Implements a backend targeting RV32GC and RV64GC, based on Linux standardizing around GC. The assembler is written to make it easy to disable usage of C, but because the sizes of some instruction sequences are compile-time constants, an additional build configuration would need to be defined to make use of it.
The assembler and disassembler cover every RV32/64GC instruction. The simulator covers all instructions except accessing CSRs and the floating point state accessible through such, include accrued exceptions and dynamic rounding mode.
Quirks:
- RISC-V is a compare-and-branch architecture, but some existing "architecture-independent" parts of the Dart compiler assume a condition code architecture. To avoid rewriting these parts, we use a peephole in the assembler to map to compare-and-branch. See Assembler::BranchIf. Luckily nothing depended on taking multiple branches on the same condition code set.
- There are no hardware overflow checks, so we must use Hacker's Delight style software checks. Often these are very cheap: if the sign of one operand is known, a single branch is needed.
- The ranges of RISC-V branches and jumps are such that we use 3 levels of generation for forward branches, instead of the 2 levels of near and far branches used on ARM[64]. Nearly all code is handled by the first two levels with 20-bits of range, with enormous regex matchers triggering the third level that uses aupic+jalr to get 32-bits of range.
- For PC-relative calls in AOT, we always generate auipc+jalr pairs with 32-bits of range, so we never generate trampolines.
- Only a subset of registers are available in some compressed instructions, so we assign the most popular uses to these registers. In particular, THR, TMP[2], CODE and PP. This has the effect of assigning CODE and PP to volatile registers in the C calling convention, whereas they are assigned preserved registers on the other architectures. As on ARM64, PP is untagged; this is so short indices can be accessed with a compressed instruction.
- There are no push or pop instructions, so combining pushes and pops is preferred so we can update SP once.
- The C calling convention has a strongly aligned stack, but unlike on ARM64 we don't need to use an alternate stack pointer. The author ensured language was added to the RISC-V psABI making the OS responsible for realigning the stack pointer for signal handlers, allowing Dart to leave the stack pointer misaligned from the C calling convention's point of view until a foreign call.
- We don't bother with the link register tracking done on ARM[64]. Instead we make use of an alternate link register to avoid inline spilling in the write barrier.
Unimplemented:
- non-trivial FFI cases
- Compressed pointers - No intention to implement.
- Unboxed SIMD - We might make use of the V extension registers when the V extension is ratified.
- BigInt intrinsics
TEST=existing tests for IL level, new tests for assembler/disassembler/simulator
Bug: https://github.com/dart-lang/sdk/issues/38587
Bug: https://github.com/dart-lang/sdk/issues/48164
Change-Id: I991d1df4be5bf55efec5371b767b332d37dfa3e0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/217289
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Reviewed-by: Slava Egorov <vegorov@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
There was a missing check for whether a backwards call is out-of-range.
The CL also removes leftover code related to [max_offset_into_target_] -
which wasn't used. The intended use is already over approximated via
[max_instructions_size_].
The CL makes the pc-relative call/tail-call distances pluggable so a
newly added test can modify them for testing in-range/out-of-range
forward/backwards calls.
TEST=vm/cc/CodeRelocator_*
Fixes https://github.com/flutter/flutter/issues/80043
Change-Id: Id4bdb7176108b61235dafb7ffc125da4a2bf07fa
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/195682
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This is a reland of b6dc4dad4d
TEST=ci
Original change's description:
> [vm/aot] Avoid using most Code objects in stack traces with --dwarf-stack-traces
>
> The following changes are done in preparation for the removal of Code
> objects in AOT with --dwarf-stack-traces:
>
> * Stack trace objects are extended to hold uword PCs (which may not
> fit into Smi range).
>
> * Scanning stack frames in GC (StackFrame::VisitObjectPointers)
> now avoids using Code objects.
> In order to find CompressedStackMaps it now calls
> ReversePc::FindCompressedStackMaps.
>
> * Singleton Code object (StubCode::UnknownDartCode()) is prepared as
> a replacement for Code objects in stack traces. It has
> PayloadStart() == 0 and Size() == kUwordMax so it includes
> arbitrary PCs.
>
> * In --dwarf-stack-traces mode, most Code objects obtained from stack
> frames are replaced with StubCode::UnknownDartCode().
> This simulates future behavior of ReversePc::Lookup when Code objects
> will be removed.
>
> Issue: https://github.com/dart-lang/sdk/issues/44852
> Change-Id: I7cec7b8b9396c9cfeca3c256a412ba4e82a7e0c4
> TEST=ci
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/182720
> Commit-Queue: Alexander Markov <alexmarkov@google.com>
> Reviewed-by: Tess Strickland <sstrickl@google.com>
> Reviewed-by: Martin Kustermann <kustermann@google.com>
Change-Id: Ia2fc4672a085cd963b7fc103369851df3603590c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/186202
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Tess Strickland <sstrickl@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This reverts commit b6dc4dad4d.
Reason for revert: broke package:vm_snapshot_analysis in Flutter
(https://github.com/flutter/flutter/issues/76313).
Original change's description:
> [vm/aot] Avoid using most Code objects in stack traces with --dwarf-stack-traces
>
> The following changes are done in preparation for the removal of Code
> objects in AOT with --dwarf-stack-traces:
>
> * Stack trace objects are extended to hold uword PCs (which may not
> fit into Smi range).
>
> * Scanning stack frames in GC (StackFrame::VisitObjectPointers)
> now avoids using Code objects.
> In order to find CompressedStackMaps it now calls
> ReversePc::FindCompressedStackMaps.
>
> * Singleton Code object (StubCode::UnknownDartCode()) is prepared as
> a replacement for Code objects in stack traces. It has
> PayloadStart() == 0 and Size() == kUwordMax so it includes
> arbitrary PCs.
>
> * In --dwarf-stack-traces mode, most Code objects obtained from stack
> frames are replaced with StubCode::UnknownDartCode().
> This simulates future behavior of ReversePc::Lookup when Code objects
> will be removed.
>
> Issue: https://github.com/dart-lang/sdk/issues/44852
> Change-Id: I7cec7b8b9396c9cfeca3c256a412ba4e82a7e0c4
> TEST=ci
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/182720
> Commit-Queue: Alexander Markov <alexmarkov@google.com>
> Reviewed-by: Tess Strickland <sstrickl@google.com>
> Reviewed-by: Martin Kustermann <kustermann@google.com>
Issue: https://github.com/dart-lang/sdk/issues/44852
Change-Id: I6f66171eecf1133363a7ce56193e782e43a20baf
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/185488
Reviewed-by: Zach Anderson <zra@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
The following changes are done in preparation for the removal of Code
objects in AOT with --dwarf-stack-traces:
* Stack trace objects are extended to hold uword PCs (which may not
fit into Smi range).
* Scanning stack frames in GC (StackFrame::VisitObjectPointers)
now avoids using Code objects.
In order to find CompressedStackMaps it now calls
ReversePc::FindCompressedStackMaps.
* Singleton Code object (StubCode::UnknownDartCode()) is prepared as
a replacement for Code objects in stack traces. It has
PayloadStart() == 0 and Size() == kUwordMax so it includes
arbitrary PCs.
* In --dwarf-stack-traces mode, most Code objects obtained from stack
frames are replaced with StubCode::UnknownDartCode().
This simulates future behavior of ReversePc::Lookup when Code objects
will be removed.
Issue: https://github.com/dart-lang/sdk/issues/44852
Change-Id: I7cec7b8b9396c9cfeca3c256a412ba4e82a7e0c4
TEST=ci
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/182720
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Tess Strickland <sstrickl@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This works around bugs in UndefinedBehaviorSanitizer and Clang.
Bug: b/28638298
Change-Id: I6be595f9664516019d28017d24559583a1ae3a21
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/144354
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Patchset 1 contains just the usings I deleted (<100 lines). All the
other diffs in the rest of the CL are just fixing all the places that
were broken by this.
Bug: https://github.com/dart-lang/sdk/issues/36839
Change-Id: I3bb4fa62ab4363ded81fd7c2815b857f91886dd6
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/108502
Commit-Queue: Liam Appelbe <liama@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
dart-bytecode, arm64: +4.742% geomean
dart-bytecode-jit-unopt, arm64: +12.73% geomean
dart2js-compile, x64: +3.635% geomean
In the polymorphic and unlinked cases, call to a stub the does a linear scan against an ICData.
In the monomorphic case, call to a prologue of the expected target function that checks the expected receiver class. There is additional indirection in the JIT version compared to the AOT version to also tick a usage counter so the inliner can make good decisions.
In the megamorphic case, call to a stub that does a hash table lookup against a MegamorphicCache.
Megamorphic call sites face a loss of precision in usage counts. The call site count is not recorded and the usage counter of the target function is used as an approximation.
Monomorphic and megamorphic calls sites are reset to the polymorphic/unlinked state on hot reload.
Monomorphic and megamorphic calls sites do not check the stepping state, so they are reset to the polymorphic/unlinked state when stepping begins and disabled.
Back-edges now increment the usage counter in addition to checking it. This ensures function with loops containing monomorphic calls will eventually cross the optimization threshold.
Fixed backwards use of kMonomorphicEntryOffset and kPolymorphicEntryOffset.
Fixed C stack overflow when bouncing between the KBC interpreter and a simulator.
Bug: https://github.com/dart-lang/sdk/issues/26780
Bug: https://github.com/dart-lang/sdk/issues/36409
Bug: https://github.com/dart-lang/sdk/issues/36731
Change-Id: I78a49cccd962703a459288e71ce246ed845df474
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/102820
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This reverts commit fde6a5917e.
Reason for revert: This commit looks to be causing new test failures on dart2js. While there are some odd results on some builders that look like infra failures, some builders (for example, dart2js-minified-strong-linux-x64-d8) show the new failing tests clearly.
Original change's description:
> [vm, compiler] Unoptimized megamorphic calls.
>
> When an instance call in unoptimized code creates more than FLAG_max_polymorphic_checks cases, switch the call to use a MegamorphicCache instead of ICData. The prevents unbounded collection of type feedback, and gives improvements on microbenchmarks in the 3-8% range for unoptimized code.
>
> It also leads to a loss of target frequency information for the optimizer, leading to different ordering for range checks in polymorphic inlining. This leads to changes on megamorphic microbenchmarks from -31% to +60%, weighted toward the negative end.
>
> In practice the frequency information seems unimportant, as dart2js has 4.01% geomean improvement.
>
> This is a step toward direct monomorphic calls in unoptimized code, which will also make use of the patching and type feedback extraction added here.
>
> Bug: https://github.com/dart-lang/sdk/issues/26780
> Bug: https://github.com/dart-lang/sdk/issues/36409
> Bug: https://github.com/dart-lang/sdk/issues/36731
> Change-Id: I29f53f23b6794c5f5f0db8b8184788cee16fd9c5
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/99270
> Reviewed-by: Alexander Markov <alexmarkov@google.com>
TBR=rmacnak@google.com,alexmarkov@google.com,ajcbik@google.com
Change-Id: Icad46b93cdf8541a00563f49da6b4ac0a4df1ba1
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: https://github.com/dart-lang/sdk/issues/26780, https://github.com/dart-lang/sdk/issues/36409, https://github.com/dart-lang/sdk/issues/36731
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103440
Reviewed-by: Teagan Strickland <sstrickl@google.com>
Commit-Queue: Teagan Strickland <sstrickl@google.com>
When an instance call in unoptimized code creates more than FLAG_max_polymorphic_checks cases, switch the call to use a MegamorphicCache instead of ICData. The prevents unbounded collection of type feedback, and gives improvements on microbenchmarks in the 3-8% range for unoptimized code.
It also leads to a loss of target frequency information for the optimizer, leading to different ordering for range checks in polymorphic inlining. This leads to changes on megamorphic microbenchmarks from -31% to +60%, weighted toward the negative end.
In practice the frequency information seems unimportant, as dart2js has 4.01% geomean improvement.
This is a step toward direct monomorphic calls in unoptimized code, which will also make use of the patching and type feedback extraction added here.
Bug: https://github.com/dart-lang/sdk/issues/26780
Bug: https://github.com/dart-lang/sdk/issues/36409
Bug: https://github.com/dart-lang/sdk/issues/36731
Change-Id: I29f53f23b6794c5f5f0db8b8184788cee16fd9c5
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/99270
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This is the next step towards preventing compiler from directly peeking
into runtime and instead interact with runtime through a well defined
surface. The goal of the refactoring to locate all places where compiler
accesses some runtime information and partion those accesses into two
categories:
- creating objects in the host runtime (e.g. allocating strings, numbers, etc)
during compilation;
- accessing properties of the target runtime (e.g. offsets of fields) to
embed those into the generated code;
This change introduces dart::compiler and dart::compiler::target namespaces.
All code in the compiler will gradually be moved into dart::compiler namespace.
One of the motivations for this change is to be able to prevent access to
globally defined host constants like kWordSize by shadowing them in the
dart::compiler namespace.
The nested namespace dart::compiler::target hosts all information about
target runtime that compiler could access, e.g. compiler::target::kWordSize
defines word size of the target which will eventually be made different
from the host kWordSize (defined by dart::kWordSize).
The API for compiler to runtime interaction is placed into compiler_api.h.
Note that we still permit runtime to access compiler internals directly -
this is not going to be decoupled as part of this work.
Issue https://github.com/dart-lang/sdk/issues/31709
Change-Id: If4396d295879391becfa6c38d4802bbff81f5b20
Reviewed-on: https://dart-review.googlesource.com/c/90242
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This CL adds support to the ImageWriter to write opaque blocks of
trampoline bytes
The AOT code relocator is adapted to allow limited range calls and
inserts trampolines if need be. The algorithm tries to minimize the
number of trampolines added, which for small applications will be 0.
The unconditional pc-relative calls have limited range:
* on ARM (+/-32 MB)
* on ARM64 (+/-128 MB)
To avoid verbose code for doubly-linked list, this CL adds double_list.h
Issue https://github.com/dart-lang/sdk/issues/33274
Change-Id: I0354cf4b2dd58ed5de25d67fc818f0603a2ec501
Reviewed-on: https://dart-review.googlesource.com/c/89283
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
If the --use-bare-instructions flag is enabled we will:
* Make call sites load the target directly from the pool (instead of
the code object) - this saves one instruction (and an indirect load)
* Ensure the object pool will have direct entry addresses by:
- Letting the clustered snapshot reader change any StubCode::UnlinkedCall()
in the object pool by it's monomorphic entry
- Change the code patcher to patch SwitchableCalls by writing the
monomorphic entry into the pool (instead of the code object)
Issue https://github.com/dart-lang/sdk/issues/33274
Change-Id: I4e41fc8e4461bde477cc559a6a4fccaaf3a350b5
Reviewed-on: https://dart-review.googlesource.com/c/86160
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Pool offsets might be not encodable as immediates. However
InstructionPattern::DecodeLoadWordFromPool can decode anything that
LoadWordFromPool can emit.
Fixes https://github.com/flutter/flutter/issues/18332
Change-Id: I8873f523bb65844e7256c0c3030edea4cc591830
Reviewed-on: https://dart-review.googlesource.com/59640
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
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 .
When throwing to a frame scheduled for lazy deopt, update the continuation pc for that frame to be the catch handler.
Weaken new assert that the deopt pc belongs to the frame's code as the deopt pc for the last eager deopt in a function lies outside the code, after the call to the deopt stub.
R=fschneider@google.com
Review URL: https://codereview.chromium.org/2392613002 .
The call sequence is very similar to a classic IC call, except the guarded class and the target are loaded indirectly from the constant pool instead of as immediates. In the monomorphic case, we call directly to the expected target with a class check in the callee. In the unlinked, polymorphic and megamorphic cases, we call a stub; these case are now call-through instead of call-and-return.
Every code, except stubs involved in switchable calls, includes the class check sequence at the beginning. So we now distinguish between a checked and an unchecked entry point. Generated code except the switchable call continues to use the unchecked entry point.
PC offsets are calculated relative to the beginning of the instruction stream, rather than either entry point.
BUG=
R=fschneider@google.com
Review URL: https://codereview.chromium.org/2226893002 .
Insert enough padding to ensure enough space for when lazy deoptimization occurs
from the last deferred code object.
Also, speed up ia32 disassembler to avoid tests timing out.
BUG=
TBR=rmacnak@google.com,
Review URL: https://codereview.chromium.org//1355953002 .
Instead of calling code object directly, call indirectly and
pass the code object in a register. The object pool is then loaded from
the code object. This is another preparation step for making generated code
relocatable.
All non-ia32 platforms:
No entry patching.
ARM:
PC marker (now code object) moves to the same place as on x64 (below saved PP, above saved FP).
R9 is now used as PP, R10 as CODE_REG.
BUG=
R=koda@google.com, rmacnak@google.com
Committed: 1d343e5a7b
Review URL: https://codereview.chromium.org//1192103004 .
Instead of calling code object directly, call indirectly and
pass the code object in a register. The object pool is then loaded from
the code object. This is another preparation step for making generated code
relocatable.
All non-ia32 platforms:
No entry patching.
ARM:
PC marker (now code object) moves to the same place as on x64 (below saved PP, above saved FP).
R9 is now used as PP, R10 as CODE_REG.
BUG=
R=rmacnak@google.com
Review URL: https://codereview.chromium.org//1192103004 .
The first invocation of a native functions goes into LinkNativeCall which
determines the final entry point and patches the object pool entry.
When running precompiled code, this makes deserializing the object
pool entries for native functions easy, they all initially point to
a single entry (LinkNativeCall).
BUG=
R=rmacnak@google.com
Review URL: https://codereview.chromium.org//1294113004 .
This adds meta-information to object pool entries to allow storing
untagged immediates or code addresses (ExternalLabel) directly.
This eliminates the need to generate extra code to preserve the LSB
when storing immediates as smis (x64, arm64).
BUG=
Review URL: https://codereview.chromium.org//1175523002.
Profiler improvements:
- Track Functions in profile and build Function based trie
- Associate code objects with functions
- Created cpu_profile.dart library
- Major speed improvements for disassembly view
- Fix truncation of disassembly comments
- Ability to get code object ticks from disassembly view
- Inlining mini-map in disassembly view.
- Remove a bunch of unused data from profile service response
- In some cases a caller PC that is better than the PC marker is inserted into the stack trace
- Inlined functions are expanded
- Ability to clear profile
- New flag '--keep_code' which keeps deoptimized code around for use by the profiler.
General fixes:
- Fix caching in service library
- Remove pubspec.yaml before running pub get
R=asiva@google.com, rmacnak@google.com
Review URL: https://codereview.chromium.org//928833003
git-svn-id: https://dart.googlecode.com/svn/branches/bleeding_edge/dart@44067 260f80e4-7a28-3924-810f-c04153c831b5
