[llvm-commits] [llvm] r171946 - in /llvm/trunk: include/llvm/CodeGen/TargetSchedule.h include/llvm/MC/MCSchedule.h include/llvm/Target/TargetSchedule.td lib/CodeGen/MachineScheduler.cpp lib/Target/ARM/ARMScheduleA9.td lib/Target/X86/X86Schedule.td lib/Target/X86/X86ScheduleAtom.td test/CodeGen/ARM/misched-inorder-latency.ll test/CodeGen/PowerPC/misched-inorder-latency.ll utils/TableGen/SubtargetEmitter.cpp

[Hal Finkel]

----- Original Message -----
> From: "Andrew Trick" <atrick at apple.com>
> To: "Hal Finkel" <hfinkel at anl.gov>
> Cc: llvm-commits at cs.uiuc.edu
> Sent: Friday, January 11, 2013 12:01:33 PM
> Subject: Re: [llvm-commits] [llvm] r171946 - in /llvm/trunk: include/llvm/CodeGen/TargetSchedule.h
> include/llvm/MC/MCSchedule.h include/llvm/Target/TargetSchedule.td lib/CodeGen/MachineScheduler.cpp
> lib/Target/ARM/ARMScheduleA9.td lib/Target/X86/X86Schedule.td lib/Target/X86/X86ScheduleAtom.td
> test/CodeGen/ARM/misched-inorder-latency.ll test/CodeGen/PowerPC/misched-inorder-latency.ll
> utils/TableGen/SubtargetEmitter.cpp
> 
> 
> On Jan 11, 2013, at 9:38 AM, Hal Finkel <hfinkel at anl.gov> wrote:
> 
> > ----- Original Message -----
> >> From: "Andrew Trick" <atrick at apple.com>
> >> To: llvm-commits at cs.uiuc.edu
> >> Sent: Tuesday, January 8, 2013 9:36:49 PM
> >> Subject: [llvm-commits] [llvm] r171946 - in /llvm/trunk:
> >> include/llvm/CodeGen/TargetSchedule.h
> >> include/llvm/MC/MCSchedule.h include/llvm/Target/TargetSchedule.td
> >> lib/CodeGen/MachineScheduler.cpp
> >> lib/Target/ARM/ARMScheduleA9.td lib/Target/X86/X86Schedule.td
> >> lib/Target/X86/X86ScheduleAtom.td
> >> test/CodeGen/ARM/misched-inorder-latency.ll
> >> test/CodeGen/PowerPC/misched-inorder-latency.ll
> >> utils/TableGen/SubtargetEmitter.cpp
> >> 
> >> Author: atrick
> >> Date: Tue Jan  8 21:36:49 2013
> >> New Revision: 171946
> >> 
> >> URL: http://llvm.org/viewvc/llvm-project?rev=171946&view=rev
> >> Log:
> >> MIsched: add an ILP window property to machine model.
> >> 
> >> This was an experimental option, but needs to be defined
> >> per-target. e.g. PPC A2 needs to aggressively hide latency.
> >> 
> >> I converted some in-order scheduling tests to A2. Hal is working
> >> on
> >> more test cases.
> > 
> > Andy, thanks!
> > 
> > Also, do you want to do the same thing with enable-aa-sched-mi?
> 
> Please add a target hook if you'd like. We still want a command line
> option for toggling this, and it doesn't need to be in the machine
> model. It's just a default policy that can be overridden by
> subtargets.

Will do, thanks! I'll certainly keep a command-line toggle to override the target default.

 -Hal

> 
> -Andy
> 

-- 
Hal Finkel
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory