Treat integer range for vector type as union of ranges of individual
elements. With this semantics, most arith ops on vectors will work out
of the box, the only special handling needed for constants and vector
elements manipulation ops.
The end goal of these changes is to be able to optimize vectorized index
calculations.
This is intended as a fast pattern rewrite driver for the cases when a
simple walk gets the job done but we would still want to implement it in
terms of rewrite patterns (that can be used with the greedy pattern
rewrite driver downstream).
The new driver is inspired by the discussion in
https://github.com/llvm/llvm-project/pull/112454 and the LLVM Dev
presentation from @matthias-springer earlier this week.
This limitation comes with some limitations:
* It does not repeat until a fixpoint or revisit ops modified in place
or newly created ops. In general, it only walks forward (in the
post-order).
* `matchAndRewrite` can only erase the matched op or its descendants.
This is verified under expensive checks.
* It does not perform folding / DCE.
We could probably relax some of these in the future without sacrificing
too much performance.
`UnsignedWhenEquivalent` doesn't really need any dialect conversion
features and switching it normal patterns makes it more composable with
other patterns-based transformations (and probably faster).
The concept of a 'program point' in the original data flow framework is
ambiguous. It can refer to either an operation or a block itself. This
representation has different interpretations in forward and backward
data-flow analysis. In forward data-flow analysis, the program point of
an operation represents the state after the operation, while in backward
data flow analysis, it represents the state before the operation. When
using forward or backward data-flow analysis, it is crucial to carefully
handle this distinction to ensure correctness.
This patch refactors the definition of program point, unifying the
interpretation of program points in both forward and backward data-flow
analysis.
How to integrate this patch?
For dense forward data-flow analysis and other analysis (except dense
backward data-flow analysis), the program point corresponding to the
original operation can be obtained by `getProgramPointAfter(op)`, and
the program point corresponding to the original block can be obtained by
`getProgramPointBefore(block)`.
For dense backward data-flow analysis, the program point corresponding
to the original operation can be obtained by
`getProgramPointBefore(op)`, and the program point corresponding to the
original block can be obtained by `getProgramPointAfter(block)`.
NOTE: If you need to get the lattice of other data-flow analyses in
dense backward data-flow analysis, you should still use the dense
forward data-flow approach. For example, to get the Executable state of
a block in dense backward data-flow analysis and add the dependency of
the current operation, you should write:
``getOrCreateFor<Executable>(getProgramPointBefore(op),
getProgramPointBefore(block))``
In case above, we use getProgramPointBefore(op) because the analysis we
rely on is dense backward data-flow, and we use
getProgramPointBefore(block) because the lattice we query is the result
of a non-dense backward data flow computation.
related dsscussion:
https://discourse.llvm.org/t/rfc-unify-the-semantics-of-program-points/80671/8
corresponding PSA:
https://discourse.llvm.org/t/psa-program-point-semantics-change/81479
This commit marks the type converter in `populate...` functions as
`const`. This is useful for debugging.
Patterns already take a `const` type converter. However, some
`populate...` functions do not only add new patterns, but also add
additional type conversion rules. That makes it difficult to find the
place where a type conversion was added in the code base. With this
change, all `populate...` functions that only populate pattern now have
a `const` type converter. Programmers can then conclude from the
function signature that these functions do not register any new type
conversion rules.
Also some minor cleanups around the 1:N dialect conversion
infrastructure, which did not always pass the type converter as a
`const` object internally.
This change makes two (related) changes:
First, it updates the tablegen option for `ListOption` to emit a
`SmallVector` instead of an `ArrayRef`. This brings `ListOption` more
inline with the traditional `Option`, where values are typically
provided using types that have storage. After this change, all options
should be fully owned by a Pass' `Options` object after it has been
fully constructed, unless the underlying type of the `Option` explicitly
indicates otherwise.
Second, it updates the generated constructors for Passes to consume
options by value instead of reference, and prefers moving options into
the pass itself. This should be more efficient for non-trivial options
objects, where the previous interface forced a copy to be materialized.
Now, at worst case the API materializes a copy (no worse than before);
at best-case, all options objects are moved into place. Ideally, we
could update the Pass constructor to take an r-value reference to the
Options object instead, but this approach will require numerous changes
to existing passes and their factory functions.
---------
Authored-by: Nikhil Kalra <nkalra@apple.com>
There are some spurious libraries which can be removed.
I'm trying to bundle MLIR/LLVM library dependencies for our own
libraries. We're utilizing cmake function to recursively collect
MLIR/LLVM related dependencies. However, we identified certain library
dependencies as redundant and safe for removal.
Instead of hardcoding all fp smaller than 32 bits are unsupported we
provide a way to pass supported floating point types as well as the
target type. fp64 and fp32 are implicitly supported.
CC: @krzysz00 @manupak
This PR adds `f4E2M1FN` type to mlir.
`f4E2M1FN` type is proposed in [OpenCompute MX
Specification](https://www.opencompute.org/documents/ocp-microscaling-formats-mx-v1-0-spec-final-pdf).
It defines a 4-bit floating point number with bit layout S1E2M1. Unlike
IEEE-754 types, there are no infinity or NaN values.
```c
f4E2M1FN
- Exponent bias: 1
- Maximum stored exponent value: 3 (binary 11)
- Maximum unbiased exponent value: 3 - 1 = 2
- Minimum stored exponent value: 1 (binary 01)
- Minimum unbiased exponent value: 1 − 1 = 0
- Has Positive and Negative zero
- Doesn't have infinity
- Doesn't have NaNs
Additional details:
- Zeros (+/-): S.00.0
- Max normal number: S.11.1 = ±2^(2) x (1 + 0.5) = ±6.0
- Min normal number: S.01.0 = ±2^(0) = ±1.0
- Min subnormal number: S.00.1 = ±2^(0) x 0.5 = ±0.5
```
Related PRs:
- [PR-95392](https://github.com/llvm/llvm-project/pull/95392) [APFloat]
Add APFloat support for FP4 data type
- [PR-105573](https://github.com/llvm/llvm-project/pull/105573) [MLIR]
Add f6E3M2FN type - was used as a template for this PR
- [PR-107999](https://github.com/llvm/llvm-project/pull/107999) [MLIR]
Add f6E2M3FN type
This PR adds `f6E2M3FN` type to mlir.
`f6E2M3FN` type is proposed in [OpenCompute MX
Specification](https://www.opencompute.org/documents/ocp-microscaling-formats-mx-v1-0-spec-final-pdf).
It defines a 6-bit floating point number with bit layout S1E2M3. Unlike
IEEE-754 types, there are no infinity or NaN values.
```c
f6E2M3FN
- Exponent bias: 1
- Maximum stored exponent value: 3 (binary 11)
- Maximum unbiased exponent value: 3 - 1 = 2
- Minimum stored exponent value: 1 (binary 01)
- Minimum unbiased exponent value: 1 − 1 = 0
- Has Positive and Negative zero
- Doesn't have infinity
- Doesn't have NaNs
Additional details:
- Zeros (+/-): S.00.000
- Max normal number: S.11.111 = ±2^(2) x (1 + 0.875) = ±7.5
- Min normal number: S.01.000 = ±2^(0) = ±1.0
- Max subnormal number: S.00.111 = ±2^(0) x 0.875 = ±0.875
- Min subnormal number: S.00.001 = ±2^(0) x 0.125 = ±0.125
```
Related PRs:
- [PR-94735](https://github.com/llvm/llvm-project/pull/94735) [APFloat]
Add APFloat support for FP6 data types
- [PR-105573](https://github.com/llvm/llvm-project/pull/105573) [MLIR]
Add f6E3M2FN type - was used as a template for this PR
This PR adds `f6E3M2FN` type to mlir.
`f6E3M2FN` type is proposed in [OpenCompute MX
Specification](https://www.opencompute.org/documents/ocp-microscaling-formats-mx-v1-0-spec-final-pdf).
It defines a 6-bit floating point number with bit layout S1E3M2. Unlike
IEEE-754 types, there are no infinity or NaN values.
```c
f6E3M2FN
- Exponent bias: 3
- Maximum stored exponent value: 7 (binary 111)
- Maximum unbiased exponent value: 7 - 3 = 4
- Minimum stored exponent value: 1 (binary 001)
- Minimum unbiased exponent value: 1 − 3 = −2
- Has Positive and Negative zero
- Doesn't have infinity
- Doesn't have NaNs
Additional details:
- Zeros (+/-): S.000.00
- Max normal number: S.111.11 = ±2^(4) x (1 + 0.75) = ±28
- Min normal number: S.001.00 = ±2^(-2) = ±0.25
- Max subnormal number: S.000.11 = ±2^(-2) x 0.75 = ±0.1875
- Min subnormal number: S.000.01 = ±2^(-2) x 0.25 = ±0.0625
```
Related PRs:
- [PR-94735](https://github.com/llvm/llvm-project/pull/94735) [APFloat]
Add APFloat support for FP6 data types
- [PR-97118](https://github.com/llvm/llvm-project/pull/97118) [MLIR] Add
f8E4M3 type - was used as a template for this PR
This PR adds `f8E3M4` type to mlir.
`f8E3M4` type follows IEEE 754 convention
```c
f8E3M4 (IEEE 754)
- Exponent bias: 3
- Maximum stored exponent value: 6 (binary 110)
- Maximum unbiased exponent value: 6 - 3 = 3
- Minimum stored exponent value: 1 (binary 001)
- Minimum unbiased exponent value: 1 − 3 = −2
- Precision specifies the total number of bits used for the significand (mantissa),
including implicit leading integer bit = 4 + 1 = 5
- Follows IEEE 754 conventions for representation of special values
- Has Positive and Negative zero
- Has Positive and Negative infinity
- Has NaNs
Additional details:
- Max exp (unbiased): 3
- Min exp (unbiased): -2
- Infinities (+/-): S.111.0000
- Zeros (+/-): S.000.0000
- NaNs: S.111.{0,1}⁴ except S.111.0000
- Max normal number: S.110.1111 = +/-2^(6-3) x (1 + 15/16) = +/-2^3 x 31 x 2^(-4) = +/-15.5
- Min normal number: S.001.0000 = +/-2^(1-3) x (1 + 0) = +/-2^(-2)
- Max subnormal number: S.000.1111 = +/-2^(-2) x 15/16 = +/-2^(-2) x 15 x 2^(-4) = +/-15 x 2^(-6)
- Min subnormal number: S.000.0001 = +/-2^(-2) x 1/16 = +/-2^(-2) x 2^(-4) = +/-2^(-6)
```
Related PRs:
- [PR-99698](https://github.com/llvm/llvm-project/pull/99698) [APFloat]
Add support for f8E3M4 IEEE 754 type
- [PR-97118](https://github.com/llvm/llvm-project/pull/97118) [MLIR] Add
f8E4M3 IEEE 754 type
This PR adds `f8E4M3` type to mlir.
`f8E4M3` type follows IEEE 754 convention
```c
f8E4M3 (IEEE 754)
- Exponent bias: 7
- Maximum stored exponent value: 14 (binary 1110)
- Maximum unbiased exponent value: 14 - 7 = 7
- Minimum stored exponent value: 1 (binary 0001)
- Minimum unbiased exponent value: 1 − 7 = −6
- Precision specifies the total number of bits used for the significand (mantisa),
including implicit leading integer bit = 3 + 1 = 4
- Follows IEEE 754 conventions for representation of special values
- Has Positive and Negative zero
- Has Positive and Negative infinity
- Has NaNs
Additional details:
- Max exp (unbiased): 7
- Min exp (unbiased): -6
- Infinities (+/-): S.1111.000
- Zeros (+/-): S.0000.000
- NaNs: S.1111.{001, 010, 011, 100, 101, 110, 111}
- Max normal number: S.1110.111 = +/-2^(7) x (1 + 0.875) = +/-240
- Min normal number: S.0001.000 = +/-2^(-6)
- Max subnormal number: S.0000.111 = +/-2^(-6) x 0.875 = +/-2^(-9) x 7
- Min subnormal number: S.0000.001 = +/-2^(-6) x 0.125 = +/-2^(-9)
```
Related PRs:
- [PR-97179](https://github.com/llvm/llvm-project/pull/97179) [APFloat]
Add support for f8E4M3 IEEE 754 type
Add an `fastMathAttr` on `arith::extf` and `arith::truncf`. If these two
ops are inserted by some promotion passes (like legalize-to-f32 /
emulate-unsupported-floats), they will be labeled as
`FastMathFlags::contract`, denoting that they can be then `eliminated by
canonicalizer`.
The `elimination` can help improve performance, while may introduce some
numerical differences.
Add an `fastMathAttr` on `arith::extf` and `arith::truncf`. If these two
ops are inserted by some promotion passes (like legalize-to-f32 /
emulate-unsupported-floats), they will be labeled as
`FastMathFlags::contract`, denoting that they can be then `eliminated by
canonicalizer`.
The `elimination` can help improve performance, while may introduce some
numerical differences.
When the integer range analysis was first develop, a pass that did
integer range-based constant folding was developed and used as a test
pass. There was an intent to add such a folding to SCCP, but that hasn't
happened.
Meanwhile, -int-range-optimizations was added to the arith dialect's
transformations. The cmpi simplification in that pass is a strict subset
of the constant folding that lived in
-test-int-range-inference.
This commit moves the former test pass into -int-range-optimizaitons,
subsuming its previous contents. It also adds an optimization from
rocMLIR where `rem{s,u}i` operations that are noops are replaced by
their left operands.
These passes have been depreciated for a long time and replaced by
one-shot bufferization. These passes are also unsafe because they do not
check for read-after-write conflicts.
Relands https://github.com/llvm/llvm-project/pull/93488 which failed on
buildbot. Fixes the failure by updating integration tests to use
one-shot-bufferize instead.
These passes have been depreciated for a long time and replaced by
one-shot bufferization. These passes are also unsafe because they do not
check for read-after-write conflicts.
The pass runs a `DataFlowSolver` and collects state information on the
input IR. Then, the rewrite driver and folding is applied. During
pattern application and folding it can happen that an Op from the input
IR is deleted and a new Op is created at the same address. When the
newly created Ops is looked up in the `DataFlowSolver` state memory, the
state of the original Op is returned.
This patch adds a method to `DataFlowSolver` which removes all state
related to a `ProgramPoint`. It also adds a listener to the Pass which
clears the state information of deleted Ops from the `DataFlowSolver`.
Fix https://github.com/llvm/llvm-project/issues/81228
Expand `arith.minsi`, `arith.minui`, `arith.maxsi`, `arith.maxui` into
`arith.cmpi` and `arith.select`.
---------
Co-authored-by: Jakub Kuderski <kubakuderski@gmail.com>
This commit generalizes and cleans up the `ValueBoundsConstraintSet`
API. The API used to provide function overloads for comparing/computing
bounds of:
- index-typed SSA value
- dimension of shaped value
- affine map + operands
This commit removes all overloads. There is now a single entry point for
each `compare` variant and each `computeBound` variant. These functions
now take a `Variable`, which is internally represented as an affine map
and map operands.
This commit also adds support for computing bounds for an affine map +
operands. There was previously no public API for that.
This commit changes the API of `ValueBoundsConstraintSet`: the stop
condition is now passed to the constructor instead of `processWorklist`.
That makes it easier to add items to the worklist multiple times and
process them in a consistent manner. The current
`ValueBoundsConstraintSet` is passed as a reference to the stop
function, so that the stop function can be defined before the the
`ValueBoundsConstraintSet` is constructed.
This change is in preparation of adding support for branches.
`CeilDivUIOp` seemed to have been added by mistake to the list of
dynamically
illegal operations in `arith-unsigned-when-equivalent`. The only illegal
operations
should be the signed operations that can be converted to their unsigned
counterpart.
Add rounding mode attribute to `arith`. This attribute can be used in
different FP `arith` operations to control rounding mode. Rounding modes
correspond to IEEE 754-specified rounding modes. Use in `arith.truncf` folding.
As this is not supported in dialects other than LLVM, conversion should fail for
now in case this attribute is present.
---------
Signed-off-by: Victor Perez <victor.perez@codeplay.com>
This commit adds the `BufferViewFlowOpInterface` to the bufferization
dialect. This interface can be implemented by ops that operate on
buffers to indicate that a buffer op result and/or region entry block
argument may be the same buffer as a buffer operand (or a view thereof).
This interface is queried by the `BufferViewFlowAnalysis`.
The new interface has two interface methods:
* `populateDependencies`: Implementations use the provided callback to
declare dependencies between operands and op results/region entry block
arguments. E.g., for `%r = arith.select %c, %m1, %m2 : memref<5xf32>`,
the interface implementation should declare two dependencies: %m1 -> %r
and %m2 -> %r.
* `mayBeTerminalBuffer`: An SSA value is a terminal buffer if the buffer
view flow analysis stops at the specified value. E.g., because the value
is a newly allocated buffer or because no further information is
available about the origin of the buffer.
Ops that implement the `RegionBranchOpInterface` or `BranchOpInterface`
do not have to implement the `BufferViewFlowOpInterface`. The buffer
dependencies can be inferred from those two interfaces.
This commit makes the `BufferViewFlowAnalysis` more accurate. For
unknown ops, it conservatively used to declare all combinations of
operands and op results/region entry block arguments as dependencies
(false positives). This is no longer the case. While the analysis is
still a "maybe" analysis with false positives (e.g., when analyzing ops
such as `arith.select` or `scf.if` where the taken branch is not known
at compile time), results and region entry block arguments of unknown
ops are now marked as terminal buffers.
This commit addresses a TODO in `BufferViewFlowAnalysis.cpp`:
```
// TODO: We should have an op interface instead of a hard-coded list of
// interfaces/ops.
```
It is no longer needed to hard-code ops.
This lowering was not correctly handling the case where saturation of
the mantissa results in an increase of the exponent value. The new code
borrows, with credit, the idea from
e1502c0cdb/c10/util/BFloat16.h (L60-L79)
and adds comments to explain the magic trick going on here and why it's
correct. Hat tip to its original author, whom I believe to be
@Maratyszcza.
A testcase was also requiring a tie to be broken upwards in a case where
"to nearest-even" required going downward. The fact that it used to pass
suggests that there was another bug in the old code.
Collection of changes with the goal of being able to convert `encoding`
to `memorySpace` during bufferization
- new API for encoder to allow implementation to select destination
memory space
- update existing bufferization implementations to support the new
interface
Many machine-learning applications (and most software written at AMD)
expect the operation that truncates floats to 8-bit floats to be
saturatinng. That is, they expect `truncf 256.0 : f32 to f8E4M3FNUZ` to
yield `240.0`, not `NaN`, and similarly for negative numbers. However,
the underlying hardware instruction that can be used for this truncation
implements overflow-to-NaN semantics.
To enable handling this usecase, we add the saturate-fp8-truncf option
to ArithToAMDGPU (off by default), which causes the requisite clamping
code to be emitted. Said clamping code ensures that Inf and NaN are
passed through exactly (and thus trancate to NaN).
Per review feedback, this commit efactors
createScalarOrSplatConstant() to the Arith dialect utilities and uses
it in this code. It also fixes naming of existing patterns and
switches from vector.extractelement/insertelement to
vector.extract/insert.
The maxnum/minnum semantics can be found at
https://llvm.org/docs/LangRef.html#llvm-minnum-intrinsic.
The revision also updates function names in lit tests to match op name.
Take arith.maxnumf as example:
```
func.func @maxnumf(%lhs: f32, %rhs: f32) -> f32 {
%result = arith.maxnumf %lhs, %rhs : f32
return %result : f32
}
```
will be expanded to
```
func.func @maxnumf(%lhs: f32, %rhs: f32) -> f32 {
%0 = arith.cmpf ugt, %lhs, %rhs : f32
%1 = arith.select %0, %lhs, %rhs : f32
%2 = arith.cmpf uno, %lhs, %lhs : f32
%3 = arith.select %2, %rhs, %1 : f32
return %3 : f32
}
```
Case 1: Both LHS and RHS are not NaN; LHS > RHS
In this case, `%1` is LHS. `%3` and `%1` have the same value, so `%3` is
LHS.
Case 2: LHS is NaN and RHS is not NaN
In this case, `%2` is true, so `%3` is always RHS.
Case 3: LHS is not NaN and RHS is NaN
In this case, `%0` is true and `%1` is LHS. `%2` is false, so `%3` and
`%1` have the same value, which is LHS.
Case 4: Both LHS and RHS are NaN:
`%1` and RHS are all NaN, so the result is still NaN.
* Declare arguments/results with `let` statements.
* Rename `transp` to `permutation`.
* Change type of `transp` from `I64ArrayAttr` to `DenseI64ArrayAttr`
(provides direct access to `ArrayRef<int64_t>` instead of `ArrayAttr`).
Note that the `Pass` suffix is added in tablegen, and as a side effect the
options are renamed from `ArithExpandOpsOptions` to `ArithExpandOpsPassOptions`.
Inserting clones requires a lot of assumptions to hold on the input IR, e.g., all writes to a buffer need to dominate all reads. This is not guaranteed by one-shot bufferization and isn't easy to verify, thus it could quickly lead to incorrect results that are hard to debug. This commit changes the mechanism of how an ownership indicator is materialized when there is not already a unique ownership present. Additionally, we don't create copies of returned memrefs anymore when we don't have ownership. Instead, we insert assert operations to make sure we have ownership at runtime, or otherwise report to the user that correctness could not be guaranteed.
Add a method to the BufferDeallocationOpInterface that allows operations to implement the interface and provide custom logic to compute the ownership indicators of values it defines. As a demonstrating example, this new method is implemented by the `arith.select` operation.
