|
|
|
@187
|
[187]
|
04/30/08 16:05:32 |
Don Clugston |
Added prod(). Use .ptr to get raw data, so it works with Bill Baxter's … |
|
|
|
@175
|
[175]
|
01/14/08 07:44:22 |
Don Clugston |
Intrinsics no longer need to be wrapped in parentheses. |
|
|
|
@173
|
[173]
|
01/11/08 02:38:02 |
Don Clugston |
* Refactored SyntaxTree?.
* Scalar folding of min, max.
* Improved … |
|
|
|
@172
|
[172]
|
01/10/08 02:48:44 |
Don Clugston |
Added sqrt intrinsic, with codegen for both X87 and SSE. |
|
|
|
@171
|
[171]
|
01/08/08 07:11:22 |
Don Clugston |
Added abs() intrinsic. Added codegen for unary minus. Bugfix for rank … |
|
|
|
@169
|
[169]
|
01/07/08 03:17:39 |
Don Clugston |
* Improved treatment of intrinsic functions.
* Added sum() intrinsic to … |
|
|
|
@168
|
[168]
|
01/04/08 07:56:21 |
Don Clugston |
Added back-end code generation for dot product for X87 and SSE/SSE2. |
|
|
|
@167
|
[167]
|
01/03/08 09:23:43 |
Don Clugston |
Added the front end code generation for dot product (this is also the way … |
|
|
|
@166
|
[166]
|
12/21/07 03:17:57 |
Don Clugston |
X87 now generates correct code for chained assignment.
X87 now does … |
|
|
|
@165
|
[165]
|
12/20/07 16:47:39 |
Don Clugston |
x87 code gen now generates code for strided vectors.
No longer tries to … |
|
|
|
@164
|
[164]
|
12/18/07 01:03:09 |
Don Clugston |
Use { to delimit compounds - no space. Allows them to be nested. |
|
|
|
@163
|
[163]
|
12/17/07 03:32:00 |
Don Clugston |
Folding of unary minus -- it gets eliminated in the majority of cases. |
|
|
|
@162
|
[162]
|
12/14/07 07:40:53 |
Don Clugston |
Scalar folding now done by a visitor -- all deprecated code removed. |
|
|
|
@161
|
[161]
|
12/14/07 03:25:52 |
Don Clugston |
The index folding visitor is now part of the main line. A preliminary … |
|
|
|
@160
|
[160]
|
12/13/07 05:23:00 |
Don Clugston |
The index folding now supports vector*matrix and matrix*matrix multiplies, … |
|
|
|
@159
|
[159]
|
12/12/07 03:12:39 |
Don Clugston |
Index folding visitor -- not linked into the main line yet. Minor … |
|
|
|
@158
|
[158]
|
12/07/07 14:51:35 |
Don Clugston |
BLADE is now object-oriented! Now uses the Visitor pattern for parsing. … |
|
|
|
@157
|
[157]
|
12/06/07 12:47:16 |
Don Clugston |
Simplify drop product.
Refactor |
|
|
|
@156
|
[156]
|
12/05/07 14:22:09 |
Don Clugston |
Fixed constant folding for dot (first attempt) |
|
|
|
@155
|
[155]
|
12/05/07 03:02:24 |
Don Clugston |
* Refactoring for strided vector length.
*Bugfix for the case of real … |
|
|
|
@153
|
[153]
|
12/01/07 03:36:16 |
Don Clugston |
Refactoring for strided vectors. |
|
|
|
@152
|
[152]
|
11/30/07 11:42:59 |
Don Clugston |
Support strided vector operations (D code only) |
|
|
|
@151
|
[151]
|
11/30/07 02:14:47 |
Don Clugston |
Now parsing and rank analysis supports combined slice & indices -- eg … |
|
|
|
@150
|
[150]
|
11/29/07 03:11:11 |
Don Clugston |
Strided array accesses are now parsed OK, and dimensionality errors are … |
|
|
|
@148
|
[148]
|
11/28/07 03:17:22 |
Don Clugston |
More support for array literals; fixed most (but not all) of the rank … |
|
|
|
@147
|
[147]
|
11/27/07 12:29:02 |
Don Clugston |
Support array literals. |
|
|
|
@146
|
[146]
|
11/26/07 02:59:29 |
Don Clugston |
A bit more support for dot product |
|
|
|
@145
|
[145]
|
11/22/07 02:18:30 |
Don Clugston |
First steps to support for dot product and other intrinsics in the front … |
|
|
|
@143
|
[143]
|
11/20/07 04:21:51 |
Don Clugston |
- *Nice* error messages. (No more template/mixin garbage!)
- $ works … |
|
|
|
@140
|
[140]
|
11/17/07 17:17:36 |
Don Clugston |
* Fixed a nasty codegen bug in X87.
* Speedup for enquote().
* Now uses … |
|
|
|
@139
|
[139]
|
11/16/07 12:37:50 |
Don Clugston |
Some CTFE speedups. Bugfix for scalar multiply folding. |
|
|
|
@138
|
[138]
|
11/15/07 14:42:22 |
Don Clugston |
Now x87 and D code generation use revised expressions. |
|
|
|
@136
|
[136]
|
11/14/07 13:04:41 |
Don Clugston |
Partial merger of the new RevisedExpression? into x87 code. |
|
|
|
@135
|
[135]
|
11/13/07 13:07:01 |
Don Clugston |
RevisedExpression? now uses a mapping from new to old variables, instead of … |
|
|
|
@133
|
[133]
|
11/13/07 04:47:28 |
Don Clugston |
Factored out the expression simplification code. Duplicate symbols now get … |