Closed Bug 551864 Opened 15 years ago Closed 15 years ago

JM: Get JM compiler working on Windows/x86

Categories

(Core :: JavaScript Engine, defect)

Product:
Core
Component:
JavaScript Engine
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED FIXED

People

(Reporter: dmandelin, Assigned: dmandelin)

References

Details

Attachments

(2 files)

Patch
9.84 KB, patch
Details | Diff | Splinter Review
Further patch for optimized builds
10.79 KB, patch
Details | Diff | Splinter Review
Mostly this means porting the asm trampolines.
Attached patch Patch — — Splinter Review
Depends on: 551865
$ ./sunspider-compare-results --shell=/cygdrive/c/sources/jaegermonkey/js/src/o pt/shell/js /tmp/jss0 /tmp/jss1 TEST COMPARISON FROM TO DETAILS ============================================================================= ** TOTAL **: 1.32x as fast 2362.8ms +/- 0.5% 1787.9ms +/- 0.3% significant ============================================================================= 3d: 1.123x as fast 326.2ms +/- 0.6% 290.5ms +/- 0.8% significant cube: 1.23x as fast 109.5ms +/- 1.0% 89.0ms +/- 1.1% significant morph: ?? 107.0ms +/- 0.6% 107.2ms +/- 1.3% not conclusive: might be *1.002x as slow* raytrace: 1.163x as fast 109.7ms +/- 0.6% 94.3ms +/- 1.1% significant access: 1.37x as fast 403.6ms +/- 1.7% 295.4ms +/- 0.9% significant binary-trees: 1.036x as fast 39.8ms +/- 1.1% 38.4ms +/- 1.0% significant fannkuch: 1.76x as fast 190.9ms +/- 3.2% 108.4ms +/- 1.8% significant nbody: 1.025x as fast 123.1ms +/- 0.5% 120.1ms +/- 0.6% significant nsieve: 1.75x as fast 49.8ms +/- 0.9% 28.5ms +/- 1.8% significant bitops: 2.03x as fast 306.8ms +/- 0.5% 151.2ms +/- 0.4% significant 3bit-bits-in-byte: 2.73x as fast 60.1ms +/- 1.6% 22.0ms +/- 1.5% significant bits-in-byte: 2.78x as fast 79.3ms +/- 0.7% 28.5ms +/- 1.3% significant bitwise-and: 1.40x as fast 58.2ms +/- 0.8% 41.5ms +/- 0.9% significant nsieve-bits: 1.84x as fast 109.2ms +/- 0.4% 59.2ms +/- 0.5% significant controlflow: 1.188x as fast 40.5ms +/- 0.9% 34.1ms +/- 0.7% significant recursive: 1.188x as fast 40.5ms +/- 0.9% 34.1ms +/- 0.7% significant crypto: 1.43x as fast 177.9ms +/- 0.4% 124.6ms +/- 0.5% significant aes: 1.54x as fast 69.6ms +/- 0.7% 45.1ms +/- 0.5% significant md5: 1.34x as fast 53.1ms +/- 0.4% 39.6ms +/- 1.3% significant sha1: 1.38x as fast 55.2ms +/- 0.5% 39.9ms +/- 0.6% significant date: 1.142x as fast 157.4ms +/- 0.8% 137.8ms +/- 0.5% significant format-tofte: 1.27x as fast 64.9ms +/- 1.7% 51.0ms +/- 0.7% significant format-xparb: 1.066x as fast 92.5ms +/- 0.7% 86.8ms +/- 0.8% significant math: 1.139x as fast 299.4ms +/- 1.3% 262.8ms +/- 0.8% significant cordic: 1.38x as fast 142.4ms +/- 2.5% 103.5ms +/- 0.5% significant partial-sums: - 99.2ms +/- 1.7% 97.8ms +/- 1.4% spectral-norm: *1.064x as slow* 57.8ms +/- 0.8% 61.5ms +/- 1.1% significant regexp: 3.84x as fast 150.3ms +/- 0.6% 39.1ms +/- 1.6% significant dna: 3.84x as fast 150.3ms +/- 0.6% 39.1ms +/- 1.6% significant string: 1.107x as fast 500.7ms +/- 1.0% 452.4ms +/- 1.0% significant base64: 1.31x as fast 50.2ms +/- 2.6% 38.4ms +/- 1.8% significant fasta: 1.121x as fast 100.1ms +/- 0.5% 89.3ms +/- 0.5% significant tagcloud: 1.033x as fast 114.8ms +/- 1.3% 111.1ms +/- 1.0% significant unpack-code: - 106.6ms +/- 0.3% 106.7ms +/- 0.6% validate-input: 1.21x as fast 129.0ms +/- 2.8% 106.9ms +/- 3.9% significant Note that 110ms of the speedup comes from regexp-dna being jitted, so that doesn't really count. It's still a 27% speedup if you take that out (c.f. the 18% I got on Mac).
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: