[go] cmd/compile: add the concept of dual instructions to ssa/_gen
Jorropo (Gerrit)
Jorropo has uploaded the change for review![]( "Open in Gerrit")
Commit message
cmd/compile: add the concept of dual instructions to ssa/_gen
commutative ops were renamed triviallyCommutative.
And I've added a dual field to opInfo, which defaults 0 for
non commutative ops, and self for triviallyCommutative ones.
Due to the diff being huge dual instructions are completely
useless for now, as this means this diff is only the field
rename and gen changes.
Change diff
Change information
- M src/cmd/compile/internal/ssa/_gen/386Ops.go
- M src/cmd/compile/internal/ssa/_gen/AMD64Ops.go
- M src/cmd/compile/internal/ssa/_gen/ARM64Ops.go
- M src/cmd/compile/internal/ssa/_gen/ARMOps.go
- M src/cmd/compile/internal/ssa/_gen/LOONG64Ops.go
- M src/cmd/compile/internal/ssa/_gen/MIPS64Ops.go
- M src/cmd/compile/internal/ssa/_gen/MIPSOps.go
- M src/cmd/compile/internal/ssa/_gen/PPC64Ops.go
- M src/cmd/compile/internal/ssa/_gen/RISCV64Ops.go
- M src/cmd/compile/internal/ssa/_gen/S390XOps.go
- M src/cmd/compile/internal/ssa/_gen/genericOps.go
- M src/cmd/compile/internal/ssa/_gen/main.go
- M src/cmd/compile/internal/ssa/_gen/rulegen.go
- M src/cmd/compile/internal/ssa/_gen/simdAMD64ops.go
- M src/cmd/compile/internal/ssa/_gen/simdgenericOps.go
- M src/cmd/compile/internal/ssa/cse.go
- M src/cmd/compile/internal/ssa/op.go
- M src/cmd/compile/internal/ssa/opGen.go
- M src/cmd/compile/internal/ssa/regalloc.go
- M src/cmd/compile/internal/ssa/schedule.go
Related details
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo uploaded new patchset![]( "Open in Gerrit")
Related details
- Jorropo
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo uploaded new patchset![]( "Open in Gerrit")
Related details
- Jorropo
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo voted Auto-Submit+1![]( "Open in Gerrit")
- Keith Randall
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
David Chase (Gerrit)
David Chase added 1 comment![]( "Open in Gerrit")
Will try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
Related details
- Jorropo
- Keith Randall
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo added 1 comment![]( "Open in Gerrit")
Will try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
I am not expecting this to get in before the freeze. Don't worry about being timely.
Related details
- David Chase
- Keith Randall
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo added 1 comment![]( "Open in Gerrit")
JorropoWill try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
I am not expecting this to get in before the freeze. Don't worry about being timely.
If conflicts are too bad I can keep the `commutative` name but I think `triviallyCommutative` is better.
Keith Randall (Gerrit)
Keith Randall added 1 comment![]( "Open in Gerrit")
JorropoWill try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
JorropoI am not expecting this to get in before the freeze. Don't worry about being timely.
If conflicts are too bad I can keep the `commutative` name but I think `triviallyCommutative` is better.
I'd really rather leave `commutative` as it is currently named and have another name for this new thing. It is kind of anticommutative, but not quite. dualCommutative or skewCommutative?
Related details
- David Chase
- Jorropo
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo added 1 comment![]( "Open in Gerrit")
JorropoWill try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
JorropoI am not expecting this to get in before the freeze. Don't worry about being timely.
Keith RandallIf conflicts are too bad I can keep the `commutative` name but I think `triviallyCommutative` is better.
I'd really rather leave `commutative` as it is currently named and have another name for this new thing. It is kind of anticommutative, but not quite. dualCommutative or skewCommutative?
- David Chase
- Keith Randall
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Keith Randall (Gerrit)
Keith Randall added 1 comment![]( "Open in Gerrit")
JorropoWill try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
JorropoI am not expecting this to get in before the freeze. Don't worry about being timely.
Keith RandallIf conflicts are too bad I can keep the `commutative` name but I think `triviallyCommutative` is better.
JorropoI'd really rather leave `commutative` as it is currently named and have another name for this new thing. It is kind of anticommutative, but not quite. dualCommutative or skewCommutative?
`commuted` ?
`asm: "CMOVQEQ", commuted: "CMOVQNE"`
Sure, an optional `commuted` or `commutedOp` field would be fine.
Related details
- David Chase
- Jorropo
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo uploaded new patchset![]( "Open in Gerrit")
- TryBots-Pass: LUCI-TryBot-Result+1 by golang...@luci-project-accounts.iam.gserviceaccount.com
Related details
- David Chase
- Jorropo
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
No-Wait-Release
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo added 1 comment![]( "Open in Gerrit")
JorropoWill try to get this reviewed in a timely way, there's loads of conflicts with machine-generated code from SIMD right now. In theory a trivial update to the generators, in practice we're already stepping on each other's toes trying to land SIMD work.
JorropoI am not expecting this to get in before the freeze. Don't worry about being timely.
Keith RandallIf conflicts are too bad I can keep the `commutative` name but I think `triviallyCommutative` is better.
JorropoI'd really rather leave `commutative` as it is currently named and have another name for this new thing. It is kind of anticommutative, but not quite. dualCommutative or skewCommutative?
Keith Randall`commuted` ?
`asm: "CMOVQEQ", commuted: "CMOVQNE"`
Sure, an optional `commuted` or `commutedOp` field would be fine.
- David Chase
- Keith Randall
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
No-Wait-Release
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Jorropo (Gerrit)
Jorropo uploaded new patchset![]( "Open in Gerrit")
- David Chase
- Jorropo
- Keith Randall
- Martin Möhrmann
Code-Review
No-Unresolved-Comments
No-Wait-Release
Review-Enforcement
TryBots-Pass
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |