x86: Update code-generator documentation #7375
Conversation
doc/compiler/x/OpCodeEncoding.md
Outdated
|
|
||
| ``` | ||
| 128 - SSE, VEX_L128, EVEX_L128 | ||
| 256 - VEX_L256, EVEX_L128 |
There was a problem hiding this comment.
Should this be EVEX_L256 ?
doc/compiler/x/OpCodeEncoding.md
Outdated
|
|
||
| ### Generating SIMD instructions | ||
|
|
||
| Generating SIMD instructions is simple. You may optionally specify the method of which to encode |
doc/compiler/x/OpCodeEncoding.md
Outdated
| generateRegRegInstruction(TR::InstOpCode::MOVDQURegReg, node, resultReg, valueReg, cg, OMR::X86::VEX_L128); | ||
| ``` | ||
|
|
||
| By not specifying an opcode encoding method, the code-generator will use method specified |
There was a problem hiding this comment.
"code-generator" -> "code generator"
It isn't our convention to hyphenate these words.
doc/compiler/x/OpCodeEncoding.md
Outdated
| ``` | ||
|
|
||
| By not specifying an opcode encoding method, the code-generator will use method specified | ||
| on opcodes definition. If the instruction is labeled as VEX_L128, and AVX is not supported, |
There was a problem hiding this comment.
"use the method specified on an opcode's definition."
doc/compiler/x/OpCodeEncoding.md
Outdated
| } Encoding; | ||
| ``` | ||
|
|
||
| #### 128-Bit vmovdqu |
There was a problem hiding this comment.
Is this included just as an example, or is there some other reason for calling out 128-bit vmovdqu ?
doc/compiler/x/OpCodeEncoding.md
Outdated
|
|
||
| To dynamically find the best encoding method you may call ```getSIMDEncoding(&cpu, vl)```. | ||
| This query uses flags stored on the instruction to determine the best encoding method for | ||
| the given instruction, cpu, and vector length. |
doc/compiler/x/OpCodeEncoding.md
Outdated
|
|
||
| #### Dynamically determining the best encoding method | ||
|
|
||
| To dynamically find the best encoding method you may call ```getSIMDEncoding(&cpu, vl)```. |
There was a problem hiding this comment.
Are there restrictions on when you can call this function? e.g., can you only call it during binary encoding?
There was a problem hiding this comment.
No, doesn't matter when/where you call it
doc/compiler/x/OpCodeEncoding.md
Outdated
|
|
||
| #### Dynamically determining the best encoding method | ||
|
|
||
| To dynamically find the best encoding method you may call ```getSIMDEncoding(&cpu, vl)```. |
There was a problem hiding this comment.
Can you mention the class this function is a member of?
| This method will return ```OMR::X86::Bad``` if the opcode is not supported at the given | ||
| vector length and will also throw an assertion failure if the instruction is missing | ||
| CPU feature requirements flags. | ||
|
|
There was a problem hiding this comment.
Precede this with "For example," here if that's what it is.
doc/compiler/x/OpCodeEncoding.md
Outdated
|
|
||
| #### Dynamically determining the best encoding method | ||
|
|
||
| To dynamically find the best encoding method you may call ```getSIMDEncoding(&cpu, vl)```. |
There was a problem hiding this comment.
Consider adding a permalink to where it is defined.
Signed-off-by: Bradley Wood <[email protected]>
|
@0xdaryl See the force-push |
No description provided.