This CL generates optimizations for masked variant of AVX512
instructions for patterns:
x.Op(y).Merge(z, mask) => OpMasked(z, x, y mask), where OpMasked is
resultInArg0.
Change-Id: Ife7ccc9ddbf76ae921a085bd6a42b965da9bc179
Reviewed-on: https://go-review.googlesource.com/c/go/+/718160
Reviewed-by: David Chase <drchase@google.com>
TryBot-Bypass: Junyang Shao <shaojunyang@google.com>
For 512-bits they are unchanged. This CL adds the optimization rules for
128/256-bits under feature check.
This CL also fixed a bug for masked load variant of instructions and
make them zeroing by default as well.
Change-Id: I6fe395541c0cd509984a81841420e71c3af732f2
Reviewed-on: https://go-review.googlesource.com/c/go/+/717822
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
These should really be machine ops only.
Change-Id: Idcc611719eff068153d88c5162dd2e0883e5e0ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/717821
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL also fixed some bugs left in CL 712181.
Change-Id: I9cb6cd9fbaef307f352809bf21b8fec3eb62721a
Reviewed-on: https://go-review.googlesource.com/c/go/+/712361
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This includes an non-public intrinsic for testing,
and a test. Optimizations using this instruction
will follow in another CL.
Change-Id: I7f7a93212249a16a30bd1379c717f8a7f9915daf
Reviewed-on: https://go-review.googlesource.com/c/go/+/708995
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Junyang Shao <shaojunyang@google.com>
API naming changes.
This CL also remove AddDotProductPairsSaturated.
Change-Id: I02e6d45268704f3ed4eaf62f0ecb7dc936b42124
Reviewed-on: https://go-review.googlesource.com/c/go/+/710935
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
AVXAES is a composite feature set, Intel did listed it as "AVXAES" in
the XED data instead of separating them.
The tests will be in the next CL.
Change-Id: I89c97261f2228b2fdafb48f63e82ef6239bdd5ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/706055
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Using this name until a better one appears:
x.Select128FromPair(3, 2, y)
Includes test for constant and variable case.
Checks for unexpected immediates (using the zeroing flag,
which is not supported for this intrinsic) and panics.
Change-Id: I9249475d6572968c127b4ee9e00328d717c07578
Reviewed-on: https://go-review.googlesource.com/c/go/+/705496
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
These are package private, and will be hidden behind
other methods in a following CL with a more general
interface.
Change-Id: Id090a5de06a0e2aed5cc60a11ff627c5e3b9c52d
Reviewed-on: https://go-review.googlesource.com/c/go/+/698577
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The previous CL left a bug in the xed parser so that the generator can
generate rules rewriting an AVX instruction to AVX512 instruction.
This CL fixes that.
Change-Id: I0df7e7dc6c936ce7add24a757ce7f44a15917fef
Reviewed-on: https://go-review.googlesource.com/c/go/+/703399
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL adds the load + const imm8 variants ofr many instructions.
Change-Id: I46116906077e33eabccc111be6d16019002f3474
Reviewed-on: https://go-review.googlesource.com/c/go/+/703395
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This CL adds the machine ops for memory-op and also their prog writing
logic.
This CL also fixes a bug in the XED parser. Previously the
merge of machine ops is not checking the CPU feature, so some AVX
instruction might have their "memFeatures" field set incorrectly.
However since that field is not used until this CL, putting the fix here
should be ok.
Change-Id: I91031cbbf63453257473dd1d2ff47f7496d1a01d
Reviewed-on: https://go-review.googlesource.com/c/go/+/701198
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
these are building blocks for transpose, not sure of their
best names yet.
Change-Id: I3800a55de9fa7fde2590ca822894c8a75387dec3
Reviewed-on: https://go-review.googlesource.com/c/go/+/698576
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The namings follow the following convention:
- If its indices are from constant, amend "Constant" to the name.
- If its indices are used by multiple groups, mend "Grouped" to the
name.
- If its indexing only the low part, amend "Lo", similarly "Hi".
Change-Id: I6a58f5dae54c882ebd59f39b5288f6f3f14d957f
Reviewed-on: https://go-review.googlesource.com/c/go/+/698296
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
This CL should complete the conversions between int and uint.
Change-Id: I46742a62214f346e014a68b9c72a9b116a127f67
Reviewed-on: https://go-review.googlesource.com/c/go/+/698236
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Commit-Queue: David Chase <drchase@google.com>
Reviewed-by: David Chase <drchase@google.com>
Downsizing conversions' truncating version complete. Saturation ver not
done.
Change-Id: I710976c2b5329e2882763d60fcef2a827213df09
Reviewed-on: https://go-review.googlesource.com/c/go/+/697975
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This new API is given the name SumAbsDiff, a slightly-longer name for
its canonical abbreviation SAD(Sum-Absolute-Differences).
This instruction has some similar semantic's one, but their semantic is much more
specific and complex: MPSADBW, VDBPSADBW. They should have a more
specific name given this fact.
Change-Id: Ied9144440f82919c3c2d45ae4ce5b961ae91a020
Reviewed-on: https://go-review.googlesource.com/c/go/+/697776
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This is not the end of such peephole optimizations, there
would need to be many of these for many simd operations.
Change-Id: I4511f6fac502bc7259c1c4414c96f56eb400c202
Reviewed-on: https://go-review.googlesource.com/c/go/+/697157
TryBot-Bypass: David Chase <drchase@google.com>
Commit-Queue: David Chase <drchase@google.com>
Reviewed-by: Junyang Shao <shaojunyang@google.com>
This also required a "always use operation with least
OverrideBase" filter in choosing the machine instructions.
The order of generated HW operations is slightly
modified because the Float version of GetElem
appears earlier in the sorted operations list,
though it is not chosen to generate the HW Op.
Change-Id: I95fa67afca9c8b6f4f18941fdcaf69afdad8055b
Reviewed-on: https://go-review.googlesource.com/c/go/+/696375
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
this also removes AVX512 versions of the operations
that would use the same names, but not run on AVX2-only
includes files generated by simdgen CL 692355
Change-Id: Iff29042245b7688133fed49a03e681e85235b8a8
Reviewed-on: https://go-review.googlesource.com/c/go/+/692335
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Junyang Shao <shaojunyang@google.com>
Generated by simdgen CL 693599
This turned out to require some additional work in
other places, including filling in missing
methods (use OverwriteBase to get FP versions).
Also includes a test.
Change-Id: I2efe8967837834745f9cae661d4d4dcbb5390b6f
Reviewed-on: https://go-review.googlesource.com/c/go/+/693758
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Junyang Shao <shaojunyang@google.com>
- Absolute -> Abs
- ApproximateReciprocal -> Reciprocal
- Other derived apis also changed.
- Round -> RoundToEven
- Other derived apis also changed.
- Drop DotProdBroadcast
- Fused(Mul|Add)(Mul|Add)? -> remove the "Fused"
- MulEvenWiden -> remove 64bit
- MulLow -> Mul, add unit
- PairDotProd -> DotProdPairs
- make AddDotProdPairs machine ops only - peepholes will be in another
CL at dev.simd.
- PopCount -> OnesCount
- Saturated* -> *Saturated
- Fix (Add|Sub)Saturated uint mappings.
- UnsignedSignedQuadDotProdAccumulate -> AddDotProdQuadruple
- The "DotProdQuadruple" instruction does not exist, so no peepholes for
this.
This CL is generated by CL 694095.
Change-Id: If4110cc04ab96240cf56f2348d35ed2a719687de
Reviewed-on: https://go-review.googlesource.com/c/go/+/694115
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is generated by CL 693336.
Change-Id: Ic1712d49fcad0544fa3c19b0249d8bc65b347104
Reviewed-on: https://go-review.googlesource.com/c/go/+/693375
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is generated by CL 693335.
Change-Id: Ie9adda526573f979ec7e4f535033ba29236cc5cb
Reviewed-on: https://go-review.googlesource.com/c/go/+/693355
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Generated by arch/internal/simdgen CL 693175
These methods are not public because of simdgen-induced name/signature
issues, and because their addition was motivated by the need for
emulation tools.
The specific name signature problems are:
1) one set of instructions has the "Masked" suffix (because of how
that is incorporated into names) and the other set does not (though I
suppose the operation could be renamed).
2) because the AVX2 instruction is bytes-only, to get the signature
right, requires "OverwriteBase" but OverwriteBase also requires
OverwriteClass and "simdgen does not support [OverwriteClass] in
inputs".
3) the default operation order is false, true, but we want this in a
"x.Merged(y, mask)" that pairs with "x.Masked(mask)" where the true
case is x and the false case is y/zero, but the default ordering for
VPBLENDVB and VPBLENDMB is false->x and true->y.
4) VPBLENDVB only comes in byte width, which causes problems
for floats.
All this may get fixed in the future, for now it is just an
implementation detail.
Change-Id: I61b655c7011e2c33f8644f704f886133c89d2f15
Reviewed-on: https://go-review.googlesource.com/c/go/+/693155
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Junyang Shao <shaojunyang@google.com>
This CL is generated by CL 693136.
Change-Id: Ifd2278d3f927efa008a14cc5e592e7c14b7120ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/693157
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
This CL is generated by CL 692219.
Change-Id: I50fa919f1edc5c6505bc6d3238f65b37fc7628b5
Reviewed-on: https://go-review.googlesource.com/c/go/+/692156
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Generated by arch/internal/simdgen CL 689735
A small number of conversions for testing purposes
Change-Id: I4d52c643d08c02794c3fea9778bb1ecbb5507de4
Reviewed-on: https://go-review.googlesource.com/c/go/+/689716
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is partially generated by CL 689775.
Change-Id: I0c36fd2a44706c88db1a1d5ea4a6d0b9f891d85f
Reviewed-on: https://go-review.googlesource.com/c/go/+/689795
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is generated by CL 687975.
Change-Id: I21707d108773cc6d8e6f07aaed60e756faa1e6cb
Reviewed-on: https://go-review.googlesource.com/c/go/+/687995
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
This CL also added some tests for them.
This CL is generated by CL 687919.
Change-Id: I9ddd2cd23bb98ecca91bfbeaffd62faa4bd85e0d
Reviewed-on: https://go-review.googlesource.com/c/go/+/687939
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
This CL is generated by CL 687915.
Change-Id: I1a2fb031c086b2b23fd135c48f8494ba5122493a
Reviewed-on: https://go-review.googlesource.com/c/go/+/687916
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
This CL does:
1. Removes ShiftRightSignExtended, default signed vectors to shift
arithmetic, and unsigned to shift logical.
2. Add the missing Shifts which were left out by YAML error in the
generator.
This CL is generated by CL 687595.
Change-Id: I663115498adb91c82e89a8476e6748794e997cfa
Reviewed-on: https://go-review.googlesource.com/c/go/+/687596
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This CL is generated by CL 686817.
Change-Id: I19b8e468594514b2b1c99f8ad766f78b5e194c80
Reviewed-on: https://go-review.googlesource.com/c/go/+/686876
TryBot-Bypass: David Chase <drchase@google.com>
Reviewed-by: David Chase <drchase@google.com>
This CL contains codes generated by CL 686556.
Change-Id: I4d7287476b478efdc186a64c12de33528c7fb0af
Reviewed-on: https://go-review.googlesource.com/c/go/+/686476
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is generated by CL 686575.
Change-Id: I1483189a1ae9bed51446fd69daab3f7b128549ae
Reviewed-on: https://go-review.googlesource.com/c/go/+/686516
Reviewed-by: David Chase <drchase@google.com>
TryBot-Bypass: David Chase <drchase@google.com>
This CL is generated by CL 686555.
Change-Id: I0efb86a919692cd97c1c5b6365d77361a30bf7cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/686496
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
This is generated by simdgen CL 684080
and should be submitted after it.
Also includes tests.
Change-Id: I1d680911134d8fb92f4deccae4ec373f3ed9f752
Reviewed-on: https://go-review.googlesource.com/c/go/+/684115
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is created by simdgen CL 684055 and
should be submitted after it.
Also includes a test.
Change-Id: I2ad7ae51d11cfc19745e866150e2eaf010d4ea49
Reviewed-on: https://go-review.googlesource.com/c/go/+/684077
Reviewed-by: Junyang Shao <shaojunyang@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL is generated by CL 684275.
Change-Id: Ie1efd0979af0ef0a56781bf9013071bf4d2c52c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/684175
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This CL is generated by CL 683475.
Change-Id: I9e3ac6aff6f711cb26ff85e4c8729d9e2cc38e7d
Reviewed-on: https://go-review.googlesource.com/c/go/+/683715
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
This CL is generated by simdgen CL 683836
and this CL should be submitted after its
generator.
Change-Id: I1aa893b185826ad1f9fb60b85c75eda31f70623b
Reviewed-on: https://go-review.googlesource.com/c/go/+/683797
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
