Re: [JS-internals] Performance impact of Debugger

2013-10-24 Thread Jim Blandy
I lose, evidently. Anyway, it's attached to the bug: https://bugzilla.mozilla.org/show_bug.cgi?id=914429 ___ dev-tech-js-engine-internals mailing list dev-tech-js-engine-internals@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-tech-js-engine-

Re: [JS-internals] Performance impact of Debugger

2013-10-24 Thread Jim Blandy
That doesn't seem to have worked. Attached: ___ dev-tech-js-engine-internals mailing list dev-tech-js-engine-internals@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-tech-js-engine-internals

Re: [JS-internals] Performance impact of Debugger

2013-10-24 Thread Jim Blandy
OMG I made a plot: plot of performance impact of Debugger with various JITs I really don't understand why we have that much impact even against the interpreter. I would like to. ___ dev-tech-js-engine-internals mailing list dev-tech-js-engine-interna

Re: [JS-internals] Performance impact of Debugger

2013-10-15 Thread Jim Blandy
On 10/14/2013 10:18 AM, Jason Orendorff wrote: On 10/14/13 12:03 PM, Jim Blandy wrote: [...]there's probably no need to separate bytecode instructions that have no visible side effects from their successors. Filed bug 926528, " JSD2: Reduce the number of positions where a breakpoint can be plac

Re: [JS-internals] Performance impact of Debugger

2013-10-14 Thread Jason Orendorff
On 10/14/13 12:03 PM, Jim Blandy wrote: > [...]there's probably no need to separate bytecode instructions that > have no visible side effects from their successors. Filed bug 926528, " JSD2: Reduce the number of positions where a breakpoint can be placed". https://bugzilla.mozilla.org/show_bug.cgi?

Re: [JS-internals] Performance impact of Debugger

2013-10-14 Thread Jim Blandy
On 10/14/2013 05:17 AM, Jan de Mooij wrote: Especially if Esprima has many short-running function calls, the 5x instead of 3x slowdown is not unexpected I think. It should be easy to temporarily disable the prologue/epilogue calls and see how much faster that makes us (search for debugMode_ in

Re: [JS-internals] Performance impact of Debugger

2013-10-14 Thread Jan de Mooij
On Sat, Oct 12, 2013 at 2:57 PM, Jason Orendorff wrote: > What's up with #2? Do the debug epilogues really cost that much? Or are > we accidentally inhibiting Baseline somehow? When debug mode is enabled we do the following: (1) Call DebugPrologue and DebugEpilogue (jit/VMFunctions.cpp) when ent

Re: [JS-internals] Performance impact of Debugger

2013-10-12 Thread Jason Orendorff
On 10/11/13 8:52 PM, Jim Blandy wrote: > Hi! I wrote a simple script to measure the impact of Debugger on > SpiderMonkey's performance. Thanks for the early results. OK, so in short: 1. disabling Ion: 3x slower 2. enabling debugger: 5x slower 3. enabling debugger with an onEnterFrame hook: 14x sl

[JS-internals] Performance impact of Debugger

2013-10-11 Thread Jim Blandy
Hi! I wrote a simple script to measure the impact of Debugger on SpiderMonkey's performance. This measures how long the Esprima JavaScript parser takes to parse itself. Here are preliminary results, running the benchmark scripts at https://github.com/jimblandy/benchmark-debugger. The variance