This is a step towards encoding all the necessary information in
the feedback slot kind instead of storing it in the IC dispatcher's
code object flags.
BUG=v8:5849, v8:5917
Review-Url: https://codereview.chromium.org/2662113005
Cr-Commit-Position: refs/heads/master@{#42859}
The String.prototype was altered after snapshot time (during
experimental natives setup), invalidating the stored map used for
fast-path checks.
BUG=
Review-Url: https://codereview.chromium.org/2663303003
Cr-Commit-Position: refs/heads/master@{#42842}
They have the same lifetime. It's a match!
Both structures are native context dependent and dealt with (creation,
clearing, gathering feedback) at the same time. By treating the spaces used
for literal boilerplates as feedback vector slots, we no longer have to keep
track of the materialized literal count elsewhere.
A follow-on CL removes even more parser infrastructure related to this count.
BUG=v8:5456
Review-Url: https://codereview.chromium.org/2655853010
Cr-Commit-Position: refs/heads/master@{#42771}
This test checks that counters accurately reflect the allocated size.
There's an edge case that can occur when, previously to the allocation,
the page does not have enough space left to allocate the requested
object - then we move on to a fresh page, fill the remaining space of
the old page with a filler object, and allocate the requested object on
the new page.
The counters will show the size of the filler object plus the requested
object size, while the test expects only the requested size.
This CL fixes that case by performing two GCs to clear out new space.
BUG=
Review-Url: https://codereview.chromium.org/2652933002
Cr-Commit-Position: refs/heads/master@{#42646}
Manipulating the signaling NaN used for the hole and uninitialized double
field sentinel in C++, e.g. with bit_cast or HeapNumber::value()/set_value(),
will change its value on ia32 (the x87 stack is used to return values and
stores to the stack silently clear the signalling bit).
BUG=v8:5495
Review-Url: https://codereview.chromium.org/2652553003
Cr-Commit-Position: refs/heads/master@{#42609}
Also update a call in cctest to check the result.
BUG=chromium:681843
Review-Url: https://codereview.chromium.org/2647573003
Cr-Commit-Position: refs/heads/master@{#42513}
Background: the first page of each space is implicitly immovable.
Recently, our builtin code objects have reached a size at which we
fill up the first page of code space during initialization. Once
that occurs, newly requested allocations of immovable code are
allocated in a large object space page of 512K.
This CL mitigates these effects by simply marking pages as immovable
during snapshot creation instead of going into LO space.
On snapshot builds, this should just work: deserialized pages are
trimmed and marked immovable when deserialization finishes.
However, non-snapshot builds and allocations of immovable CEntryStub
code at runtime are still affected.
BUG=v8:5831
Review-Url: https://codereview.chromium.org/2635973002
Cr-Commit-Position: refs/heads/master@{#42411}
Literal arrays and feedback vectors for a function can be garbage
collected if we don't have a rooted closure for the function, which
happens often. It's expensive to come back from this (recreating
boilerplates and gathering feedback again), and the cost is
disproportionate if the function was inlined into optimized code.
To guard against losing these arrays when we need them, we'll now
create literal arrays when creating the feedback vector for the outer
closure, and root them strongly in that vector.
BUG=v8:5456
Review-Url: https://codereview.chromium.org/2620753003
Cr-Original-Commit-Position: refs/heads/master@{#42258}
Committed: 3188780410
Review-Url: https://codereview.chromium.org/2620753003
Cr-Commit-Position: refs/heads/master@{#42264}
Reason for revert:
gc stress:
https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20gc%20stress/builds/8105
also on mac
Original issue's description:
> [TypeFeedbackVector] Root literal arrays in function literals slots
>
> Literal arrays and feedback vectors for a function can be garbage
> collected if we don't have a rooted closure for the function, which
> happens often. It's expensive to come back from this (recreating
> boilerplates and gathering feedback again), and the cost is
> disproportionate if the function was inlined into optimized code.
>
> To guard against losing these arrays when we need them, we'll now
> create literal arrays when creating the feedback vector for the outer
> closure, and root them strongly in that vector.
>
> BUG=v8:5456
>
> Review-Url: https://codereview.chromium.org/2620753003
> Cr-Commit-Position: refs/heads/master@{#42258}
> Committed: 3188780410TBR=bmeurer@chromium.org,mstarzinger@chromium.org,yangguo@chromium.org,mvstanton@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5456
Review-Url: https://codereview.chromium.org/2626863004
Cr-Commit-Position: refs/heads/master@{#42260}
Literal arrays and feedback vectors for a function can be garbage
collected if we don't have a rooted closure for the function, which
happens often. It's expensive to come back from this (recreating
boilerplates and gathering feedback again), and the cost is
disproportionate if the function was inlined into optimized code.
To guard against losing these arrays when we need them, we'll now
create literal arrays when creating the feedback vector for the outer
closure, and root them strongly in that vector.
BUG=v8:5456
Review-Url: https://codereview.chromium.org/2620753003
Cr-Commit-Position: refs/heads/master@{#42258}
Downside: this adds all kinds of weird includes in the .cc files.
(See design doc linked in the bug.)
BUG=v8:5402
Review-Url: https://codereview.chromium.org/2622503002
Cr-Commit-Position: refs/heads/master@{#42140}
(Missing includes discovered during the objects.h splitting work.)
BUG=v8:5402
Review-Url: https://codereview.chromium.org/2610643002
Cr-Commit-Position: refs/heads/master@{#42029}
Reason for revert:
Speculative revert because of blocked roll: https://codereview.chromium.org/2596013002/
Original issue's description:
> [TypeFeedbackVector] Root literal arrays in function literals slots
>
> Literal arrays and feedback vectors for a function can be garbage
> collected if we don't have a rooted closure for the function, which
> happens often. It's expensive to come back from this (recreating
> boilerplates and gathering feedback again), and the cost is
> disproportionate if the function was inlined into optimized code.
>
> To guard against losing these arrays when we need them, we'll now
> create literal arrays when creating the feedback vector for the outer
> closure, and root them strongly in that vector.
>
> BUG=v8:5456
>
> Review-Url: https://codereview.chromium.org/2504153002
> Cr-Commit-Position: refs/heads/master@{#41893}
> Committed: 93df094081TBR=bmeurer@chromium.org,mlippautz@chromium.org,mvstanton@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5456
Review-Url: https://codereview.chromium.org/2597163002
Cr-Commit-Position: refs/heads/master@{#41917}
Literal arrays and feedback vectors for a function can be garbage
collected if we don't have a rooted closure for the function, which
happens often. It's expensive to come back from this (recreating
boilerplates and gathering feedback again), and the cost is
disproportionate if the function was inlined into optimized code.
To guard against losing these arrays when we need them, we'll now
create literal arrays when creating the feedback vector for the outer
closure, and root them strongly in that vector.
BUG=v8:5456
Review-Url: https://codereview.chromium.org/2504153002
Cr-Commit-Position: refs/heads/master@{#41893}
Reason for revert:
LiveEdit is broken in some cases.
Original issue's description:
> Store SharedFunctionInfos of a Script in a FixedArray indexed by their ID
>
> Now that SharedFunctionInfos have a unique ID (and the IDs are dense),
> we can use them as an index into an array, instead of using a
> WeakFixedArray where we have to do a linear scan.
>
> Hooking up liveedit is a bit more involved, see
> https://docs.google.com/presentation/d/1FtNa3U7WsF5bPhY9uGoJG5Y9hnz5VBDabfOWpb4unWI/edit
> for an overview
>
> BUG=v8:5589
> R=verwaest@chromium.org,jgruber@chromium.org
>
> Committed: https://crrev.com/6595e7405769dc9d49e9568d61485efc6d468baf
> Cr-Commit-Position: refs/heads/master@{#41600}
TBR=jgruber@chromium.org,verwaest@chromium.org,yangguo@chromium.org,jochen@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=v8:5589,chromium:673950
NOPRESUBMIT=true
Review-Url: https://codereview.chromium.org/2578433002
Cr-Commit-Position: refs/heads/master@{#41684}
come from the runtime.
This patch fixes an issue of heap growing to max capacity when incremental
marking is finished but cannot finalize due to GC stack guard not triggering.
It can happen if all allocations come from the runtime, for example,
from JSON parser or compiler.
Now before expanding the heap we check if we are above the allocation limit
and the incremental marking needs to be finalized. If so we do not expand
the heap and force GC, which will finalize the incremental marking.
The check is performed for paged spaces and large-object space.
BUG=chromium:670675
Review-Url: https://codereview.chromium.org/2552613004
Cr-Commit-Position: refs/heads/master@{#41524}
Adds a bytecode_age field to BytecodeArray objects. This is incremented each
time the bytecode array is marked by GC, and reset to zero if the bytecode
is executed.
This is used to enable the CompilationCache for interpreted functions,
where Interpreted entries are evicted once the bytecode becomes old.
BUG=chromium:666275,v8:4680
Review-Url: https://codereview.chromium.org/2534763003
Cr-Commit-Position: refs/heads/master@{#41356}
MarkingParity was used to avoid performing an operation on an object if it was
marked multiple times. We no longer mark things multiple times, so this concept
is no longer required.
BUG=chromium:666275
Review-Url: https://codereview.chromium.org/2529173002
Cr-Commit-Position: refs/heads/master@{#41354}
If code is flushed on a SFI, we can still use the bytecode if it was compiled,
since this never gets flushed.
This fixes a DCHECK where we were trying to compile the bytecode multiple
times after the baseline code was flushed.
BUG=chromium:668133
Review-Url: https://codereview.chromium.org/2526243002
Cr-Commit-Position: refs/heads/master@{#41274}
Adds the marking logic to mark the young generation.
BUG=chromium:651354
Review-Url: https://codereview.chromium.org/2498583002
Cr-Commit-Position: refs/heads/master@{#41104}
Adds an IsInterpreted() function to both SharedFunctionInfo and JSFunction.
This is used to fix the test-heap code-aging tests since Ignition doesn't
age code.
BUG=v8:4680
Review-Url: https://codereview.chromium.org/2481433002
Cr-Commit-Position: refs/heads/master@{#40868}
This is an experiment to check whether the heuristics is still useful.
BUG=
Review-Url: https://codereview.chromium.org/2482163002
Cr-Commit-Position: refs/heads/master@{#40833}
This enables Ignition unconditionally for all code that is destined for
optimization with TurboFan. This ensures all optimization attempts will
go through the BytecodeGraphBuilder and that the AstGraphBuilder pipe is
dried out in practice.
patch from issue 2427953002 at patchset 120001 (http://crrev.com/2427953002#ps120001)
R=mvstanton@chromium.org,rmcilroy@chromium.org
Review-Url: https://codereview.chromium.org/2453973004
Cr-Commit-Position: refs/heads/master@{#40663}
Native setters (see AccessorInfo in accessors.h) didn't have the ability
to return a result value. As a consequence of this, for instance, Reflect.set
on the length property of arrays had the wrong behavior:
var y = [];
Object.defineProperty(y, 0, {value: 42, configurable: false})
Reflect.set(y, 'length', 0)
The Reflect.set call used to return true. Now it returns false as
required by the spec.
BUG=v8:5401
Review-Url: https://codereview.chromium.org/2397603003
Cr-Commit-Position: refs/heads/master@{#40579}
This patch moves management of marking deque backing store into the
MarkingDeque class, which will simplify unmapping of backing store in
concurrent task.
BUG=
Review-Url: https://codereview.chromium.org/2439063002
Cr-Commit-Position: refs/heads/master@{#40523}
Additionally, remove all code related to the old-style slots filtering and black area end markers.
BUG=chromium:648568
Review-Url: https://chromiumcodereview.appspot.com/2440683002
Cr-Commit-Position: refs/heads/master@{#40494}
This enables Ignition unconditionally for all code that is destined for
optimization with TurboFan. This ensures all optimization attempts will
go through the BytecodeGraphBuilder and that the AstGraphBuilder pipe is
dried out in practice.
R=mvstanton@chromium.org
Review-Url: https://chromiumcodereview.appspot.com/2427953002
Cr-Commit-Position: refs/heads/master@{#40462}
When the test is executed in user mode qemu for mips64, it fails because
the buffer is too small. Increasing the buffer fixes the problem.
BUG=
TEST=cctest/test-mark-compact/RegressJoinThreadsOnIsolateDeinit
Review-Url: https://codereview.chromium.org/2416573002
Cr-Commit-Position: refs/heads/master@{#40278}
This reverts commit 7db0ecdec3.
Manual revert since automatic revert is too large for the web interface.
BUG=
TBR=bmeurer@chromium.org,mstarzinger@chromium.org,yangguo@chromium.org,ahaas@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
Review-Url: https://codereview.chromium.org/2396353002
Cr-Commit-Position: refs/heads/master@{#40082}
The scavenger should never consider mark bits for promotion/copy as this creates
weird livetimes at the start of incremental marking. E.g. consider an object
marked black by the marker at the start of incremental marking. A scavenge would
promote it to the old generation although it could --and for short-living
objects actually does-- become unreachable during marking
Also, keeping this invariant significantly simplifies young generation mark
compacting as we can compare against the scavenging decision without keeping
different sets of markbits.
BUG=chromium:651354
R=hpayer@chromium.org
Review-Url: https://codereview.chromium.org/2397713002
Cr-Commit-Position: refs/heads/master@{#40026}
On PPC64 linux the OS page size is 64KB, therefore when the
snapshot is created the serialized heap already has LO_SPACE
allocated(the allocation goes beyond the 1st page of
CODE_SPACE and hence LO_SPACE is allocated).
I've updated the testcase to check if the delta
of the LO_SPACE is zero.
R=mlippautz@chromium.org, ulan@chromium.org, vogelheim@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2394893002
Cr-Commit-Position: refs/heads/master@{#40014}
Reason for revert:
No real improvement as we still lack the ability to promote from
scavenges/young gen GCs.
Let's keep this in mind for later.
Original issue's description:
> [heap] Remove border page
>
> A page now belongs either the nursery *or* the intermediate gen. The page that
> contained objects of both spaces is removed in this change.
>
> BUG=chromium:636331
>
> Committed: https://crrev.com/42ece47446f0dbd3779d6e0e00dce97a1931a9f9
> Cr-Commit-Position: refs/heads/master@{#39778}
TBR=ulan@chromium.org,hpayer@chromium.org
BUG=chromium:636331
Review-Url: https://codereview.chromium.org/2383443002
Cr-Commit-Position: refs/heads/master@{#39854}
Reason for revert:
OOMs in nosnap debug:
https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20nosnap%20-%20debug/builds/9572
Original issue's description:
> [heap] New heuristics for starting of incremental marking.
>
> The motivation for this patch is to move more marking work to tasks.
> This is done by postponing the start of incremental marking until
> a marking task is running.
>
> This patch introduces a soft and a hard limits for incremental marking.
> When the soft limit is reached, the marking task is scheduled.
> If the hard limit is reached before the task is running, then
> incremental marking is started without waiting for the task.
>
> BUG=chromium:616434
> LOG=NO
>
> Committed: https://crrev.com/55683ddd2a32e0dfb8df66271fbf53e3618cce9d
> Cr-Commit-Position: refs/heads/master@{#39831}
TBR=hpayer@chromium.org,ulan@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:616434
Review-Url: https://codereview.chromium.org/2375983002
Cr-Commit-Position: refs/heads/master@{#39833}
The motivation for this patch is to move more marking work to tasks.
This is done by postponing the start of incremental marking until
a marking task is running.
This patch introduces a soft and a hard limits for incremental marking.
When the soft limit is reached, the marking task is scheduled.
If the hard limit is reached before the task is running, then
incremental marking is started without waiting for the task.
BUG=chromium:616434
LOG=NO
Review-Url: https://codereview.chromium.org/2364923002
Cr-Commit-Position: refs/heads/master@{#39831}
This patch simplifies code for speeding up marking and
removes write barrier counter.
The step size is now computed based in two parts:
- bytes to mark in order to keep up with allocation,
- bytes to mark in order to make progress.
BUG=chromium:616434, chromium:646139, chromium:644819
LOG=NO
Review-Url: https://codereview.chromium.org/2359903002
Cr-Commit-Position: refs/heads/master@{#39827}
A page now belongs either the nursery *or* the intermediate gen. The page that
contained objects of both spaces is removed in this change.
BUG=chromium:636331
Review-Url: https://codereview.chromium.org/2209583002
Cr-Commit-Position: refs/heads/master@{#39778}
This patch changes incremental marking work scheduling from combination
of idle/delayed tasks to ordinary short-running tasks and moves
more marking work from V8.Execute to tasks by accounting how much
bytes were marked in tasks.
BUG=chromium:616434
LOG=NO
Review-Url: https://codereview.chromium.org/2321553002
Cr-Commit-Position: refs/heads/master@{#39348}
Now callers of Heap::CollectGarbage* functions need to
specify the reason as an enum value instead of a string.
Subsequent CL will add stats counter for GC reason.
BUG=
Review-Url: https://codereview.chromium.org/2310143002
Cr-Commit-Position: refs/heads/master@{#39239}
This patch
- extracts the logic of keeping track of allocated bytes
from the actual incremental marking step.
- replaces OldSpaceStep with a check for incremental marking start.
- removes the force_marking parameter of AdvanceIncrementalMarking.
BUG=chromium:616434
LOG=NO
Review-Url: https://codereview.chromium.org/2304123003
Cr-Commit-Position: refs/heads/master@{#39213}
This way we avoid the cyclic dependency between objects.h and heap.h and still
have one definition. Add a static assert that this size is indeed smaller than
the payload of a page.
Follow ups can finally remove the dependency on spaces.h for all heap.h users.
R=ulan@chromium.org,bmeurer@chromium.org,vogelheim@chromium.og
Review-Url: https://codereview.chromium.org/2311203002
Cr-Commit-Position: refs/heads/master@{#39206}
Replace first page size in the snapshots with a heap logic that trims pages
after deserialization. The snapshot provided page sizes was just an
approximation, while the heap knows exactly where to trim.
Furthermore, trim the pages directly after deserialization, leaving no wiggle
room for further objects. This avoids pollution of the immortal immovable pages
with regular objects, e.g. Contexts. The downside is that we potentially require
expanding the space with a new page.
BUG=chromium:636331
Review-Url: https://codereview.chromium.org/2311963002
Cr-Commit-Position: refs/heads/master@{#39200}
Rebuilding (after touching certain files) is crazy slow because
includes are out of control. Many of these files we need to rebuild are
cctests which pull in more includes than they need.
BUG=v8:5294
Review-Url: https://codereview.chromium.org/2304553002
Cr-Commit-Position: refs/heads/master@{#39080}
Reason for revert:
Tanks pretty much alle metrics across the board. Probably LO space limit too low but needs investigation.
Original issue's description:
> [heap] Switch to 500k pages
>
> Decrease regular heap object size to 400k. In a follow up, we can now get rid of
> the new space border page while keeping the 1M minimum new space size.
>
> This reverts commit 1617043c10.
>
> BUG=chromium:636331
>
> Committed: https://crrev.com/2101e691caeef656eb91f1c98620b3955d337c83
> Cr-Commit-Position: refs/heads/master@{#38916}
TBR=ulan@chromium.org,verwaest@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=chromium:636331
NOPRESUBMIT=true
Review-Url: https://codereview.chromium.org/2289493002
Cr-Commit-Position: refs/heads/master@{#38960}
Decrease regular heap object size to 400k. In a follow up, we can now get rid of
the new space border page while keeping the 1M minimum new space size.
This reverts commit 1617043c10.
BUG=chromium:636331
Review-Url: https://codereview.chromium.org/2278653003
Cr-Commit-Position: refs/heads/master@{#38916}
Before this patch all tracing scopes in incremental marking would be reset
during a gc tracer start/stop cycle. This patch handles scopes the same way it
does other incremental marking metrics.
Also:
- Align finalization metric with regular marking metric.
- Smaller cleanups
BUG=chromium:639818
Review-Url: https://codereview.chromium.org/2273673002
Cr-Commit-Position: refs/heads/master@{#38834}
Before this patch all tracing scopes in incremental marking would be reset
during a gc tracer start/stop cycle. This patch handles scopes the same way it
does other incremental marking metrics.
Also:
- Align finalization metric with regular marking metric.
- Smaller cleanups
BUG=chromium:639818
R=jochen@chromium.org
Review-Url: https://codereview.chromium.org/2264033002
Cr-Commit-Position: refs/heads/master@{#38822}
Reason for revert:
Tanks octane
Original issue's description:
> [heap] Improve size profiling for ArrayBuffer tracking
>
> Eagerly account for retained sizes during ArrayBuffer tracking. Following up on this,
> we can now do Scavenges if the amount of memory retained from new space is too large.
>
> BUG=chromium:621829
> R=jochen@chromium.org,hpayer@chromium.org
>
> Committed: https://crrev.com/28e13bd6a75c9467dae43043e7b741a1387d5252
> Cr-Commit-Position: refs/heads/master@{#38731}
TBR=jochen@chromium.org,hpayer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:621829
Review-Url: https://codereview.chromium.org/2261513003
Cr-Commit-Position: refs/heads/master@{#38739}
Eagerly account for retained sizes during ArrayBuffer tracking. Following up on this,
we can now do Scavenges if the amount of memory retained from new space is too large.
BUG=chromium:621829
R=jochen@chromium.org,hpayer@chromium.org
Review-Url: https://codereview.chromium.org/2210263002
Cr-Commit-Position: refs/heads/master@{#38731}
Slots in ConsString/SlicedString can point to an evacutaion candidate.
The MakeExternal function makes in-place conversion to external string.
After the conversion we can have a recorded slot containing an external
pointer. As long as the external pointer is aligned, this is not a
problem. We clear the recorded slots to fix verify-heap checks.
BUG=chromium:631969
LOG=NO
Finalizing CL: https://codereview.chromium.org/2199863002/
Review-Url: https://codereview.chromium.org/2242183003
Cr-Commit-Position: refs/heads/master@{#38653}
Decrease regular heap object size to 400k. In a follow up, we can now get rid of
the new space border page while keeping the 1M minimum new space size.
BUG=chromium:636331
This reverts commit 555c961990.
Review-Url: https://codereview.chromium.org/2232653003
Cr-Commit-Position: refs/heads/master@{#38608}
