Closed Bug 527246 Opened 16 years ago Closed 16 years ago

Nanojit should have load/store 32-bit float operations

Categories

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

Product:
Tamarin Graveyard
Component:
Baseline JIT (CodegenLIR)
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
VERIFIED FIXED

People

(Reporter: stejohns, Assigned: stejohns)

References

Details

It would be immensely useful to have an operation that loads a 32-bit float and extends it to 64-bit, and an operation that coerces a 64-bit double to 32-bit and stores in that form. To be sure, these operations can be done with existing nanojit, but (e.g.) SSE2 allows these to be done via single instructions, so providing support at the NJ level would allow for much more efficient code for these operations.
Blocks: 458279
Prototype patch for this is included in a patch for bug 527083.
fixed as part of https://bugzilla.mozilla.org/show_bug.cgi?id=527083
Status: NEW → RESOLVED
Closed: 16 years ago
Resolution: --- → FIXED
Assignee: nobody → stejohns
Engineering work item. Marking as verified.
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.