3b3febea820cbf6360061d8bcd9c709f43dcf415: Bug 1364364 - Part 4: Add test case. r?smaug draft
Samael Wang <freesamael@gmail.com> - Wed, 30 Aug 2017 17:00:12 +0800 - rev 664559
Push 79723 by bmo:sawang@mozilla.com at Thu, 14 Sep 2017 03:01:27 +0000
Bug 1364364 - Part 4: Add test case. r?smaug MozReview-Commit-ID: BozQDPdx0MI
d51c63c7974c22b7a70cbad98bf8e1cd22314a99: Bug 1364364 - Part 3: Clear dynamic subframe entries in RemoveFromBFCacheSync/Async. r?smaug draft
Samael Wang <freesamael@gmail.com> - Thu, 24 Aug 2017 15:17:39 +0800 - rev 664558
Push 79723 by bmo:sawang@mozilla.com at Thu, 14 Sep 2017 03:01:27 +0000
Bug 1364364 - Part 3: Clear dynamic subframe entries in RemoveFromBFCacheSync/Async. r?smaug MozReview-Commit-ID: F8OTxbWIp5O
9148ef1f8195f23bbcdbd4ce4a62a85a5e1c00e8: Bug 1389470 - fix for zero size memory leak. r?kinetik draft
Alfredo.Yang <ayang@mozilla.com> - Thu, 14 Sep 2017 10:29:43 +0800 - rev 664557
Push 79722 by bmo:ayang@mozilla.com at Thu, 14 Sep 2017 02:32:13 +0000
Bug 1389470 - fix for zero size memory leak. r?kinetik MozReview-Commit-ID: F1hBnukiwKN
e1986cac32c7c169e588d6e8135308a104252a7d: Bug 1399372 - update the principal only once per OnDataAvailable() call. draft
JW Wang <jwwang@mozilla.com> - Tue, 29 Aug 2017 17:57:00 +0800 - rev 664556
Push 79721 by jwwang@mozilla.com at Thu, 14 Sep 2017 02:15:48 +0000
Bug 1399372 - update the principal only once per OnDataAvailable() call. MozReview-Commit-ID: BWyDmKglWsl
7fbd7a84416c8e129665ecce4f111c0346d33f10: Bug 1397101 - Only use a thread local arena for small sizes. r?njn draft
Mike Hommey <mh+mozilla@glandium.org> - Thu, 14 Sep 2017 07:36:39 +0900 - rev 664555
Push 79720 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 02:03:42 +0000
Bug 1397101 - Only use a thread local arena for small sizes. r?njn The bin-unused count in memory reports indicates how much memory is used by runs of small and sub-page allocations that is not actually allocated. This is generally thought as an indicator of fragmentation. While this is generally true, with the use of thread local arenas by stylo, combined with how stylo allocates memory, it ends up also being an indicator of wasted memory. For instance, over the lifetime of an AWSY iteration, there are only a few allocations that ends up in the bucket for 2048 allocated bytes. In the "worst" case, there's only one. But the run size for such allocations is 132KiB. Which means just because we're allocating one buffer of size between 1024 and 2048 bytes, we end up wasting 130+KiB. Per thread. Something similar happens with size classes of 512 and 1024, where the run size is respectively 32KiB and 64KiB, and where there's at most a handful of allocations of each class ever happening per thread. Overall, an allocation log from a full AWSY iteration reveals that there are only 448 of 860700 allocations happening on the stylo arenas that involve sizes above (and excluding) 512 bytes, so 0.05%. While there are improvements that can be done to mozjemalloc so that it doesn't waste more than one page per sub-page size class, they are changes that are too low-level to land at this time of the release cycle. However, considering the numbers above and the fact that the stylo arenas are only really meant to avoid lock contention during the heavy parallel work involved, a short term, low risk, strategy is to just delegate all sub-page (> 512, < 4096) and large (>= 4096) to the main arena. Technically speaking, only sub-page allocations are causing this waste, but it's more consistent to just delegate everything above 512 bytes. This should save 132KiB + 64KiB = 196KiB per stylo thread.
27313c255db72920dcf0cd87f32648c09e9a6604: Bug 1397101 - Reduce the number of dirty pages we allow to be kept in thread local arenas. r?njn draft
Mike Hommey <mh+mozilla@glandium.org> - Thu, 14 Sep 2017 07:26:30 +0900 - rev 664554
Push 79720 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 02:03:42 +0000
Bug 1397101 - Reduce the number of dirty pages we allow to be kept in thread local arenas. r?njn Until bug 1361258, there was only ever one mozjemalloc arena, and the number of dirty pages we allow to be kept dirty, fixed to 1MB per arena, was, in fact, 1MB for an entire process. With stylo using thread local arenas, we now can have multiple arenas per process, multiplying that number of dirty pages. While those dirty pages may be reused later on, when other allocations end up filling them later on, the fact that a relatively large number of them is kept around for each stylo thread (in proportion to the amount of memory ever allocated by stylo), combined with the fact that the memory use from stylo depends on the workload generated by the pages being visited, those dirty pages may very well not be used for a rather long time. This is less of a problem with the main arena, used for most everything else. So, for each arena except the main one, we decrease the number of dirty pages we allow to be kept around to 1/8 of the current value. We do this by introducing a per-arena configuration of that maximum number.
c7b1cd3ec017a351831bb0abf738d70656d17601: Bug 1383907 - Enable WebVR reftests on macOS; r?kip, jgilbert draft
Daosheng Mu <daoshengmu@gmail.com> - Fri, 25 Aug 2017 00:30:31 +0800 - rev 664553
Push 79719 by bmo:dmu@mozilla.com at Thu, 14 Sep 2017 01:44:35 +0000
Bug 1383907 - Enable WebVR reftests on macOS; r?kip, jgilbert MozReview-Commit-ID: 2OpA95QWI4d
7a1a6934ac4c137f58c79d98484fffb20c571389: Bug 1393659 - Fix inconsistent handling of max_post_bytes and max_request_bytes r?markh draft
Thom Chiovoloni <tchiovoloni@mozilla.com> - Wed, 13 Sep 2017 18:02:41 -0700 - rev 664552
Push 79718 by bmo:tchiovoloni@mozilla.com at Thu, 14 Sep 2017 01:31:44 +0000
Bug 1393659 - Fix inconsistent handling of max_post_bytes and max_request_bytes r?markh MozReview-Commit-ID: 4jwpAYNuoQj
07092bab376310867fa1a87210ba6c3eddb9cc8e: Bug 1398993 - Disable stylo in e10s parent process. r=xidorn draft
Cameron McCormack <cam@mcc.id.au> - Wed, 13 Sep 2017 17:13:11 +0800 - rev 664551
Push 79717 by bmo:cam@mcc.id.au at Thu, 14 Sep 2017 01:21:46 +0000
Bug 1398993 - Disable stylo in e10s parent process. r=xidorn nsLayoutStatics::Initialize is sometimes too early to know whether a process is an e10s parent process, or a non-e10s main process, because some prefs get loaded later on. So we unconditionally initialize some Servo data in nsLayoutStatics::Initialize, but we still check later on whether we are really a non-e10s main process or e10s content process when deciding whether to preload Servo style sheets, choose our document backend type, etc. MozReview-Commit-ID: 93tPCvuTdzl
161c06ef2c6d36d494a76435f67af53d33869c9a: Bug 1395387 - Reconcile WebExtension page actions and Photon page actions. draft
Drew Willcoxon <adw@mozilla.com> - Wed, 13 Sep 2017 18:12:04 -0700 - rev 664550
Push 79716 by dwillcoxon@mozilla.com at Thu, 14 Sep 2017 01:12:44 +0000
Bug 1395387 - Reconcile WebExtension page actions and Photon page actions. MozReview-Commit-ID: GLWqHebJNBY
e47a6715f29ef6847f76d2c4195cd0a1bc4480ac: Bug 1393659 - Fix inconsistent handling of max_post_bytes and max_request_bytes r?markh draft
Thom Chiovoloni <tchiovoloni@mozilla.com> - Wed, 13 Sep 2017 18:02:41 -0700 - rev 664549
Push 79715 by bmo:tchiovoloni@mozilla.com at Thu, 14 Sep 2017 01:10:10 +0000
Bug 1393659 - Fix inconsistent handling of max_post_bytes and max_request_bytes r?markh MozReview-Commit-ID: 4jwpAYNuoQj
91eb4bd3032adb0392a7d13f08a14651c0c3a162: Bug 1393659 - Fix inconsistent handling of max_post_bytes and max_request_bytes r?markh draft
Thom Chiovoloni <tchiovoloni@mozilla.com> - Wed, 13 Sep 2017 18:02:41 -0700 - rev 664548
Push 79714 by bmo:tchiovoloni@mozilla.com at Thu, 14 Sep 2017 01:02:53 +0000
Bug 1393659 - Fix inconsistent handling of max_post_bytes and max_request_bytes r?markh MozReview-Commit-ID: 4jwpAYNuoQj
3d20b9c62540ad9e873a175f9ddbfd1c00f390cf: Bug 1397101 - Only use a thread local arena for small sizes. r?njn draft
Mike Hommey <mh+mozilla@glandium.org> - Thu, 14 Sep 2017 07:36:39 +0900 - rev 664547
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Bug 1397101 - Only use a thread local arena for small sizes. r?njn The bin-unused count in memory reports indicates how much memory is used by runs of small and sub-page allocations that is not actually allocated. This is generally thought as an indicator of fragmentation. While this is generally true, with the use of thread local arenas by stylo, combined with how stylo allocates memory, it ends up also being an indicator of wasted memory. For instance, over the lifetime of an AWSY iteration, there are only a few allocations that ends up in the bucket for 2048 allocated bytes. In the "worst" case, there's only one. But the run size for such allocations is 132KiB. Which means just because we're allocating one buffer of size between 1024 and 2048 bytes, we end up wasting 130+KiB. Per thread. Something similar happens with size classes of 512 and 1024, where the run size is respectively 32KiB and 64KiB, and where there's at most a handful of allocations of each class ever happening per thread. Overall, an allocation log from a full AWSY iteration reveals that there are only 448 of 860700 allocations happening on the stylo arenas that involve sizes above (and excluding) 512 bytes, so 0.05%. While there are improvements that can be done to mozjemalloc so that it doesn't waste more than one page per sub-page size class, they are changes that are too low-level to land at this time of the release cycle. However, considering the numbers above and the fact that the stylo arenas are only really meant to avoid lock contention during the heavy parallel work involved, a short term, low risk, strategy is to just delegate all sub-page (> 512, < 4096) and large (>= 4096) to the main arena. Technically speaking, only sub-page allocations are causing this waste, but it's more consistent to just delegate everything above 512 bytes. This should save 132KiB + 64KiB = 196KiB per stylo thread.
44ae0224d5c0456d08bc550e7341718d033715dc: Bug 1397101 - Reduce the number of dirty pages we allow to be kept in thread local arenas. r?njn draft
Mike Hommey <mh+mozilla@glandium.org> - Thu, 14 Sep 2017 07:26:30 +0900 - rev 664546
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Bug 1397101 - Reduce the number of dirty pages we allow to be kept in thread local arenas. r?njn Until bug 1361258, there was only ever one mozjemalloc arena, and the number of dirty pages we allow to be kept dirty, fixed to 1MB per arena, was, in fact, 1MB for an entire process. With stylo using thread local arenas, we now can have multiple arenas per process, multiplying that number of dirty pages. While those dirty pages may be reused later on, when other allocations end up filling them later on, the fact that a relatively large number of them is kept around for each stylo thread (in proportion to the amount of memory ever allocated by stylo), combined with the fact that the memory use from stylo depends on the workload generated by the pages being visited, those dirty pages may very well not be used for a rather long time. This is less of a problem with the main arena, used for most everything else. So, for each arena except the main one, we decrease the number of dirty pages we allow to be kept around to 1/8 of the current value. We do this by introducing a per-arena configuration of that maximum number.
3b18e792cc3eb0bf381a422511c5bdde66f10af8: Backed out changeset 1061923744e8 (bug 1396784) for leaks in devtools' browser_boxmodel_computed-accordion-state.js. r=backout
Sebastian Hengst <archaeopteryx@coole-files.de> - Thu, 14 Sep 2017 00:27:19 +0200 - rev 664545
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Backed out changeset 1061923744e8 (bug 1396784) for leaks in devtools' browser_boxmodel_computed-accordion-state.js. r=backout
7d4a2f21a7b3b04981f8b4c6cb8ddd36cb6deaa8: Backed out changeset 99f6b31d5f44 (bug 1396784)
Sebastian Hengst <archaeopteryx@coole-files.de> - Thu, 14 Sep 2017 00:26:31 +0200 - rev 664544
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Backed out changeset 99f6b31d5f44 (bug 1396784)
9d99ac9ae088bcee2b7f89f2aff10ff0bea73f56: merge mozilla-central to autoland. r=merge a=merge
Sebastian Hengst <archaeopteryx@coole-files.de> - Thu, 14 Sep 2017 00:11:28 +0200 - rev 664543
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
merge mozilla-central to autoland. r=merge a=merge
3ce5a9419d5b7a976d442901c04995c22e014bf1: Bug 1396274 - Disable Activity Stream about:home in browser_aboutHome.js r=Mardak
Ursula Sarracini - Wed, 13 Sep 2017 18:06:33 -0400 - rev 664542
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Bug 1396274 - Disable Activity Stream about:home in browser_aboutHome.js r=Mardak MozReview-Commit-ID: LXzgaAhNjnV
01b9bea7d2ba241fba343aedf6decdcf2fac851f: Bug 1397005 - Set size of panel text r=liuche
sajattack <sajattack@gmail.com> - Mon, 11 Sep 2017 17:33:53 -0700 - rev 664541
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Bug 1397005 - Set size of panel text r=liuche MozReview-Commit-ID: Ik6PFbF5Loc
cfd5924a10e353e60b2dc93837954b446e3d422e: Bug 1397930 - Use the same Safe Browsing client ID on all official builds. r=gcp
Francois Marier <francois@mozilla.com> - Wed, 13 Sep 2017 14:04:26 -0700 - rev 664540
Push 79713 by bmo:mh+mozilla@glandium.org at Thu, 14 Sep 2017 00:59:03 +0000
Bug 1397930 - Use the same Safe Browsing client ID on all official builds. r=gcp MozReview-Commit-ID: 5der4PiHWyv
(0) -300000 -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 tip