Bug 142036 – DFG SSA stack accesses shouldn't speak of VariableAccessDatas

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Filip Pizlo]

VariableAccessData is super useful in ThreadedCPS because it allows us to region about things that span basic blocks.  It also allows us to do stepwise type inference on variables.  But it's purely harmful in SSA.  SSA shouldn't speak of it.

The main effect of this is that it will allow us to easily introduce new stack accesses.

[Filip Pizlo]

Created attachment 247390 [details]
work in progress

[Filip Pizlo]

Created attachment 247402 [details]
the patch

[WebKit Commit Bot]

Attachment 247402 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:102:  Place brace on its own line for function definitions.  [whitespace/braces] [4]
ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:112:  Place brace on its own line for function definitions.  [whitespace/braces] [4]
ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:224:  Place brace on its own line for function definitions.  [whitespace/braces] [4]
ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:233:  Place brace on its own line for function definitions.  [whitespace/braces] [4]
ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:395:  Place brace on its own line for function definitions.  [whitespace/braces] [4]
ERROR: Source/JavaScriptCore/dfg/DFGStackLayoutPhase.cpp:193:  More than one command on the same line  [whitespace/newline] [4]
ERROR: Source/JavaScriptCore/CMakeLists.txt:209:  There should be exactly one empty line instead of 0 between "disassembler/ARM64/A64DOpcode.cpp" and "disassembler/ARMv7/ARMv7DOpcode.cpp".  [list/emptyline] [5]
Total errors found: 7 in 31 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Filip Pizlo]

(In reply to comment #3)
> Attachment 247402 [details] did not pass style-queue:
> 
> 
> ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:102:  Place
> brace on its own line for function definitions.  [whitespace/braces] [4]
> ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:112:  Place
> brace on its own line for function definitions.  [whitespace/braces] [4]
> ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:224:  Place
> brace on its own line for function definitions.  [whitespace/braces] [4]
> ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:233:  Place
> brace on its own line for function definitions.  [whitespace/braces] [4]
> ERROR: Source/JavaScriptCore/dfg/DFGPutStackSinkingPhase.cpp:395:  Place
> brace on its own line for function definitions.  [whitespace/braces] [4]

No.

> ERROR: Source/JavaScriptCore/dfg/DFGStackLayoutPhase.cpp:193:  More than one
> command on the same line  [whitespace/newline] [4]

Fixed.

> ERROR: Source/JavaScriptCore/CMakeLists.txt:209:  There should be exactly
> one empty line instead of 0 between "disassembler/ARM64/A64DOpcode.cpp" and
> "disassembler/ARMv7/ARMv7DOpcode.cpp".  [list/emptyline] [5]

No.

> Total errors found: 7 in 31 files
> 
> 
> If any of these errors are false positives, please file a bug against
> check-webkit-style.

[Michael Saboff]

Comment on attachment 247402 [details]
the patch

r=me

[Filip Pizlo]

Landed in http://trac.webkit.org/changeset/180691