SpiderMonkey

Versions

TraceMonkey

TraceMonkey^[11] was the first JIT compiler written for the JavaScript language. Initially introduced as an option in a beta release and introduced in Brendan Eich's blog on August 23, 2008,^[12] the compiler became part of the mainline release as part of SpiderMonkey in Firefox 3.5, providing "performance improvements ranging between 20 and 40 times faster" than the baseline interpreter in Firefox 3.^[13]

Instead of compiling whole functions, TraceMonkey was a tracing JIT, which operates by recording control flow and data types during interpreter execution. This data then informed the construction of trace trees, highly specialized paths of native code.

Improvements to JägerMonkey eventually made TraceMonkey obsolete, especially with the development of the SpiderMonkey type inference engine. TraceMonkey is absent from SpiderMonkey from Firefox 11 onward.^[14]

JägerMonkey

JägerMonkey, internally named MethodJIT, was a whole-method JIT compiler designed to improve performance in cases where TraceMonkey could not generate stable native code.^[15][16] It was first released in Firefox 4 and eventually entirely supplanted TraceMonkey. It has itself been replaced by IonMonkey.

JägerMonkey operated very differently from other compilers in its class: while typical compilers worked by constructing and optimizing a control-flow graph representing the function, JägerMonkey instead operated by iterating linearly forward through SpiderMonkey bytecode, the internal function representation. Although this prohibits optimizations that require instruction reordering, JägerMonkey compiling has the advantage of being very fast, which is useful for JavaScript since recompiling due to changing variable types is frequent.

Mozilla implemented a number of critical optimizations in JägerMonkey, most importantly polymorphic inline caches and type inference.^[17]

The difference between TraceMonkey and JägerMonkey JIT techniques and the need for both was explained in a hacks.mozilla.org article. A more in-depth explanation of the technical details was provided by Chris Leary, one of SpiderMonkey's developers, in a blog post Archived 9 December 2012 at archive.today. More technical information can be found in other developer's blogs: dvander, dmandelin.

IonMonkey

IonMonkey was a JavaScript JIT compiler of Mozilla, which was aimed to enable many new optimizations that were impossible with the prior JägerMonkey architecture.^[18]

IonMonkey was a more traditional compiler: it translated SpiderMonkey bytecode into a control-flow graph, using static single assignment form (SSA) for the intermediate representation. This architecture enabled well-known optimizations from other programming languages to be used for JavaScript, including type specialization, function inlining, linear-scan register allocation, dead code elimination, and loop-invariant code motion.^[19]

The compiler can emit fast native code translations of JavaScript functions on the ARM, x86, and x86-64 platforms. It has been the default engine since Firefox 18.^[20]

OdinMonkey

OdinMonkey is the name of Mozilla's new optimization module for asm.js, an easily compilable subset of JavaScript. OdinMonkey itself is not a JIT compiler, it uses the current JIT compiler. It's included with Firefox from release 22.

WarpMonkey

The WarpMonkey JIT replaces the former IonMonkey engine from version 83.^[21] It is able to inline other scripts and specialize code based on the data and arguments being processed. It translates the bytecode and Inline Cache data into a Mid-level Intermediate Representation (Ion MIR) representation. This graph is transformed and optimized before being lowered to a Low-level Intermediate Representation (Ion LIR). This LIR performs register allocation and then generates native machine code in a process called Code Generation. The optimizations here assume that a script continues to see data similar what has been seen before. The Baseline JITs are essential to success here because they generate ICs that match observed data. If after a script is compiled with Warp, it encounters data that it is not prepared to handle it performs a bailout. The bailout mechanism reconstructs the native machine stack frame to match the layout used by the Baseline Interpreter and then branches to that interpreter as though we were running it all along. Building this stack frame may use special side-table saved by Warp to reconstruct values that are not otherwise available.^[22]