Commit Graph

35 Commits

Author SHA1 Message Date
Daco Harkes a8b93d9e78 [vm/ffi] DBC support on x64 Linux/MacOS hosts
Issue: https://github.com/dart-lang/sdk/issues/35773

Change-Id: Ib58fb8df6c5c18e604fbf2a156ab69025e5f57d7
Cq-Include-Trybots: luci.dart.try:vm-kernel-linux-debug-simdbc64-try, vm-kernel-linux-release-simdbc64-try, vm-kernel-mac-debug-simdbc64-try, vm-kernel-mac-release-simdbc64-try, vm-kernel-reload-mac-debug-simdbc64-try, vm-kernel-reload-mac-release-simdbc64-try, vm-kernel-linux-debug-ia32-try, vm-dartkb-linux-debug-simarm64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/97221
Reviewed-by: Samir Jindel <sjindel@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
2019-05-09 11:46:03 +00:00
Vyacheslav Egorov a9ce969e53 [vm] Decouple growable_array.h and zone.h from thread.h
- Introduce a slimmed down version of thread.h, which just depends on the
Zone and StackResource.
- Introduce a layering check that would prevent the coupling in the future.

This is the first step towards decoupling compiler from runtime.

There are multiple reasons to introduce the decoupling but the main
reason currently is to introduce a controlled surface through which
compiler reaches into runtime to catch any places where runtime word size
might influence the compiler and then enable building compiler that
targets 32-bit runtime but is embedded into a 64-bit runtime.

Issue https://github.com/dart-lang/sdk/issues/31709

Change-Id: Id63ebbaddca55dd097298e51c90d957a73fa476e
Reviewed-on: https://dart-review.googlesource.com/c/87182
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
2019-01-11 20:47:10 +00:00
Ryan Macnak 30896b87f1 [vm] Fix crash when enum initialization hits a stack overflow.
Give the compiler more headroom.

Have the simulators and interpreters distinguish between the true stack limit and the limit for trigger an overflow exception.

Bug: https://github.com/dart-lang/sdk/issues/35224
Bug: https://github.com/flutter/flutter/issues/25041
Change-Id: I048d73b649b4ce1f3d9b5a33d898ee9cfe54f6ca
Reviewed-on: https://dart-review.googlesource.com/c/86821
Reviewed-by: Aart Bik <ajcbik@google.com>
Reviewed-by: Régis Crelier <regis@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
2018-12-10 23:40:21 +00:00
Ryan Macnak 4463aa0038 Revert "[vm] Fix crash when enum initialization hits a stack overflow."
This reverts commit e59312dafa.

Reason for revert: Broken arm and arm64 simulators

Original change's description:
> [vm] Fix crash when enum initialization hits a stack overflow.
> 
> Give the compiler more headroom.
> 
> Have the KBC and DBC interpreters distinguish between the true stack limit and the limit for trigger an overflow exception.
> 
> Bug: https://github.com/dart-lang/sdk/issues/35224
> Bug: https://github.com/flutter/flutter/issues/25041
> Change-Id: I85e72e1d3cceb70858c53567a4c8d3635e42165b
> Reviewed-on: https://dart-review.googlesource.com/c/86684
> Reviewed-by: Régis Crelier <regis@google.com>
> Commit-Queue: Ryan Macnak <rmacnak@google.com>

TBR=rmacnak@google.com,regis@google.com,ajcbik@google.com

Change-Id: I923cca32d34de37b761196f536873aaf81dcbc92
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: https://github.com/dart-lang/sdk/issues/35224, https://github.com/flutter/flutter/issues/25041
Reviewed-on: https://dart-review.googlesource.com/c/86782
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
2018-12-10 18:52:10 +00:00
Ryan Macnak e59312dafa [vm] Fix crash when enum initialization hits a stack overflow.
Give the compiler more headroom.

Have the KBC and DBC interpreters distinguish between the true stack limit and the limit for trigger an overflow exception.

Bug: https://github.com/dart-lang/sdk/issues/35224
Bug: https://github.com/flutter/flutter/issues/25041
Change-Id: I85e72e1d3cceb70858c53567a4c8d3635e42165b
Reviewed-on: https://dart-review.googlesource.com/c/86684
Reviewed-by: Régis Crelier <regis@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
2018-12-10 18:29:43 +00:00
Martin Kustermann 08a4e020b3 [VM] Get rid of duplicate return-type-cid information from method recognizer
When @pragma('vm:exact-result-type') was introduced, the return types
specified in the recognized method lists were **duplicated** (they were
in the method recognizer macro lists as well as in the dart source code).

This CL removes this duplication by removing the result types from the
method recognizer lists.

Change-Id: I2082110cf77fc3b97d1a541cb153abb8c17990d0
Reviewed-on: https://dart-review.googlesource.com/c/84220
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Samir Jindel <sjindel@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
2018-11-22 11:53:27 +00:00
Ben Konyi 07852532e3 [ VM ] Additional cleanup in Dart_Initialize and Dart_Cleanup
Change-Id: I6dc02b3d9de16cc176eb97613bc0c7f0bb9b16eb
Reviewed-on: https://dart-review.googlesource.com/77013
Commit-Queue: Ben Konyi <bkonyi@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Auto-Submit: Ben Konyi <bkonyi@google.com>
2018-09-28 23:18:59 +00:00
Régis Crelier 14f3982a1f [VM runtime] Remove broken LongJumpScope::IsSafeToJump().
Over the years, this debug-only code evolved to check that the current stack
pointer is below the last exit frame, which is obviously always the case.

Change-Id: I9b5e1a5462df5b3392b4541e40a9c57d33d25055
Reviewed-on: https://dart-review.googlesource.com/70960
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Régis Crelier <regis@google.com>
2018-08-21 19:29:09 +00:00
Martin Kustermann cf1de7d46c [VM] Replace hand-written assembly prologues with IR
As part of the prologue changes we get rid of the empty context as well.

Issue https://github.com/dart-lang/sdk/issues/31495

Change-Id: I707e23c631bcfbbad6c91c4963d0c10f7a0be625
Reviewed-on: https://dart-review.googlesource.com/25320
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
2017-12-13 16:04:34 +00:00
Ryan Macnak 4f612b749a [vm] Only support instruction tracing for DBC in debug mode.
Flutter uses DBC in release mode.

dart-simdbc-checked linux-armv7hf
ParticleSystemUpdate (Odroid-C2) -32.73% (-0.8 noise)
Tracer (Odroid-C2)                17.72% (0.0 noise)
StarryStrings (Odroid-C2)         45.95% (0.1 noise)
Richards (Odroid-C2)              49.89% (0.1 noise)
DeltaBlueClosures (Odroid-C2)     54.56% (0.1 noise)
DeltaBlue (Odroid-C2)             54.27% (3.7 noise)
ParticleSystemPaint (Odroid-C2)   56.29% (4.8 noise)

Change-Id: I42f9697c7a66f2d8bfb665c32aed90e833d48a75
Reviewed-on: https://dart-review.googlesource.com/21283
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Zach Anderson <zra@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
2017-11-15 23:42:01 +00:00
Ryan Macnak 204f7f6ffa [vm] Visit DBC interpreter's pp and argdesc slot.
Enables compaction with DBC.

dart-simdbc-checked linux-armvhf (closest to Flutter's usage):
ParticleSystemPaint (Odroid-C2)  -3.454% (-0.3 noise)
ParticleSystemUpdate (Odroid-C2) -9.009% (-0.2 noise)
DeltaBlue (Odroid-C2)            -1.655% (-0.1 noise)
DeltaBlueClosures (Odroid-C2)    -2.415% (-0.0 noise)
StarryStrings (Odroid-C2)        -2.445% (-0.0 noise)
Richards (Odroid-C2)            -0.9325% (-0.0 noise)
Tracer (Odroid-C2)              -0.6846% (-0.0 noise)

Bug: https://github.com/dart-lang/sdk/issues/30978
Change-Id: Iff7dfad25db0a1e12fd02157c7c4f970ea88779d
Reviewed-on: https://dart-review.googlesource.com/16689
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
2017-11-15 22:07:14 +00:00
Ryan Macnak b85a7fecd0 Reapply "[vm] Prefer stack bounds from the system thread library for overflow checks."
Mark some profiler tests on Windows as failing.

Bug: https://github.com/dart-lang/sdk/issues/31137
Change-Id: I46a5d4838443e5cfdac6043ea8dcfeb3d05d4823
Reviewed-on: https://dart-review.googlesource.com/14920
Reviewed-by: Zach Anderson <zra@google.com>
2017-10-18 21:15:49 +00:00
Dmitry Stefantsov a227b168da Revert "Reapply "[vm] Prefer stack bounds from the system thread library for overflow checks.""
This reverts commit 5a022b9db6.

The reverted CL caused some redness on some vm and vm-precomp bots.

TBR=rmacnak@google.com
Change-Id: I9be670f3c3e341ec6d95082f288bc569834e38c7
Reviewed-on: https://dart-review.googlesource.com/13460
Reviewed-by: Dmitry Stefantsov <dmitryas@google.com>
Reviewed-by: Karl Klose <karlklose@google.com>
Commit-Queue: Dmitry Stefantsov <dmitryas@google.com>
2017-10-12 07:32:30 +00:00
Ryan Macnak 5a022b9db6 Reapply "[vm] Prefer stack bounds from the system thread library for overflow checks."
Fix accounting of guard pages to move the stack limit up instead of down.

Change-Id: I2f6577a12cf95707c186884f0bd3def03eac12ed
Reviewed-on: https://dart-review.googlesource.com/13381
Reviewed-by: Zach Anderson <zra@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
2017-10-12 02:17:11 +00:00
Ryan Macnak 6ecf6dd0a3 Revert "[vm] Prefer stack bounds from the system thread library for overflow checks."
This reverts commit af251ddfec.

Reason for revert: Crashes on Windows.

Original change's description:
> [vm] Prefer stack bounds from the system thread library for overflow checks.
> 
> Before this change, we would assume all stacks have the size the VM specifies for threads it creates, but this won't be accurate for the initial thread or threads created by the embedder.
> 
> Change-Id: I6605a88d6666a6f9d47fbe047d3fdd02aa22c80a
> Reviewed-on: https://dart-review.googlesource.com/10209
> Reviewed-by: Zach Anderson <zra@google.com>

TBR=rmacnak@google.com,zra@google.com

Change-Id: I96eaccdc8edf6e3d319827a63d168534bad0518e
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/13106
Reviewed-by: Ryan Macnak <rmacnak@google.com>
2017-10-11 21:17:28 +00:00
Ryan Macnak af251ddfec [vm] Prefer stack bounds from the system thread library for overflow checks.
Before this change, we would assume all stacks have the size the VM specifies for threads it creates, but this won't be accurate for the initial thread or threads created by the embedder.

Change-Id: I6605a88d6666a6f9d47fbe047d3fdd02aa22c80a
Reviewed-on: https://dart-review.googlesource.com/10209
Reviewed-by: Zach Anderson <zra@google.com>
2017-10-11 19:35:44 +00:00
Vyacheslav Egorov 8a179fb953 [VM, Compiler] Move compiler to a separate folder.
New folder structure (nested under vm/):

- compiler/
-   jit/         - JIT specific code
-   aot/         - AOT specific code
-   backend/     - all middle-end and back-end code (IL, flow graph)
-   assembler/   - assemblers and disassemblers
-   frontend/    - front ends (AST -> IL, Kernel -> IL)

compiler/README.md would be the documentation root for the compiler
pipeline

Bug: https://github.com/dart-lang/sdk/issues/30575
Change-Id: I2dfd9688793bff737f7632ddc77fca766875ce36
Reviewed-on: https://dart-review.googlesource.com/2940
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
2017-09-04 15:15:18 +00:00
Zachary Anderson 6cd8a79078 VM: Re-format to use at most one newline between functions
R=asiva@google.com

Review-Url: https://codereview.chromium.org/2974233002 .
2017-07-13 08:08:37 -07:00
Alexander Markov b04b95cfd2 Implement Dart VM option '--trace_sim_after' in DBC simulator
R=zra@google.com

Review-Url: https://codereview.chromium.org/2962353003 .
2017-06-30 15:29:10 -07:00
Todd Turnidge 7d6ba6bf1f Make rewind work on DBC
Closes #27878

R=johnmccutchan@google.com

Review-Url: https://codereview.chromium.org/2636443004 .
2017-03-08 13:15:54 -08:00
Vyacheslav Egorov 10588e7bec VM: [DBC] Remove sp_ field from the DBC simulator
This field was always equal to fp_.

R=zra@google.com

Review-Url: https://codereview.chromium.org/2570103003 .
2016-12-13 19:14:13 +01:00
John McCutchan f125cc7d0b Improve the casing of Stackmap and Stacktrace.
- [x] Stackmap -> StackMap
- [x] Stacktrace -> StackTrace

This makes us consistent with the Dart language (StackTrace) and
corrects the casing of StackMap.

BUG=
R=rmacnak@google.com

Review-Url: https://codereview.chromium.org/2572563004 .
2016-12-12 14:51:30 -08:00
Todd Turnidge 2d22e85cf0 Revert "Revert "JumpToFrame refactor"" + Fix
This reverts commit 1cdb3d2155.

I have also added in a fix for the lazy-deopt in exception handler case.

BUG=
R=johnmccutchan@google.com

Review URL: https://codereview.chromium.org/2506503002 .
2016-11-21 08:49:18 -08:00
Todd Turnidge 1cdb3d2155 Revert "JumpToFrame refactor"
This reverts commit 5ed13d3298.

BUG=

Review URL: https://codereview.chromium.org/2503623003 .
2016-11-14 15:57:01 -08:00
Todd Turnidge 5ed13d3298 JumpToFrame refactor
- Refactor the JumpToExceptionHandle code so that it is now built from
  two pieces: JumpToFrame and RunExceptionHandler.

- Refactor the Simulator::Longjmp() code so that it is no longer
  exception-specific.  Instead it uses the RunExceptionHandler stub.

This makes it so that the JumpToFrame stub and Simulator::JumpToFrame
have the same semantics.  This will make it easier to land the Rewind
changes I am working on.

There are some oddities for dbc.

BUG=
R=johnmccutchan@google.com

Review URL: https://codereview.chromium.org/2503653002 .
2016-11-14 14:38:21 -08:00
Zachary Anderson a1bcf051d8 clang-format runtime/vm
R=johnmccutchan@google.com

Review URL: https://codereview.chromium.org/2481873005 .
2016-11-08 13:54:47 -08:00
Zachary Anderson 103881d01c Make header include guards great again
i.e. #ifndef VM_WHATEVER -> #ifndef RUNTIME_VM_WHATEVER

This lets us remove a hack from the PRESUBMIT.py script that existed
for reasons that are no longer valid, and sets us up to add some
presubmit checks for the GN build.

R=asiva@google.com, rmacnak@google.com

Review URL: https://codereview.chromium.org/2450713004 .
2016-10-26 00:26:03 -07:00
Ryan Macnak e3420eab29 Revert "Pass new pool pointer to the JumpToException stub instead of reloading in through the frame's Code object."
This reverts commit 3413e052b3.

Review URL: https://codereview.chromium.org/2374173002 .
2016-09-28 13:00:00 -07:00
Ryan Macnak 3413e052b3 Pass new pool pointer to the JumpToException stub instead of reloading in through the frame's Code object.
This is prep work for lazy deopt without code patching.

R=fschneider@google.com

Review URL: https://codereview.chromium.org/2376843002 .
2016-09-28 10:15:09 -07:00
Ryan Macnak 59c52a5b64 Partially implement DBC profiler.
- Fix setting exit frame info to indicate when we are executing Dart code.
 - Copy local pc to memory on call and return to make it accessible to the profiler. Since the profiler cannot access the local pc, the profiler is not accurate at the instruction level.
 - Teach the profiler's stack walker that the stack grows up for Dart frames on DBC.

R=johnmccutchan@google.com

Review URL: https://codereview.chromium.org/2250823002 .
2016-08-19 16:48:34 -07:00
Ryan Macnak f6a320b973 DBC: Implement coverage by adding missing count increments to unoptimized static and instance calls.
These counts also factor into inlining decisions.

R=zra@google.com

Review URL: https://codereview.chromium.org/2187423002 .
2016-07-28 16:43:47 -07:00
Ryan Macnak 2e9c5f9a91 Reapply "VM: Add result cid information for recognized methods."
- _HashVMBase._index is nullable
 - _HashVMBase._data is a regular Array
 - DBC fixes from Zach.

R=fschneider@google.com

Review URL: https://codereview.chromium.org/2033553002 .
2016-06-02 09:59:48 -07:00
Vyacheslav Egorov 655bc90489 Enable optimizer pipeline for DBC.
Most of the infrastructure is fixed to work with DBC stack layout:

- register allocator allocates DBC registers with the limitation that we allocate only 20 registers and bail out if anything needs spilling (there is no use implementing spilling on DBC because registers are memory locations themselves). We should be able to bump number of CPU registers on DBC up to 256 but this requires major surgery in some parts - so I postponed this;
- lazy deoptimization is implemented, eager deoptimization is not - because we don't emit any code that actually requires it. it's a minor change to support it once we have a target;
- stack scanning respects stack maps built by registers allocator;

We bailout from all unsupported instructions.

R=zra@google.com

Review URL: https://codereview.chromium.org/1992963002 .
2016-05-24 14:35:50 +02:00
Zachary Anderson 3d7202bad3 DBC: Adds simdbc64 target, adds arm64 arithmetic overflow logic
There were differences between linux gcc, android gcc,
MacOS clang, and 32 vs. 64-bit w.r.t the __builtin
functions for detecting arithmetic overflow. I couldn't get
them all working at the same time. Instead, I removed them,
and changed to always use the inline assembly. This works
in all the configurations above.

This change also adds a simdbc64 target for building simdbc
for 64-bit, and sets up the android targets. simdbc targets
arm, and simdbc64 targets arm64. You can build them with:

$ ./tools/build.py -m release -a simdbc{64} --os=android runtime

R=iposva@google.com

Review URL: https://codereview.chromium.org/1904153003 .
2016-04-25 12:48:34 -07:00
Vyacheslav Egorov ee0f608ce4 Dart Byte Code interpreter.
This version is Clang/GCC only and does not support Windows because it uses computed goto's.

Only unoptimized mode is supported.

Architecture is described in constants_dbc.h and stack_frame_dbc.h.

R=fschneider@google.com, zra@google.com

Review URL: https://codereview.chromium.org/1858283002 .
2016-04-18 23:02:01 +02:00