RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-06-22 22:22:10 +00:00
Code Issues Activity Actions
58669 commits 1 branch 0 tags 230 MiB
Commit graph

5076 commits

Author SHA1 Message Date
Andreas Kling
c2aad0f573 LibJS/Bytecode: Move PutByValue impl to CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
9c93d100d1 LibJS/JIT: Compile the NewFunction bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
9f61cda27e LibJS/Bytecode: Move NewFunction impl into CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
a913ac5799 LibJS/JIT: Compile the NewArray bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
6a16783c66 LibJS/JIT: Compile the SetVariable bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
393d90abe1 LibJS/Bytecode: Move SetVariable implementation to CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
e946440ed3 LibJS/JIT: Compile the TypeofVariable bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
d368dc5d25 LibJS/Bytecode: Move TypeofVariable impl to CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
c65aecd878 LibJS/JIT: Compile all the unary bytecode instructions 2023-10-27 19:07:22 +02:00
Andreas Kling
640455b1d2 LibJS/JIT: Compile the Call bytecode instruction 
I've left a FIXME about dealing with some throwsy cases.
 2023-10-27 19:07:22 +02:00
Andreas Kling
7fc35fde09 LibJS/Bytecode: Move throw_if_needed_for_call to CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
b56ecc7e34 LibJS/Bytecode: Move perform_call helper to CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
d866780235 LibJS/JIT: Compile the NewObject bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
580249d650 LibJS/JIT: Compile the PutById bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
10bf25999c LibJS/JIT: Move VM pointers from R8,R9,R10 to R13,R14,R15 
This way they don't clash with the SysV ABI function argument registers.
 2023-10-27 19:07:22 +02:00
Andreas Kling
decc221109 LibJS/JIT: Save and restore callee-saved registers in jitted code 2023-10-27 19:07:22 +02:00
Andreas Kling
3974ce2069 LibJS/JIT: Compile the GetGlobal bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
12898f5aef LibJS/JIT: Compile the Decrement bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
966b6f78a6 LibJS/JIT: Compile the GetByValue bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
e8190105db LibJS/Bytecode: Move GetByValue implementation to CommonImplementations 2023-10-27 19:07:22 +02:00
Andreas Kling
1c0efbec6b LibJS/JIT: Compile the ResolveThisBinding bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
b2602a4bae LibJS/JIT: Compile the ToNumeric bytecode instruction 2023-10-27 19:07:22 +02:00
Andreas Kling
0f735b3502 LibJS/JIT: Log both success and failure from the JIT compiler 
These logs will eventually go away, once the JIT compiler can always
compile everything. :^)
 2023-10-27 19:07:22 +02:00
Andreas Kling
310bcd4717 LibJS/JIT: Don't keep trying to JIT unsupported bytecode executables 
We now only try jitting each Bytecode::Executable once, and then cache
the resulting NativeExecutable.
 2023-10-27 19:07:22 +02:00
Andreas Kling
6a6ef6670c LibJS/JIT: Support the GetById bytecode op 
We can now do basic property (get) access in jitted code! :^)
 2023-10-27 19:07:22 +02:00
Andreas Kling
8905682a16 LibJS/Bytecode: Begin moving shareable (JIT+Interpreter) stuff somewhere 
There are a lot of native C++ functions that will be used by both the
bytecode interpreter and jitted code. Let's put them in their own file
instead of having them in Interpreter.cpp.
 2023-10-27 19:07:22 +02:00
Andreas Kling
b923ca392d LibJS/JIT: Support all the binary bytecode ops :^) 
(And use the X macro to avoid repeating ourselves!)
 2023-10-27 19:07:22 +02:00
Andreas Kling
c2fe7af095 LibJS/JIT: Support the NewString bytecode op 
This necessitated making the JIT::Compiler aware of the current
Bytecode::Executable, since that's where all the string literals are
held, but that seems like a good thing.
 2023-10-27 19:07:22 +02:00
Andreas Kling
efe58ebf2f LibJS/JIT: Support the Return bytecode op 2023-10-27 19:07:22 +02:00
Andreas Kling
45be2a8f72 LibJS/JIT: Add support for Add/Sub/Mul/Div bytecode ops 2023-10-27 19:07:22 +02:00
Geo25rey
891b071654 LibJS/JIT: Add support for "throw" keyword 2023-10-27 19:07:22 +02:00
Andreas Kling
ed0d2bce83 LibJS/JIT: Handle exceptions in LessThan :^) 2023-10-27 19:07:22 +02:00
Andreas Kling
9dd5be0186 LibJS/JIT: Compile the EnterUnwindContext and LeaveUnwindContext ops 
These push a "valid" unwind context on the stack and check_exception()
now knows how to jump to the (catch) handler if present.

(finally) finalizers will require some more work, but with this change,
we now have basic support for try...catch. :^)
 2023-10-27 19:07:22 +02:00
Andreas Kling
9d35016284 LibJS/JIT: Add Assembler::jump(Operand) and verify_not_reached() 
Yet another jump() overload, and also a way to crash if we end up
somewhere we shouldn't be while in jitted code.
 2023-10-27 19:07:22 +02:00
Andreas Kling
e3560c2545 LibJS/JIT: Propagate exceptions in the simplest case :^) 
We now establish a stack of "unwind contexts" similar to what the
bytecode interpreter does, but here, it's a stack of structs with
addresses to the catch and finally blocks.

Unwind contexts also have a "valid" flag, and the root unwind context
(always present, pushed on JIT code entry) has valid=false, which we
interpret in check_exception() as "return and let our caller deal with
the exception".

Anything in Compiler that may generate an exception should now also
call check_exception() ASAP to emit the code for handling this.
 2023-10-27 19:07:22 +02:00
Andreas Kling
3523f9f722 LibJS/JIT: Add patchable absolute references to basic blocks 2023-10-27 19:07:22 +02:00
Andreas Kling
2e6c02f088 LibJS/JIT: Support Imm32 operands in Assembler::push() 2023-10-27 19:07:22 +02:00
Andreas Kling
71e41418f6 LibJS/JIT: Do "enter & leave" sequence in x86_64 machine code 
This ensures that the stack pointer is restored before we return from
the jitted code.
 2023-10-27 19:07:22 +02:00
Andreas Kling
814b07a9c2 LibJS/JIT: Add Assembler::jump_if_equal() 
And also factor out cmp() so we don't have to repeat it.
 2023-10-27 19:07:22 +02:00
Andreas Kling
a7bad26b63 LibJS/JIT: Add missing unistd.h include (fixes Clang build) 2023-10-27 19:07:22 +02:00
Andreas Kling
c21978ead6 LibJS/JIT: Remove unused functions from Assembler 2023-10-27 19:07:22 +02:00
Andreas Kling
1e8c6d3b06 LibJS/JIT: Use mov() helper in Assembler::native_call() 2023-10-27 19:07:22 +02:00
Andreas Kling
e7ce3d271a LibJS/JIT: Add Assembler::add() and sub() helpers 2023-10-27 19:07:22 +02:00
Andreas Kling
10dcb4d53a LibJS/JIT: Add Assembler::push() and pop() helpers 2023-10-27 19:07:22 +02:00
Andreas Kling
acece9057e LibJS/JIT: Make Assembler::Reg represent X86 registers 
And move the generic register aliases to JIT::Compiler.
 2023-10-27 19:07:22 +02:00
Andreas Kling
f9041c7b31 LibJS/JIT: Fast path for boolean JS::Value in compile_to_boolean() 2023-10-27 19:07:22 +02:00
Andreas Kling
babdc0a25b LibJS: Work-in-progress JIT compiler :^) 2023-10-27 19:07:22 +02:00
Andreas Kling
2e23f00a2f LibJS/Bytecode: Move environment coordinate caches to Executable 
Moving them out of the respective instructions allows the bytecode
stream to be immutable.
 2023-10-27 07:26:37 +02:00
Jesús (gsus) Lapastora
2086b8df9c LibJS/Date: Ensure YearFromTime(t) holds invariant after approximation 
As of https://tc39.es/ecma262/#sec-yearfromtime, YearFromTime(t) should
return `y` such that `TimeFromYear(YearFromTime(t)) <= t`. This wasn't
held, since the approximation contained decimal digits that would nudge
the final value in the wrong direction.

Adapted from Kiesel:
6548a85743

Co-authored-by: Linus Groh <mail@linusgroh.de>
 2023-10-23 09:26:55 -04:00
Jelle Raaijmakers
c58193bafa LibJS: Support large number of decimals in Number.prototype.toFixed 
The spec asks us to perform some calculations that quickly exceed an
`u64`, but instead of jumping through hoops we can rely on our AK
implementation of floating point formatting to come up with the
correctly rounded result.

Note that most other JS engines seem to diverge from the spec as well
and fall back to a generic dtoa path.
 2023-10-18 19:39:30 -04:00