1668 commits

Author	SHA1	Message	Date
Timothy Flynn	6b4b7a54de	LibJS: Revert ArrayIterator and RegExpStringIterator to manual iterators ... This is a normative change in the ECMA-262 spec. See: de62e8d This did not actually seem to affect our implementation as we were not using generators here to begin with. So this patch is basically just adding spec comments.	2025-04-30 07:29:34 -04:00
Andreas Kling	7a600e60bc	LibJS: Mark catch parameter as initialized local when applicable ... Otherwise we'll emit a TDZ check for every `catch` parameter immediately after initializing it. This check would never fail, so it's redundant.	2025-04-30 09:38:40 +02:00
Andreas Kling	00f7a6f9e0	LibJS: Remove unused bytecode VM register reservation ... We were not actually using the "saved exception" register for anything, but we were clearing it on every function entry.	2025-04-30 09:38:40 +02:00
Andreas Kling	1f0c67cc12	Revert "LibJS: Add StackFrame to avoid indirection in VM register access" ... This reverts commit 36bb2824a6. Although this was faster on my M3 MacBook Pro, other Apple machines disagree, including our benchmark runner. So let's revert it.	2025-04-29 16:08:42 +02:00
Andreas Kling	36bb2824a6	LibJS: Add StackFrame to avoid indirection in VM register access ... This is a simple trick to generate better native code for access to registers, locals, and constants. Before this change, each access had to first dereference the member pointer in Interpreter, and then get to the values. Now we always have a pointer directly to the values on hand. Here's how it looks: class StackFrame { public: Value get(Operand) const; void set(Operand, Value); private: Value m_values[]; }; And we just place one of these as a window on top of the execution context's array of values (registers, locals, and constants).	2025-04-29 14:23:03 +02:00
Timothy Flynn	4d70f6ce1c	LibJS: Ensure iterator parameter validation closes underlying iterator ... This is a normative change in the ECMA-262 spec. See: 9552f29 f2bad00	2025-04-29 07:33:08 -04:00
Timothy Flynn	908349f8fe	LibJS: Add a TRY_OR_CLOSE_ITERATOR macro for IfAbruptCloseIterator ... Behaves just like TRY_OR_REJECT for promises.	2025-04-29 07:33:08 -04:00
Timothy Flynn	568524f8ba	LibJS: Close sync iterator when async wrapper yields rejection ... This is a normative change in the ECMA-262 spec. See: ff129b1	2025-04-29 07:33:08 -04:00
Timothy Flynn	15faaeb2bb	LibJS: Remove [[VarNames]] from GlobalEnvironment ... This is a normative change in the ECMA-262 spec. See: ed75310	2025-04-29 07:33:08 -04:00
Timothy Flynn	9c85a16aeb	LibJS: Standardize spec prose for Math.sumPrecise	2025-04-29 07:33:08 -04:00
Timothy Flynn	9674210ef8	LibJS: Update spec steps / links for the Float16Array proposal ... This proposal has reached stage 4 and been merged into the main ECMA-262 spec. See: d430ace	2025-04-29 07:33:08 -04:00
Timothy Flynn	adf6024805	LibJS: Update spec steps / links for the RegExp.escape proposal ... This proposal has reached stage 4 and been merged into the main ECMA-262 spec. See: e2da759	2025-04-29 07:33:08 -04:00
Timothy Flynn	2401764697	LibJS: Update spec steps / links for the JSON modules proposal ... This proposal has reached stage 4 and been merged into the main ECMA-262 spec. See: 3feb1ba	2025-04-29 07:33:08 -04:00
Timothy Flynn	3867a192a1	LibJS: Update spec steps / links for the import-assertions proposal ... This proposal has reached stage 4 and been merged into the main ECMA-262 spec. See: 4e3450e	2025-04-29 07:33:08 -04:00
Andreas Kling	6061da3382	LibJS: Overload Instruction::length() on variable-length instructions ... This allows the DISPATCH_NEXT() macro to do the math to find the next instruction boundary point without performing a function call.	2025-04-29 02:09:35 +02:00
Andreas Kling	a2b7e04da3	LibJS: Defer looking up the realm in ordinary_call_evaluate_body() ... We don't actually need the realm for normal function calls, so we can avoid looking it up on the EC stack in that case.	2025-04-29 02:09:35 +02:00
Andreas Kling	0f1be720bb	LibJS: Mark exception path [[unlikely]] in ordinary_call_evaluate_body()	2025-04-29 02:09:35 +02:00
Andreas Kling	58925887ce	LibJS: Inline VM::bytecode_interpreter()	2025-04-29 02:09:35 +02:00
Andreas Kling	35275651e3	LibJS: Skip stack overflow check in ESFO::prepare_for_ordinary_call() ... We already do a stack overflow check when entering run_bytecode(), which is the first thing that happens when we actually invoke the ESFO executable.	2025-04-29 02:09:35 +02:00
Andreas Kling	95ba74d934	LibJS: Remove redundant VERIFY in run_executable() ... Getting the running_execution_context() already verifies that the execution context stack is non-empty, we don't need to do it separately here as well.	2025-04-29 02:09:35 +02:00
Andreas Kling	6de1a0aeaf	LibJS: Don't cache a pointer to accumulator register in run_bytecode() ... The old accumulator register is really only used to pass the end completion to the caller of run_bytecode() nowadays. As such, we don't need to cache a pointer to it for fast access. One less thing to do on run_bytecode() entry.	2025-04-29 02:09:35 +02:00
Andreas Kling	942ce2162d	LibJS: Mark stack overflow path in run_bytecode() [[unlikely]]	2025-04-29 02:09:35 +02:00
Andreas Kling	4d17707b26	LibJS: Store bytecode VM program counter in ExecutionContext ... This way it's always automatically correct, and we don't have to manually flush it in push_execution_context(). ~7% speedup on the MicroBench/call* tests :^)	2025-04-28 21:12:48 +02:00
Andreas Kling	233097c250	LibJS: Inline part of VM::run_queued_promise_jobs() ... Most of the time there are no queued promise jobs to run after exiting a stack frame. By moving the check inline, leaving a function call gets a measurable speedup in the common case.	2025-04-28 10:39:42 -04:00
Andreas Kling	b4554c01db	LibJS: Mark ESFO path for [[Call]] on a class constructor [[unlikely]] ... This is an exception path that's not supposed to be called normally, so let's mark it unlikely.	2025-04-28 10:39:42 -04:00
Andreas Kling	6ec4d0f5ba	LibJS: Mark stack overflow exception code path as [[unlikely]] ... This is supposed to be exceedingly rare, so a great candidate for [[unlikely]] annotation.	2025-04-28 10:39:42 -04:00
Andreas Kling	074ca5d5b4	LibJS: Make ESFO::ordinary_call_evaluate_body() return TCO<Value> ... This matches what the caller wants to return and allows us to simplify a bunch of logic around returning a value or throwing.	2025-04-28 12:44:49 +02:00
Andreas Kling	670e439e1e	LibJS: Put FLATTEN on ECMAScriptFunctionObject.[[Call]] ... This makes function calls ~5% faster in micro-benchmarks on my MBP. Basically free money on the table. Let's take it!	2025-04-28 12:44:49 +02:00
Andreas Kling	d0d87d3aed	LibJS: Demote some overly paranoid VERIFY()s in ESFO [[Call]] flow	2025-04-28 12:44:49 +02:00
Andreas Kling	403ae86fd9	LibJS: Pass VM& to ECMAScriptFunctionObject [[Call]] helpers ... This avoids fetching the VM from the Cell::private_data() repeatedly.	2025-04-28 12:44:49 +02:00
devgianlu	5f1a30197c	LibCrypto: Remove the concept of invalid big integers ... This concept is rarely used in codebase and very much error-prone if you forget to check it. Instead, make it so that operations that would produce invalid integers return an error instead.	2025-04-28 12:05:26 +02:00
devgianlu	dd0cced92f	LibJS: Prevent huge memory allocations for bigint left shift	2025-04-28 12:05:26 +02:00
devgianlu	a019efb24b	LibCrypto+LibJS: Remove {Signed,Unsigned}BigInteger to_base_deprecated ... Use `to_base` instead.	2025-04-28 12:05:26 +02:00
Andreas Kling	a05be67e4a	LibJS: Let invokers (callers) of [[Call]] allocate ExecutionContext ... Instead of letting every [[Call]] implementation allocate an ExecutionContext, we now make that a responsibility of the caller. The main point of this exercise is to allow the Call instruction to write function arguments directly into the callee ExecutionContext instead of copying them later. This makes function calls significantly faster: - 10-20% faster on micro-benchmarks (depending on argument count) - 4% speedup on Kraken - 2% speedup on Octane - 5% speedup on JetStream	2025-04-28 01:23:56 +02:00
Andreas Kling	93788f8057	LibJS: Add parentheses to ALLOCATE_EXECUTION_CONTEXT_ON_NATIVE_STACK() ... Just to stop clang-tidy from complaining about it every time.	2025-04-28 01:23:56 +02:00
Andreas Kling	373307db5b	LibJS: Mark arguments local as initialized after creating it ... This avoids emitting an unnecessary TDZ check for the first time we access the arguments object.	2025-04-28 01:23:56 +02:00
Shannon Booth	9e44d86915	LibJS: Remove value_or from JS::Value ... This is no longer used after 3cf5053.	2025-04-27 11:14:54 -04:00
Timothy Flynn	317cf65eb0	LibJS: Avoid naming conflict between Object's and Error's is_error ... Object defines an is_error virtual method to be overridden by Error for fast-is. This is the same name as the Error.isError constructor method. Rename the former to avoid conflicts, as GCC 15 just started warning on this.	2025-04-26 09:04:45 -06:00
Aliaksandr Kalenik	2d732b2251	LibJS: Skip allocating locals for arguments that allowed to be local ... This allows us to get rid of instructions that move arguments to locals and allocate smaller JS::Value vector in ExecutionContext by reusing slots that were already allocated for arguments. With this change for following function: ```js function f(x, y) { return x + y; } ``` we now produce following bytecode: ``` [ 0] 0: Add dst:reg6, lhs:arg0, rhs:arg1 [ 10] Return value:reg6 ``` instead of: ``` [ 0] 0: GetArgument 0, dst:x~1 [ 10] GetArgument 1, dst:y~0 [ 20] Add dst:reg6, lhs:x~1, rhs:y~0 [ 30] Return value:reg6 ```	2025-04-26 11:02:29 +02:00
Aliaksandr Kalenik	3f04d18ef7	LibJS: Add new operand type for function arguments ... This allows us to directly access passed arguments instead of copying them to register/local first using GetArgument instruction.	2025-04-26 11:02:29 +02:00
Aliaksandr Kalenik	81a3bfd492	LibJS: Allow using locals if arguments is used in strict mode ... Previously we blocked using locals for function arguments whenever `arguments` was mentioned in function body, however, this is not necessary in strict mode, where mutations to the arguments object are not reflected in the function arguments and vice versa.	2025-04-25 21:08:24 +02:00
Shannon Booth	5290ebfe19	LibJS: Switch Agent [[CanBlock]] slot to a enum member ... It turns out it was a mistake to make this a virtual since ServiceWorkerAgents are effectively the exact same as DedicatedWorkerAgents and SharedWorkerAgents just with [[CanBlock]] set to false.	2025-04-25 14:07:51 +02:00
Shannon Booth	7dd7e5b438	LibJS+LibWeb: Defer initialization of the Agent after VM constructor ... This helps unwind a niggly depedency where the VM owns and constructs the Heap and the Agent. But the agent wants to have customized construction that depends on the heap. Solve this by defering the initialization of the Agent to after we have constructed the VM and the heap.	2025-04-25 14:07:51 +02:00
Shannon Booth	8263a9863f	LibJS+LibWeb: Do not return error from VM::create ... This never returns an error to propogate, also allowing ErrorOr to be removed from creating the main thread VM.	2025-04-25 14:07:51 +02:00
aplefull	223c9c91e6	LibJS: Implement rawJSON and isRawJSON functions	2025-04-24 09:33:49 -04:00
Andrew Kaster	9bae24cc4a	LibJS: Add and use ValidateNonRevokedProxy AO ... This refactor is from two editorial changes to the spec from a while back. 44d1cae2b2 21ffeee869	2025-04-24 10:37:39 +02:00
Aliaksandr Kalenik	e48645c83f	LibJS: Cache arguments span in ExecutionContext ... Allows us to avoid doing math in ExecutionContext::argument()	2025-04-24 10:30:52 +02:00
Aliaksandr Kalenik	ff751173ac	LibJS: Delete unused m_arguments member in Interpreter	2025-04-24 10:30:52 +02:00
Aliaksandr Kalenik	a329868c1b	LibJS: Allocate ExecutionContext memory using alloca() when possible ... This should be faster than heap allocation. However, heap allocation is still necessary in some cases, such as with generators and async functions.	2025-04-24 10:30:52 +02:00
Aliaksandr Kalenik	5a92929282	LibJS: Put vector of regs+consts+locals+args in tail of ExecutionContext ... By doing that we avoid doing separate allocation for each such vector, which was really expensive on js heavy websites. For example this change helps to get EC allocation down from ~17% to ~2% on Google Maps. This comes at cost of adding extra complexity to custom execution context allocator, because EC no longer has fixed size and we need to maintain a list of buckets.	2025-04-24 10:30:52 +02:00