1. 20 Feb, 2019 1 commit
    • Jaroslav Sevcik's avatar
      Reland "Fix accessor update of non-extensible maps." · 0a069d94
      Jaroslav Sevcik authored
      This is a reland of 1a3a2bc3,
      fixed an infinite loop in Map::TryUpdateSlow and added
      a relevant test.
      
      Original change's description:
      > Fix accessor update of non-extensible maps.
      >
      > When installing getter/setter of non-extensible map with existing
      > setter/getter of the same name, we introduce a new transition
      > (so we have two transitions with the same name!). This triggers
      > an assertion in map updater.
      >
      > This fix carefully checks that on the back-pointer path from
      > non-extensible map to the extensible map there are only
      > integrity level transitions. Otherwise, we just bail out.
      >
      > Bug: chromium:932953
      > Change-Id: I02e91c3b652428a84a9f5c58b6691ea9b1fc44d6
      > Reviewed-on: https://chromium-review.googlesource.com/c/1477067
      > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > Commit-Queue: Jaroslav Sevcik <jarin@chromium.org>
      > Cr-Commit-Position: refs/heads/master@{#59667}
      
      Bug: chromium:932953
      Change-Id: I015ee3795f816c8eabb5b5c5cb0ee30f365cc972
      Reviewed-on: https://chromium-review.googlesource.com/c/1477675Reviewed-by: 's avatarIgor Sheludko <ishell@chromium.org>
      Commit-Queue: Jaroslav Sevcik <jarin@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#59715}
      0a069d94
  2. 19 Feb, 2019 2 commits
    • Leszek Swirski's avatar
      Revert "Fix accessor update of non-extensible maps." · 4cbdf97d
      Leszek Swirski authored
      This reverts commit 1a3a2bc3.
      
      Reason for revert: Only CL in a reverted roll (https://chromium-review.googlesource.com/c/chromium/src/+/1477147)
      
      Original change's description:
      > Fix accessor update of non-extensible maps.
      > 
      > When installing getter/setter of non-extensible map with existing
      > setter/getter of the same name, we introduce a new transition
      > (so we have two transitions with the same name!). This triggers
      > an assertion in map updater.
      > 
      > This fix carefully checks that on the back-pointer path from
      > non-extensible map to the extensible map there are only
      > integrity level transitions. Otherwise, we just bail out.
      > 
      > Bug: chromium:932953
      > Change-Id: I02e91c3b652428a84a9f5c58b6691ea9b1fc44d6
      > Reviewed-on: https://chromium-review.googlesource.com/c/1477067
      > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > Commit-Queue: Jaroslav Sevcik <jarin@chromium.org>
      > Cr-Commit-Position: refs/heads/master@{#59667}
      
      TBR=jarin@chromium.org,ishell@chromium.org
      
      Change-Id: I9264423c605cebef87beb6c0f066e90b59faae48
      No-Presubmit: true
      No-Tree-Checks: true
      No-Try: true
      Bug: chromium:932953
      Reviewed-on: https://chromium-review.googlesource.com/c/1477219Reviewed-by: 's avatarLeszek Swirski <leszeks@chromium.org>
      Commit-Queue: Leszek Swirski <leszeks@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#59680}
      4cbdf97d
    • Jaroslav Sevcik's avatar
      Fix accessor update of non-extensible maps. · 1a3a2bc3
      Jaroslav Sevcik authored
      When installing getter/setter of non-extensible map with existing
      setter/getter of the same name, we introduce a new transition
      (so we have two transitions with the same name!). This triggers
      an assertion in map updater.
      
      This fix carefully checks that on the back-pointer path from
      non-extensible map to the extensible map there are only
      integrity level transitions. Otherwise, we just bail out.
      
      Bug: chromium:932953
      Change-Id: I02e91c3b652428a84a9f5c58b6691ea9b1fc44d6
      Reviewed-on: https://chromium-review.googlesource.com/c/1477067Reviewed-by: 's avatarIgor Sheludko <ishell@chromium.org>
      Commit-Queue: Jaroslav Sevcik <jarin@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#59667}
      1a3a2bc3
  3. 26 Dec, 2018 1 commit
  4. 20 Dec, 2018 1 commit
  5. 07 Dec, 2018 1 commit
  6. 05 Dec, 2018 2 commits
  7. 29 Nov, 2018 2 commits
  8. 27 Nov, 2018 1 commit
  9. 22 Nov, 2018 1 commit
    • Shiyu Zhang's avatar
      Reland "Create a fast path to get migration target when updating map" · 69232ba6
      Shiyu Zhang authored
      This is a reland of 6ec90ece
      
      Original change's description:
      > Reland "Create a fast path to get migration target when updating map"
      >
      > This is a reland of c285380c
      >
      > Original change's description:
      > > Create a fast path to get migration target when updating map
      > >
      > > During map updating, store the pointer to new map in the
      > > raw_transitions slot of the old map that is deprecated from map
      > > transition tree. Thus, we can get the migration target directly
      > > instead of TryReplayPropertyTransitions when updating map.
      > >
      > > This can improve Speedometer2.0 Elm-TodoMVC case by ~5% on ATOM
      > > Chromebook and ~9% on big-core Ubuntu.
      > >
      > > Change-Id: I56f9ce5183bbdd567b964890f623ef0ceed9b7db
      > > Reviewed-on: https://chromium-review.googlesource.com/1233433
      > > Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      > > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > > Cr-Commit-Position: refs/heads/master@{#56303}
      >
      > Change-Id: Idf0b7716b92a6a15bfe58721c2c34dbd02b31137
      > Reviewed-on: https://chromium-review.googlesource.com/c/1270261
      > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      > Cr-Commit-Position: refs/heads/master@{#56588}
      
      Change-Id: Iad23455f66c83340430370be6670bac493c638a1
      Reviewed-on: https://chromium-review.googlesource.com/c/1309305
      Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      Reviewed-by: 's avatarIgor Sheludko <ishell@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#57705}
      69232ba6
  10. 13 Nov, 2018 1 commit
  11. 26 Oct, 2018 1 commit
  12. 24 Oct, 2018 1 commit
  13. 15 Oct, 2018 1 commit
    • Igor Sheludko's avatar
      Revert "Reland "Create a fast path to get migration target when updating map"" · 1898944b
      Igor Sheludko authored
      This reverts commit 6ec90ece.
      
      Reason for revert: causes a lot of Canary crashes (chromium:895208).
      GC relies on an the fact that the transition array stays alive while it's owner map
      is alive (this is needed in order to properly transfer descriptor array ownership
      to the parent map when the map owning a shared descriptor array dies). We need to
      rethink a way of caching the migration target shortcut.
      
      Original change's description:
      > Reland "Create a fast path to get migration target when updating map"
      >
      > This is a reland of c285380c
      >
      > Original change's description:
      > > Create a fast path to get migration target when updating map
      > >
      > > During map updating, store the pointer to new map in the
      > > raw_transitions slot of the old map that is deprecated from map
      > > transition tree. Thus, we can get the migration target directly
      > > instead of TryReplayPropertyTransitions when updating map.
      > >
      > > This can improve Speedometer2.0 Elm-TodoMVC case by ~5% on ATOM
      > > Chromebook and ~9% on big-core Ubuntu.
      > >
      > > Change-Id: I56f9ce5183bbdd567b964890f623ef0ceed9b7db
      > > Reviewed-on: https://chromium-review.googlesource.com/1233433
      > > Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      > > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > > Cr-Commit-Position: refs/heads/master@{#56303}
      >
      > Change-Id: Idf0b7716b92a6a15bfe58721c2c34dbd02b31137
      > Reviewed-on: https://chromium-review.googlesource.com/c/1270261
      > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      > Cr-Commit-Position: refs/heads/master@{#56588}
      
      TBR=ishell@chromium.org,shiyu.zhang@intel.com
      
      # Not skipping CQ checks because original CL landed > 1 day ago.
      
      Change-Id: Ie7e9b98395b041a1095da549d1cd71d7180a4888
      Bug: chromium:895208
      Reviewed-on: https://chromium-review.googlesource.com/c/1280223
      Commit-Queue: Igor Sheludko <ishell@chromium.org>
      Reviewed-by: 's avatarIgor Sheludko <ishell@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#56641}
      1898944b
  14. 12 Oct, 2018 1 commit
    • Shiyu Zhang's avatar
      Reland "Create a fast path to get migration target when updating map" · 6ec90ece
      Shiyu Zhang authored
      This is a reland of c285380c
      
      Original change's description:
      > Create a fast path to get migration target when updating map
      >
      > During map updating, store the pointer to new map in the
      > raw_transitions slot of the old map that is deprecated from map
      > transition tree. Thus, we can get the migration target directly
      > instead of TryReplayPropertyTransitions when updating map.
      >
      > This can improve Speedometer2.0 Elm-TodoMVC case by ~5% on ATOM
      > Chromebook and ~9% on big-core Ubuntu.
      >
      > Change-Id: I56f9ce5183bbdd567b964890f623ef0ceed9b7db
      > Reviewed-on: https://chromium-review.googlesource.com/1233433
      > Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > Cr-Commit-Position: refs/heads/master@{#56303}
      
      Change-Id: Idf0b7716b92a6a15bfe58721c2c34dbd02b31137
      Reviewed-on: https://chromium-review.googlesource.com/c/1270261Reviewed-by: 's avatarIgor Sheludko <ishell@chromium.org>
      Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
      Cr-Commit-Position: refs/heads/master@{#56588}
      6ec90ece
  15. 30 Sep, 2018 2 commits
  16. 17 Sep, 2018 1 commit
  17. 14 Sep, 2018 2 commits
    • Marja Hölttä's avatar
      Revert "[in-place weak refs] Fix MaybeObject function names" · 3a79fe23
      Marja Hölttä authored
      This reverts commit ad72d195.
      
      Reason for revert: Build failures on *san
      
      Original change's description:
      > [in-place weak refs] Fix MaybeObject function names
      > 
      > E.g., "ToWeakHeapObject" was misleading, since it didn't convert to a weak heap
      > object, instead returned a weakly pointed heap object. Change the function names
      > (in this case, to "GetHeapObjectIfWeak") to reflect this.
      > 
      > Also make casts explicit, if a MaybeObject is an Object, we can call cast<Object>().
      > 
      > BUG=v8:7308
      > 
      > Change-Id: I4ef078572b4f4415afe7e2e706d3bd684e16e47d
      > Reviewed-on: https://chromium-review.googlesource.com/1219025
      > Reviewed-by: Igor Sheludko <ishell@chromium.org>
      > Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
      > Reviewed-by: Yang Guo <yangguo@chromium.org>
      > Reviewed-by: Andreas Haas <ahaas@chromium.org>
      > Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
      > Commit-Queue: Marja Hölttä <marja@chromium.org>
      > Cr-Commit-Position: refs/heads/master@{#55906}
      
      TBR=ulan@chromium.org,marja@chromium.org,yangguo@chromium.org,ahaas@chromium.org,tebbi@chromium.org,ishell@chromium.org
      
      Change-Id: I054b578518e3f6fd7dbcddf0b56cc018726c1e7a
      No-Presubmit: true
      No-Tree-Checks: true
      No-Try: true
      Bug: v8:7308
      Reviewed-on: https://chromium-review.googlesource.com/1226874Reviewed-by: 's avatarMarja Hölttä <marja@chromium.org>
      Commit-Queue: Marja Hölttä <marja@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#55918}
      3a79fe23
    • Marja Hölttä's avatar
      [in-place weak refs] Fix MaybeObject function names · ad72d195
      Marja Hölttä authored
      E.g., "ToWeakHeapObject" was misleading, since it didn't convert to a weak heap
      object, instead returned a weakly pointed heap object. Change the function names
      (in this case, to "GetHeapObjectIfWeak") to reflect this.
      
      Also make casts explicit, if a MaybeObject is an Object, we can call cast<Object>().
      
      BUG=v8:7308
      
      Change-Id: I4ef078572b4f4415afe7e2e706d3bd684e16e47d
      Reviewed-on: https://chromium-review.googlesource.com/1219025Reviewed-by: 's avatarIgor Sheludko <ishell@chromium.org>
      Reviewed-by: 's avatarTobias Tebbi <tebbi@chromium.org>
      Reviewed-by: 's avatarYang Guo <yangguo@chromium.org>
      Reviewed-by: 's avatarAndreas Haas <ahaas@chromium.org>
      Reviewed-by: 's avatarUlan Degenbaev <ulan@chromium.org>
      Commit-Queue: Marja Hölttä <marja@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#55906}
      ad72d195
  18. 18 Jul, 2018 1 commit
  19. 13 Jul, 2018 1 commit
  20. 12 Jul, 2018 1 commit
    • Leszek Swirski's avatar
      [cleanup] Remove Isolate parameter from object print · 13b899a5
      Leszek Swirski authored
      With ReadOnlyRoots and GetIsolate on JSReceiver, we can remove almost
      every isolate parameter from <Object>::Print. The remaining ones, like
      Map, are special-caseable for read-only maps, and as a result we can
      remove isolate parameters from <Object>::Print entirely.
      
      This patch also opportunistically cleans up a few places where isolates
      were only needed for Object::Print, such as TransitionAccessors and
      DescriptorArrays.
      
      TBR=yangguo@chromium.org,mstarzinger@chromium.org
      
      Bug: v8:7786
      Change-Id: Id44bd53b9893e679eea5f37b9548257595a1bfd9
      Reviewed-on: https://chromium-review.googlesource.com/1133385Reviewed-by: 's avatarLeszek Swirski <leszeks@chromium.org>
      Reviewed-by: 's avatarDan Elphick <delphick@chromium.org>
      Commit-Queue: Leszek Swirski <leszeks@chromium.org>
      Cr-Commit-Position: refs/heads/master@{#54401}
      13b899a5
  21. 04 Jul, 2018 1 commit
  22. 21 Jun, 2018 1 commit
  23. 05 Jun, 2018 1 commit
  24. 29 May, 2018 1 commit
  25. 19 Apr, 2018 1 commit
  26. 18 Apr, 2018 2 commits
  27. 05 Apr, 2018 1 commit
  28. 28 Mar, 2018 2 commits
  29. 23 Mar, 2018 1 commit
  30. 19 Mar, 2018 1 commit
  31. 15 Jan, 2018 1 commit
  32. 14 Dec, 2017 1 commit
  33. 18 Oct, 2017 1 commit