Closed Bug 569838 Opened 14 years ago Closed 6 years ago

x86 code quality is generally improved if we use FASTCALL for all jit helpers.

Categories

(Tamarin Graveyard :: Baseline JIT (CodegenLIR), defect, P2)

Product:
Tamarin Graveyard
Component:
Baseline JIT (CodegenLIR)
Platform:
x86
All
Type:
defect

Tracking

(Not tracked)

Status:
RESOLVED WONTFIX
Milestone:
Q4 12 - Ellis

People

(Reporter: edwsmith, Assigned: phohense)

References

Details

(Whiteboard: PACMAN) 

(needs experimental evidence to back up this claim).
Severity: normal → minor
Priority: -- → P2
Whiteboard: PACMAN
Target Milestone: --- → flash10.2
Yes and please include architecture / processor specifics with any data, as there is most likely variance amongst the chips.
x86 is the only applicable architecture, and msvc/gcc are the only applicable compilers.  All other CPU types have a register rich calling convention that doesn't change for plain functions vs method calls. (and FASTCALL is an empy macro for them).
Right, I was referring to variants in x86 chip architectures that might manifest as deltas in performance.  For example, dual-core vs. core2 vs Nehalem based chips; where each may exhibit slightly different effects from this change.
Flags: flashplayer-bug-
Flags: flashplayer-injection-
Target Milestone: Q3 11 - Serrano → Q1 12 - Brannan
Depends on: 686664
Assignee: nobody → phohense
Target Milestone: Q1 12 - Brannan → Q4 12 - Ellis
Tamarin is a dead project now. Mass WONTFIX.
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
Tamarin isn't maintained anymore. WONTFIX remaining bugs.
You need to log in before you can comment on or make changes to this bug.