
  • Tab Center Redux
  • Violentmonkey -> AAK-Cont Userscript For uBlock Origin
My system is:
  • Windows 7 Home Premium SP1 (64-bit)
  • 4 core CPU: AMD A6-3620 APU with Radeon(tm) HD Graphics 2.20 GHz
  • gpu: AMD Radeon HD 6530D (as reported by CPU-Z)
  • 8 GB RAM (7.49 GB usable)
  • Antivirus: Norton Security Suite (but with "community watch" turned off; it interferes with Nightly updates). I started using it 8 months ago when Avast Internet Security kept on dropping too many packets for distant websites. (Yes, I know antivirus provided by one's ISP generally has poor reputation, but it proved less troublesome than Avast Internet Security.)
Do we have any commonalities for these CPU spikes?

Do we have things we can rule out?
  • Operating system - this thread already lists Windows 10, Windows 7 and linux.
  • Processors - i7 and AMD are both mentioned
The throbber one kicks in while page loading like in the first report, #-o
not sure what causes the idle CPU spike, :evil:
User avatar
Mark12547
Posts: 327
Joined: May 13th, 2017, 11:36 am
Location: Oregon, United States, Earth

Re: The first official 20171108 builds are not yet out

Post by Mark12547 »

In my case, no throbbers were involved. First two times I was reading pages that had already loaded and no throbber. This morning I was in the middle of reading a page while another tab in another window was playing a YouTube video and no throbbers were active.

Also, the throbber consuming too much CPU has been reported on really slow machines, with a lot more discussion on the original throbber Bug, where one of the developers had to go out and purchase a laptop as similar as possible as the Reporter of the original throbber Bug so the problem could be reproduced on a slow machine. Discussions on the original throbber Bug got so long that they decided to spawn a new throbber Bug without the hundreds of comments.

I'm going to ask on the Firefox subreddit to see if others have observed the CPU spikes since last night's Nightly update and if we can discover any commonalities.
Lurtz
Posts: 359
Joined: June 12th, 2016, 12:25 pm

Re: The first official 20171108 builds are not yet out

Post by Lurtz »

WebRender MVP was just delayed to Firefox 60. Not surprising, still a bit sad.
User avatar
streetwolf
Posts: 2700
Joined: August 21st, 2011, 8:07 am
Location: NJ (USA)

Re: The first official 20171108 builds are not yet out

Post by streetwolf »

Lurtz wrote:WebRender MVP was just delayed to Firefox 60. Not surprising, still a bit sad.
Please provide a link to the source of this information.
Intel i9-13900K | ASUS ROG MAXIMUS Z790 HERO DDR5 | 64GB CORSAIR VENGEANCE DDR5 @ 6400 Mhz.
H100i ELITE CAPELLIX XT Liquid CPU Cooler | PNY 12GB GeForce RTX 3080 Ti | 2 CORSAIR 2TB MP600 PRO XT GEN 4
HX1200 PLATINUM PSU | XENEON 32" IPS UHD 144Hz | BenQ 32" UHD | MS Windows 11 Pro
bjherbison
Posts: 1039
Joined: October 6th, 2003, 5:40 am
Location: Bolton, MA, US
Contact:

Re: The first official 20171108 builds are not yet out

Post by bjherbison »

I see a mention of a delay beyond 59, but the mention says till TBD.

https://groups.google.com/forum/#!msg/m ... YqDMSxAwAJ
Josa
Posts: 7360
Joined: July 28th, 2009, 4:52 pm

Re: The first official 20171108 builds are out

Post by Josa »

Locked