Viewing Issue Simple Details
[ Jump to Notes ]
|
[ View Advanced ]
[ Issue History ]
[ Print ]
|
ID |
Category |
Severity |
Reproducibility |
Date Submitted |
Last Update |
0007583 |
[Squeak] Kernel |
minor |
always |
12-13-10 21:40 |
12-31-10 06:40 |
|
Reporter |
matthewf |
View Status |
public |
|
Assigned To |
|
Priority |
normal |
Resolution |
open |
|
Status |
new |
|
Product Version |
|
|
Summary |
0007583: Float does not use FloatMathPlugin for bit-consistent float math across platforms |
Description |
Croquet requires that all float operations be bit-reproducable across platforms, and the FloatMathPlugin was made to ensure this works. However, the squeak image does not currently ensure this. Attached is a file-in from Croquet that puts FloatMathPlugin (which is in the VM but apparently unused) to good use. It is also available in the inbox: Kernel-mtf.527 |
Additional Information |
To test this, load Croquet-Kernel (attached), and ignore the initialization errors. Run the test case CroquetVMTests in Croquet-Kernel-Tests. All the float tests in it will pass after the attached changeset is filed in. There are 3 failing and 2 error test cases when I run this on a clean trunk image, but those are unrelated to this issue |
|
Attached Files |
UseFloatMathPlugin.2.cs [^] (11,508 bytes) 12-13-10 21:40
Croquet-Kernel-mtf.86.mcz [^] (83,357 bytes) 12-13-10 21:58 |
|