This is a partial reland of https://crrev.com/c/3597106
With this change, an existing Script from the compilation cache can be
reused after its top-level SharedFunctionInfo was discarded, but only if
the new script is parsed on the main thread (not deserialized from code
cache data, and not parsed on a background thread).
Bug: v8:12808
Change-Id: I1edaee2095306a89e2c3b91f2fd01ac053f3c770
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3689348
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#81472}
MinorMC maintained a separate marking state to support interleaved GCs.
Since MinorMC now assumes that interleaving is not possible, MinorMC can
use the same marking state as the full GC.
Bug: v8:12612
Change-Id: Ibeb7df2eb24e448f811b497c9d16b3b132f87ec2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3735163
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Omer Katz <omerkatz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81468}
Unaglined allocations are not fully supported in V8.
- Set USE_ALLOCATION_ALIGNMENT_BOOL to false for documentation
- Verify HeapObject address alignment requirements with --verify-heap
- Move address alignment to right after allocation in the deserializer
- Use object_size in the CheckAlignment helper to get a chance to
figure out which allocation path we took
Bug: chromium:1330861, v8:8875
Change-Id: Iffd02d869923ccec133618250dfefb0480b02741
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3717995
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81358}
This CL introduces a compile flag v8_enable_inner_pointer_resolution_osb
behind which lies the experimental implementation of the object start
bitmap. It disassociates the object start bitmap from the compile flag
v8_enable_conservative_stack_scanning. At the moment the former flag is
a prerequisite for the latter, as conservative stack scanning requires
some mechanism for inner pointer resolution and the object start bitmap
provides one such mechanism.
Bug: v8:12851
Change-Id: I24c6b389453fbaefc79ae50c34c5ec7a1bf23347
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3717322
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Nikolaos Papaspyrou <nikolaos@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81295}
Remove finalization step of incremental marking. The step was
historically used to process embedder/weak work on the main thread
before invoking the atomic pause. Remove the infrastructure as the
step is not needed anymore and actually required a safepoint.
Change-Id: I208767bbac3d9a06a0b3c67aa9779f8a5fa07328
Bug: v8:12775
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3702801
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81234}
This is a partial reland of https://crrev.com/c/3597106 , except for the
changes in compiler.cc, which are just the minimal possible changes to
make the code compile.
With this change, it is possible that a call to
CompilationCache::LookupScript returns any of:
1. A Script and a toplevel SharedFunctionInfo (cache hit)
2. A Script but no toplevel SharedFunctionInfo (partial cache hit)
3. Nothing (cache miss)
Bug: v8:12808
Change-Id: Id33a4cd0cb28562d6b862fbb113ea9d03f255b2b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3687425
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#81193}
... which might fail because usual operator== for tagged values compares
only lower 32 bits of the pointer.
Bug: v8:11880, v8:12958
Change-Id: I0978d6c510424aecfee2f044c40ea424b6cb3ab9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3695593
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Auto-Submit: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81096}
CodePageCollectionMemoryModificationScope now increases a per-thread
counter and inserts unprotected code chunks into a thread-local set
of chunks. This information is moved from Heap into LocalHeap.
We can't use kMaxWriteUnprotectCounter on the unprotect counter on the
MemoryChunk anymore, since e.g. for concurrent Sparkplug N threads might
now allocate a code object on the same page and since
CodePageCollectionMemoryModificationScope doesn't know about the
other threads anymore, each thread has to increase that counter by 1.
We DCHECK that nesting depth now in the scope's constructor instead.
We still need to remove chunks from `unprotected_memory_chunks_` when
freeing an executable MemoryChunk during GC. Fortunately we can still do
this, since all threads are in a safepoint during GC and we can remove
the chunk from each thread-local set without any synchronization.
Bug: chromium:1330887
Change-Id: Icefc61b8d8de113d8dcfb1cf64122d12dd9798c4
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3688516
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81047}
This is a reland of commit abcb6bb8b4.
The data race is fixed by using atomic operations.
Original change's description:
> [heap] Avoid dynamic updates of FLAG_gc_interval
>
> Flags will be protected from updates after V8 initialization (in the
> future). This CL avoids any updates of the --gc-interval flag during
> runtime, and instead updates a static field on the HeapAllocator
> directly.
>
> R=mlippautz@chromium.org
>
> Bug: v8:12887
> Change-Id: I17a495cae50a46d59a8159c6ece1558d4d61b949
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3687691
> Commit-Queue: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80998}
Bug: v8:12887
Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_rel_ng
Change-Id: Ib5b537500413a627d9b2509354d20906e0474d8e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3695380
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81008}
This reverts commit abcb6bb8b4.
Reason for revert: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux64%20TSAN%20-%20isolates/20029/overview
Original change's description:
> [heap] Avoid dynamic updates of FLAG_gc_interval
>
> Flags will be protected from updates after V8 initialization (in the
> future). This CL avoids any updates of the --gc-interval flag during
> runtime, and instead updates a static field on the HeapAllocator
> directly.
>
> R=mlippautz@chromium.org
>
> Bug: v8:12887
> Change-Id: I17a495cae50a46d59a8159c6ece1558d4d61b949
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3687691
> Commit-Queue: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80998}
Bug: v8:12887
Change-Id: I18310a3f515506d617f42be7a208013957625eaf
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3695559
Reviewed-by: Manos Koukoutos <manoskouk@chromium.org>
Owners-Override: Manos Koukoutos <manoskouk@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81002}
Flags will be protected from updates after V8 initialization (in the
future). This CL avoids any updates of the --gc-interval flag during
runtime, and instead updates a static field on the HeapAllocator
directly.
R=mlippautz@chromium.org
Bug: v8:12887
Change-Id: I17a495cae50a46d59a8159c6ece1558d4d61b949
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3687691
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80998}
... as a pair of Code and CodeDataContainer.
In order to stop creating and using trampoline Code objects for
builtins we need a different way to represent an "embedded builtin"
code lookup result of builtin trampoline Code objects.
We can't switch to CodeT for this purpose because GC still needs to
be able to locate not yet evacuated Code objects in order to update
old code pointers on the stack once Code objects are moved.
Bug: v8:11880
Change-Id: I296636a6728a11c8e3220b3fee43fd12ff633c1b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3684813
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80937}
This CL introduces the following changes to the experimental
implementation of the object start bitmap, that is evaluated as
a mechanism for resolving inner pointers (behind the flag
v8_enable_conservative_stack_scanning):
- Manually iterate through page objects, instead of using the
PagedSpaceObjectIterator, for performance (avoid calling
MakeHeapIterable all the time) and to simplify the handling
of filler objects.
- Clear bits when reusing evacuated pages of the new space.
- Use the cage base to iterate correctly through code objects.
- Introduce a method for verifying the validity of the object
start bitmap.
- Minor fixes, additional checks and cleanup.
Bug: v8:12851
Change-Id: I245937ffe6f4b53c4c2dcf5126e8836aec4dc79e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3675099
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Commit-Queue: Nikolaos Papaspyrou <nikolaos@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80869}
The abstractions in this CL include:
1) Using EvacuatePrologue to handle age mark updating in
SemiSpaceNewSpace.
2) Using IsPromotionCandidate to check if a page contains
the current age mark.
3) EnsureCurrentCapacity instead of Rebalance.
4) Delegate page promotions in mark-compact.cc to the
NewSpace implementation.
Bug: v8:12612
Change-Id: Ied83261d661a8e61a11bf33b1d7a2103ac99a853
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3644966
Commit-Queue: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80846}
This reverts commit 5f0e7dd44f.
Reason for revert: Turns out this isn't true: https://source.chromium.org/chromium/chromium/src/+/main:v8/src/ic/ic.cc;l=946
Original change's description:
> [ic] Shadow smi_handler on LoadHandler to always be Smi
>
> StoreHandlers can store either a Code object or a Smi, so the
> smi_handler field on DataHandler is a Code|Smi. But, LoadHandlers can
> only store Smis here so add a shadowing smi_handler definition which
> only allows / returns Smis.
>
> Change-Id: Icc1c508db268950990b9c84b9e682b9a9b1f43e5
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3663745
> Commit-Queue: Igor Sheludko <ishell@chromium.org>
> Auto-Submit: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> Reviewed-by: Igor Sheludko <ishell@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80775}
Change-Id: Ia2c14141ab5929be3ded883f0cecb40645a72dfd
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3672419
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Owners-Override: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#80776}
StoreHandlers can store either a Code object or a Smi, so the
smi_handler field on DataHandler is a Code|Smi. But, LoadHandlers can
only store Smis here so add a shadowing smi_handler definition which
only allows / returns Smis.
Change-Id: Icc1c508db268950990b9c84b9e682b9a9b1f43e5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3663745
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80775}
This is a reland of commit a76072217a
The bug exposed by landing this change the first time has been fixed
separately in https://crrev.com/c/3654413 .
Original change's description:
> Disable recompilation of existing Scripts from Isolate compilation cache
>
> My previous change https://crrev.com/c/3597106 led to some performance
> regressions in time spent on parsing and compilation. This change
> disables the ability to recompile an existing uncompiled Script, as an
> attempt to both fix the regressions and isolate which part of the
> previous change was the cause of those problems.
>
> Bug: v8:12808, chromium:1325566, chromium:1325567, chromium:1325601
> Change-Id: Ifa086bf27070da8f4b3c0e4415af5ca7b6706b0a
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3652252
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
> Cr-Commit-Position: refs/heads/main@{#80616}
Bug: v8:12808, chromium:1325566, chromium:1325567, chromium:1325601
Change-Id: Ib31864bef90ff3340d1dfd4e25e21bef121f2d49
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3655011
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#80645}
This reverts commit a76072217a.
Reason for revert: fails on GC Stress bot:
https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux%20-%20gc%20stress/38512/overview
Original change's description:
> Disable recompilation of existing Scripts from Isolate compilation cache
>
> My previous change https://crrev.com/c/3597106 led to some performance
> regressions in time spent on parsing and compilation. This change
> disables the ability to recompile an existing uncompiled Script, as an
> attempt to both fix the regressions and isolate which part of the
> previous change was the cause of those problems.
>
> Bug: v8:12808, chromium:1325566, chromium:1325567, chromium:1325601
> Change-Id: Ifa086bf27070da8f4b3c0e4415af5ca7b6706b0a
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3652252
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
> Cr-Commit-Position: refs/heads/main@{#80616}
Bug: v8:12808, chromium:1325566, chromium:1325567, chromium:1325601
Change-Id: I0f56163856d04ff49da96b0fb344fb59b5501a40
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3653273
Auto-Submit: Adam Klein <adamk@chromium.org>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#80617}
My previous change https://crrev.com/c/3597106 led to some performance
regressions in time spent on parsing and compilation. This change
disables the ability to recompile an existing uncompiled Script, as an
attempt to both fix the regressions and isolate which part of the
previous change was the cause of those problems.
Bug: v8:12808, chromium:1325566, chromium:1325567, chromium:1325601
Change-Id: Ifa086bf27070da8f4b3c0e4415af5ca7b6706b0a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3652252
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#80616}
This is a follow-up CL to https://crrev.com/c/3623542.
When updating pointers during a full GC, a page might not be swept
already. In such cases there might be invalid objects in free memory.
Since these objects might be dead, their maps might have been reclaimed
already as well.
The previous CL cached the size of invalid objects in order to avoid
accessing an invalid object's map. However, as soon as a slot is within
an invalid object, we also need to check whether this slot is still a
tagged pointer which would require map access. This CL checks marking
bits on invalid objects to skip that check on such invalid objects.
Bug: v8:12578, chromium:1316289
Change-Id: Ie1d736f897a2994dbed7bfb95ed37732cd3b0882
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596123
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80609}
This adds a new struct "OOMDetails" which is passed to the
OOMErrorCallback. It currently holds the "is_heap_oom" bool that was
also passed before, plus an optional "detail" string.
The struct can later be extended without having to change the signature
of the OOMErrorCallback. Removing fields will have to follow the
standard deprecation rules, but this is also easily possible without the
hassle for this initial change.
We modify the deprecated OOMErrorCallback definition and un-deprecate it,
which can be seen as removing a deprecated API and adding a new one in
one CL.
R=mlippautz@chromium.org, jkummerow@chromium.org
Bug: chromium:1323177
Change-Id: Ic4c2cb5856906ebd664626fe463d8e96cb99b0a5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3647827
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80565}
This is a reland of commit 2b79eefed3
A DCHECK was using map[key] and inadvertently inserted into the map
that way.
Original change's description:
> Reland^2: [heap] Store size with invalidated object
>
> This is a reland of commit 23b2d571a7
>
> When updating pointers during a full GC, a page might not be swept
> already. In such cases there might be invalid objects and slots
> recorded in free memory. Updating tagged slots in free memory is fine
> even though not strictly necessary.
>
> However, the GC also needs to calculate the size of potentially dead
> invalid objects in order to be able to check whether a slot is within
> that object. But since that object is dead, its map might be dead as
> well which makes size calculation impossible on such objects. The CL
> changes this to cache the size of invalid objects. A follow-up CL will
> also check the marking bit of invalid objects.
>
> Reason for reverts:
>
> Revert #2: In-object slack tracking on JSObjects doesn't update the
> cached size of invalidated objects. The fix here was to stop
> invalidating recorded slots on JSObjects at all and avoid that problem
> completely (see https://crrev.com/c/3620274).
>
> Revert #1: Not all size changes go through NotifyObjectLayoutChange, so
> https://crrev.com/c/3607992 introduced NotifyObjectSizeChange as a
> bottleneck for object size changes/right-trimming. This method is
> now used to update the size of invalidated objects.
>
> Bug: v8:12578, chromium:1316289
> Change-Id: I0478d04601c0270ddb39419ca6cf98719951eb4d
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3623542
> Reviewed-by: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Patrick Thier <pthier@chromium.org>
> Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80344}
Bug: v8:12578, chromium:1316289
Change-Id: Ibcc04c209213c584860a7c473082526cb4e53c59
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3627635
Reviewed-by: Patrick Thier <pthier@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80542}
This CL adds to the existing experimental implementation of the
object start bitmap, that is evaluated as a mechanism for resolving
inner pointers (behind the flag v8_enable_conservative_stack_scanning).
It fixes method ObjectStartBitmap::FindBasePtr to ensure that the
correct base pointer is returned, even if the bitmap is not fully
populated (e.g., with object evacuation or inline object allocation).
This method now recalculates the part of the bitmap that is
required for returning the correct result, by iterating through
objects of the page. A special constructor has been introduced to the
PagedSpaceObjectIterator for this purpose.
It also moves the existing inline methods of ObjectStartBitmap to a
new -inl.h header file, to avoid circular dependencies.
Bug: v8:12851
Change-Id: Iabd0df020bee3bb63ef9d4888591b25d24d79dd9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3641179
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Nikolaos Papaspyrou <nikolaos@chromium.org>
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80538}
NewSpace is renamed to SemiSpaceNewSpace and NewSpaceBase is renamed to
NewSpace (the new PagedSpace new space implementation will be named
PagedNewSpace).
Most usecases are updated to use the base class rather than the concrete
semi space based implementation. To that end, the base class is extended
with additional virtual methods (for delegating to the concrete class).
This CL follows these guidelines:
(*) If at a method callsite we should know the exact new space
implementation we use, we cast to the concrete class. This is the case
for example for callsites in scavenger.*.
(*) If a method is called from outside the heap implementation or should
be present regardless of the concrete implementation, that method is
made virtual.
(*) Other cases are usually methods that are specific to a concrete
implementation but the concrete implementation is not known at the
callsite and there's no clear way to nicely abstract the method. In such
cases we cast to the concrete SemiSpaceNewSpace implementation for now
and we will revisit these cases once PagedNewSpace exists.
Bug: v8:12612
Change-Id: I7b85626774ce0d785b0257bf8d32b9f50eeaf292
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3625975
Commit-Queue: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80482}
Once the root SharedFunctionInfo for any Script gets its bytecode
flushed, the Isolate's compilation cache currently evicts that entry, to
reduce memory usage. However, the associated Script is likely still
alive, since scripts often declare functions which outlive the initial
evaluation of the script. If an identical script is loaded later, a
duplicate Script is created for it, which can waste memory.
In this change, I propose that the compilation cache keys can refer
weakly to the Script. When the root SharedFunctionInfo gets old, instead
of deleting the cache entry entirely, we can just drop the strong
reference to the SharedFunctionInfo. A subsequent lookup in the cache
will retrieve the Script instead of the root SharedFunctionInfo,
indicating an opportunity to save some memory by reusing the existing
Script.
Eventually, all callers to CompilationCache::LookupScript should reuse
the Script if possible. This change implements only the easy case of
reusing the Script for synchronous parsing. Follow-up changes will be
required for the TODO comments left by this change.
Bug: v8:12808
Change-Id: Ia8b0389441a682de9a43e73329049fd2e7835d3d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3597106
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80472}
The issue is that a thread_local variable used in RwxMemoryWriteScope
can't be directly accessed from another component, so the workaround is
to avoid inlining accesses the variable into other components.
Bug: v8:12797, chromium:1324333
Change-Id: I0f83358ac0c663c92ef7b3dff54a068472d61aed
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3641169
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Auto-Submit: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80470}
In some cases we were still trying to change permissions of RWX pages
which is not allowed.
Bug: v8:12797
Change-Id: I9f4ca319d842c524fc6f60bfd3bb3726f8d3029d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3635719
Auto-Submit: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80442}
This CL cleans up the existing experimental implementation of
conservative stack scanning. It retains the object start bitmap, to
evaluate it as a mechanism for resolving inner pointers, and the
conservative stack scanning visitor (which is currently not used).
The flag v8_enable_conservative_stack_scanning is kept and will be
used for experimental purposes. It currently does not imply any
other flag.
Bug: v8:10614
Bug: v8:12851
Change-Id: Id0ae0f437ed2601eed9ec634d2d1dd2f030d814e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3602516
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Commit-Queue: Nikolaos Papaspyrou <nikolaos@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80365}
This reverts commit 2b79eefed3.
Reason for revert: Speculative revert because of https://crbug.com/1322305
Original change's description:
> Reland^2: [heap] Store size with invalidated object
>
> This is a reland of commit 23b2d571a7
>
> When updating pointers during a full GC, a page might not be swept
> already. In such cases there might be invalid objects and slots
> recorded in free memory. Updating tagged slots in free memory is fine
> even though not strictly necessary.
>
> However, the GC also needs to calculate the size of potentially dead
> invalid objects in order to be able to check whether a slot is within
> that object. But since that object is dead, its map might be dead as
> well which makes size calculation impossible on such objects. The CL
> changes this to cache the size of invalid objects. A follow-up CL will
> also check the marking bit of invalid objects.
>
> Reason for reverts:
>
> Revert #2: In-object slack tracking on JSObjects doesn't update the
> cached size of invalidated objects. The fix here was to stop
> invalidating recorded slots on JSObjects at all and avoid that problem
> completely (see https://crrev.com/c/3620274).
>
> Revert #1: Not all size changes go through NotifyObjectLayoutChange, so
> https://crrev.com/c/3607992 introduced NotifyObjectSizeChange as a
> bottleneck for object size changes/right-trimming. This method is
> now used to update the size of invalidated objects.
>
> Bug: v8:12578, chromium:1316289
> Change-Id: I0478d04601c0270ddb39419ca6cf98719951eb4d
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3623542
> Reviewed-by: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Patrick Thier <pthier@chromium.org>
> Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80344}
Bug: v8:12578, chromium:1316289
Change-Id: I6dd2dd13583123b9cb7933b15e2113fddc0c0b0b
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3627511
Owners-Override: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Auto-Submit: Dominik Inführ <dinfuehr@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80357}
This is a reland of commit 23b2d571a7
When updating pointers during a full GC, a page might not be swept
already. In such cases there might be invalid objects and slots
recorded in free memory. Updating tagged slots in free memory is fine
even though not strictly necessary.
However, the GC also needs to calculate the size of potentially dead
invalid objects in order to be able to check whether a slot is within
that object. But since that object is dead, its map might be dead as
well which makes size calculation impossible on such objects. The CL
changes this to cache the size of invalid objects. A follow-up CL will
also check the marking bit of invalid objects.
Reason for reverts:
Revert #2: In-object slack tracking on JSObjects doesn't update the
cached size of invalidated objects. The fix here was to stop
invalidating recorded slots on JSObjects at all and avoid that problem
completely (see https://crrev.com/c/3620274).
Revert #1: Not all size changes go through NotifyObjectLayoutChange, so
https://crrev.com/c/3607992 introduced NotifyObjectSizeChange as a
bottleneck for object size changes/right-trimming. This method is
now used to update the size of invalidated objects.
Bug: v8:12578, chromium:1316289
Change-Id: I0478d04601c0270ddb39419ca6cf98719951eb4d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3623542
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Reviewed-by: Patrick Thier <pthier@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80344}
To be consistent with the all the other tiers and avoid confusion, we
rename --opt to ---turbofan, and --always-opt to --always-turbofan.
Change-Id: Ie23dc8282b3fb4cf2fbf73b6c3d5264de5d09718
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3610431
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80336}
This reverts commit 23b2d571a7.
Reason for revert: Breaks the V8 roll https://ci.chromium.org/ui/p/chromium/builders/try/linux-rel/1000394/
Original change's description:
> Reland "[heap] Store size with invalidated object"
>
> This is a reland of commit 5d235def26
>
> The previous version of this CL got reverted because the cached
> size of an invalidated object wasn't up-to-date when performing a GC.
>
> Not all size changes go through NotifyObjectLayoutChange, so
> https://crrev.com/c/3607992 introduced NotifyObjectSizeChange as a
> bottleneck for object size changes/right-trimming. This method is
> now used to update the size of invalidated objects.
>
> Original change's description:
> > [heap] Store size with invalidated object
> >
> > When updating pointers during a full GC, a page might not be swept
> > already. In such cases there might be invalid objects and slots recorded
> > in free memory. Updating tagged slots in free memory is fine even though
> > it is superfluous work.
> >
> > However, the GC also needs to calculate the size of potentially dead
> > invalid objects in order to be able to check whether a slot is within
> > that object. But since that object is dead, its map might be dead as
> > well which makes size calculation impossible on such objects. The CL
> > changes this to cache the size of invalid objects. A follow-up CL will
> > also check the marking bit of invalid objects.
> >
> > Bug: v8:12578, chromium:1316289
> > Change-Id: Ie773d0862a565982957e0dc409630d76552d1a32
> > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3599482
> > Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> > Reviewed-by: Jakob Linke <jgruber@chromium.org>
> > Reviewed-by: Patrick Thier <pthier@chromium.org>
> > Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> > Cr-Commit-Position: refs/heads/main@{#80169}
>
> Bug: v8:12578, chromium:1316289
> Change-Id: I1f7c6070b8e7d116aeb1a8d03d4f87927ab40872
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3608632
> Reviewed-by: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Patrick Thier <pthier@chromium.org>
> Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80262}
Bug: v8:12578, chromium:1316289
Change-Id: I88b73ebe09bb923ba4ac57b0dbdceb08a1badd99
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3616730
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Auto-Submit: Igor Sheludko <ishell@chromium.org>
Owners-Override: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80291}
This is a reland of commit 5d235def26
The previous version of this CL got reverted because the cached
size of an invalidated object wasn't up-to-date when performing a GC.
Not all size changes go through NotifyObjectLayoutChange, so
https://crrev.com/c/3607992 introduced NotifyObjectSizeChange as a
bottleneck for object size changes/right-trimming. This method is
now used to update the size of invalidated objects.
Original change's description:
> [heap] Store size with invalidated object
>
> When updating pointers during a full GC, a page might not be swept
> already. In such cases there might be invalid objects and slots recorded
> in free memory. Updating tagged slots in free memory is fine even though
> it is superfluous work.
>
> However, the GC also needs to calculate the size of potentially dead
> invalid objects in order to be able to check whether a slot is within
> that object. But since that object is dead, its map might be dead as
> well which makes size calculation impossible on such objects. The CL
> changes this to cache the size of invalid objects. A follow-up CL will
> also check the marking bit of invalid objects.
>
> Bug: v8:12578, chromium:1316289
> Change-Id: Ie773d0862a565982957e0dc409630d76552d1a32
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3599482
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Reviewed-by: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Patrick Thier <pthier@chromium.org>
> Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80169}
Bug: v8:12578, chromium:1316289
Change-Id: I1f7c6070b8e7d116aeb1a8d03d4f87927ab40872
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3608632
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Reviewed-by: Patrick Thier <pthier@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80262}
This is a reland of commit 9d31f8663a
There were issues with --future flag implications on M1.
Original change's description:
> [rwx][mac] Support fast W^X permission switching on Apple Silicon (M1)
>
> ... for V8 code space. The feature is currently disabled.
>
> In order to use fast W^X permission switching we must allocate
> executable pages with readable writable executable permissions (RWX).
> However, MacOS on ARM64 ("Apple M1"/Apple Silicon) prohibits further
> permission changing of RWX memory pages. This means that the code page
> headers must be allocated with RWX permissions too because otherwise
> it wouldn't be possible to allocate a large code page over the freed
> regular code page and vice versa.
>
> When enabled, the new machinery works as follows:
>
> 1) when memory region is reserved for allocating executable pages, the
> whole region is committed with RWX permissions and then decommitted,
> 2) since reconfiguration of RWX page permissions is not allowed on
> MacOS on ARM64 ("Apple M1"/Apple Silicon), there must be no attempts
> to change them,
> 3) the request to set RWX permissions in the executable page region
> just recommits the pages without changing permissions (see (1), they
> were already allocated as RWX and then discarded),
> 4) in order to make executable pages inaccessible one must use
> OS::DiscardSystemPages() instead of OS::DecommitPages() or
> setting permissions to kNoAccess because the latter two are not
> allowed by the MacOS (see (2)).
> 5) since code space page headers are allocated as RWX pages it's also
> necessary to switch between W^X modes when updating the data in the
> page headers (i.e. when marking, updating stats, wiring pages in
> lists, etc.). The new CodePageHeaderModificationScope class is used
> in the respective places. On unrelated configurations it's a no-op.
>
> The fast permission switching can't be used for V8 configuration with
> enabled pointer compression and disabled external code space because
> a) the pointer compression cage has to be reserved with MAP_JIT flag
> which is too expensive,
> b) in case of shared pointer compression cage if the code range will
> be deleted while the cage is still alive then attempt to configure
> permissions of pages that were previously set to RWX will fail.
>
> This also CL extends the unmapper unit tests with permissions tracking
> for discarded pages.
>
> Bug: v8:12797
> Change-Id: Idb28cbc481306477589eee9962d2e75167d87c61
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3579303
> Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Commit-Queue: Igor Sheludko <ishell@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80238}
Bug: v8:12797
Change-Id: I0fe86666f31bad37d7074e217555c95900d2afba
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3610433
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80259}
This reverts commit 9d31f8663a.
Reason for revert: crashes on Mac/arm64 bots:
https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac%20-%20arm64%20-%20debug/5923/overview
Original change's description:
> [rwx][mac] Support fast W^X permission switching on Apple Silicon (M1)
>
> ... for V8 code space. The feature is currently disabled.
>
> In order to use fast W^X permission switching we must allocate
> executable pages with readable writable executable permissions (RWX).
> However, MacOS on ARM64 ("Apple M1"/Apple Silicon) prohibits further
> permission changing of RWX memory pages. This means that the code page
> headers must be allocated with RWX permissions too because otherwise
> it wouldn't be possible to allocate a large code page over the freed
> regular code page and vice versa.
>
> When enabled, the new machinery works as follows:
>
> 1) when memory region is reserved for allocating executable pages, the
> whole region is committed with RWX permissions and then decommitted,
> 2) since reconfiguration of RWX page permissions is not allowed on
> MacOS on ARM64 ("Apple M1"/Apple Silicon), there must be no attempts
> to change them,
> 3) the request to set RWX permissions in the executable page region
> just recommits the pages without changing permissions (see (1), they
> were already allocated as RWX and then discarded),
> 4) in order to make executable pages inaccessible one must use
> OS::DiscardSystemPages() instead of OS::DecommitPages() or
> setting permissions to kNoAccess because the latter two are not
> allowed by the MacOS (see (2)).
> 5) since code space page headers are allocated as RWX pages it's also
> necessary to switch between W^X modes when updating the data in the
> page headers (i.e. when marking, updating stats, wiring pages in
> lists, etc.). The new CodePageHeaderModificationScope class is used
> in the respective places. On unrelated configurations it's a no-op.
>
> The fast permission switching can't be used for V8 configuration with
> enabled pointer compression and disabled external code space because
> a) the pointer compression cage has to be reserved with MAP_JIT flag
> which is too expensive,
> b) in case of shared pointer compression cage if the code range will
> be deleted while the cage is still alive then attempt to configure
> permissions of pages that were previously set to RWX will fail.
>
> This also CL extends the unmapper unit tests with permissions tracking
> for discarded pages.
>
> Bug: v8:12797
> Change-Id: Idb28cbc481306477589eee9962d2e75167d87c61
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3579303
> Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Commit-Queue: Igor Sheludko <ishell@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80238}
Bug: v8:12797
Change-Id: Ic07948e036db36326d464a2a901d052aa060a406
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3611665
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Auto-Submit: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80239}
... for V8 code space. The feature is currently disabled.
In order to use fast W^X permission switching we must allocate
executable pages with readable writable executable permissions (RWX).
However, MacOS on ARM64 ("Apple M1"/Apple Silicon) prohibits further
permission changing of RWX memory pages. This means that the code page
headers must be allocated with RWX permissions too because otherwise
it wouldn't be possible to allocate a large code page over the freed
regular code page and vice versa.
When enabled, the new machinery works as follows:
1) when memory region is reserved for allocating executable pages, the
whole region is committed with RWX permissions and then decommitted,
2) since reconfiguration of RWX page permissions is not allowed on
MacOS on ARM64 ("Apple M1"/Apple Silicon), there must be no attempts
to change them,
3) the request to set RWX permissions in the executable page region
just recommits the pages without changing permissions (see (1), they
were already allocated as RWX and then discarded),
4) in order to make executable pages inaccessible one must use
OS::DiscardSystemPages() instead of OS::DecommitPages() or
setting permissions to kNoAccess because the latter two are not
allowed by the MacOS (see (2)).
5) since code space page headers are allocated as RWX pages it's also
necessary to switch between W^X modes when updating the data in the
page headers (i.e. when marking, updating stats, wiring pages in
lists, etc.). The new CodePageHeaderModificationScope class is used
in the respective places. On unrelated configurations it's a no-op.
The fast permission switching can't be used for V8 configuration with
enabled pointer compression and disabled external code space because
a) the pointer compression cage has to be reserved with MAP_JIT flag
which is too expensive,
b) in case of shared pointer compression cage if the code range will
be deleted while the cage is still alive then attempt to configure
permissions of pages that were previously set to RWX will fail.
This also CL extends the unmapper unit tests with permissions tracking
for discarded pages.
Bug: v8:12797
Change-Id: Idb28cbc481306477589eee9962d2e75167d87c61
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3579303
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80238}
Add flag --always-use-string-forwarding-table to always use the
forwarding table (usually only used for shared strings) instead of
ThinString migrations initially (during GC strings will be migrated
to normal ThinStrings). The goal is to get more coverage of this code
that is designed for shared strings.
Bug: v8:12007
Change-Id: I7eb2e5ccf0018c4ac349611aebe337d8288de5c8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3536650
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Shu-yu Guo <syg@chromium.org>
Commit-Queue: Patrick Thier <pthier@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80206}
Introduce a bottleneck for right-trimming an object. In a subsequent CL
we will use this method to update the cached size of invalidated
objects.
This CL also tries to clean-up the various CreateFillerObjectAt
methods. CreateFillerObjectAtRaw is now the internal method for all
these methods. After moving right-trimming to NotifyObjectSizeChange,
both CreateFillerObjectAt and CreateFillerObjectAtBackground don't need
those arguments for clearing slots or memory anymore.
Bug: v8:12578, chromium:1316289
Change-Id: I6ff0bfaced3e0a1765152700e68a4ad33a155723
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3607992
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80200}
It's necessary to support fast W^X permission switching on MacOS on
ARM64 ("Apple M1"/Apple Silicon) where permission modification of RWX
pages to anything else is prohibited.
On all the other architectures/platforms RecommitPages() is equivalent
to SetPermissions().
The new API will be used in a follow-up CLs.
Bug: v8:12797
Change-Id: Id0d8b8c42c81b80cd8fa6b47c227680d7d1f9b10
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3606231
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Samuel Groß <saelo@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80190}
This reverts commit 5d235def26.
Reason for revert: Speculative revert because of https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux%20-%20gc%20stress/38153/overview
Original change's description:
> [heap] Store size with invalidated object
>
> When updating pointers during a full GC, a page might not be swept
> already. In such cases there might be invalid objects and slots recorded
> in free memory. Updating tagged slots in free memory is fine even though
> it is superfluous work.
>
> However, the GC also needs to calculate the size of potentially dead
> invalid objects in order to be able to check whether a slot is within
> that object. But since that object is dead, its map might be dead as
> well which makes size calculation impossible on such objects. The CL
> changes this to cache the size of invalid objects. A follow-up CL will
> also check the marking bit of invalid objects.
>
> Bug: v8:12578, chromium:1316289
> Change-Id: Ie773d0862a565982957e0dc409630d76552d1a32
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3599482
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Reviewed-by: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Patrick Thier <pthier@chromium.org>
> Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80169}
Bug: v8:12578, chromium:1316289
Change-Id: I6949412c5d6e1aa15718d027043d9528137a60a0
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3605812
Auto-Submit: Nico Hartmann <nicohartmann@chromium.org>
Owners-Override: Nico Hartmann <nicohartmann@chromium.org>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#80173}
When updating pointers during a full GC, a page might not be swept
already. In such cases there might be invalid objects and slots recorded
in free memory. Updating tagged slots in free memory is fine even though
it is superfluous work.
However, the GC also needs to calculate the size of potentially dead
invalid objects in order to be able to check whether a slot is within
that object. But since that object is dead, its map might be dead as
well which makes size calculation impossible on such objects. The CL
changes this to cache the size of invalid objects. A follow-up CL will
also check the marking bit of invalid objects.
Bug: v8:12578, chromium:1316289
Change-Id: Ie773d0862a565982957e0dc409630d76552d1a32
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3599482
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Reviewed-by: Patrick Thier <pthier@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80169}
This CL extends BoundedPageAllocator with PageFreeingMode parameter
which controls how pages should be freed: by setting permissions to
kNoAccess (preferred) or by discarding pages (Apple Silicon specific
behavior for RWX pages). The latter mode allows to ensure that once
pages are configured with RWX permissions they are never reconfigured
to anything else again.
The new mode will be used in a follow-up CL.
Bug: v8:12797
Change-Id: I3277f56ea6fee9c9b38b1682e68c22e66e9a02a4
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3606228
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80162}
Normally, taking a heap snapshot in the near heap limit would
result in a full GC, then the overhead of the promotions would
cause another invocation of the heap limit callback and it can
raise the limit in the second call to avoid an OOM, so we test
that the callback can indeed raise the limit this way in this
case. When there is only one generation, however, there would
not be the overhead of promotions so the callback may not be
triggered again during the generation of the heap snapshot.
In that case we only need to check that the callback is called
and it can perform GC-triggering operations jsut fine there.
Bug: v8:12815
Change-Id: If244417624b56bc068aed480fb3391d26c19005a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3600357
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Joyee Cheung <joyee@igalia.com>
Cr-Commit-Position: refs/heads/main@{#80094}
