[ARITH] Enhance Canonical Simplify for LE #15471
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This PR enhances the canonical simplifier to support the following patterns:
x0 * s0 + x1 * s1 + ... + xn + c < 0, let d = gcd(s0, s1, ..., s{n-1}, c)
1. if can prove -d < xn < d, then we can simplify
the expression to x0 * (s0/d) + x1 * (s1/d) + ... + x{n-1} * (s{n-1}/d) < c/d,
e.g. `x * 8 + y < 16` where `y` \in [0, 8), we can simplify it to `x < 2`
2. if xn is in pattern of yn % m, where m % d == 0, convert it to yn // d % (m/d)
e.g. `x1 * 64 + (x2 * 8 + x3) % 64 < 120`, `x3` \in [0, 8), we can simplify it to
`x1 * 8 + (x2 * 8 + x3) // 8 % 8 < 15` ==> `x1 * 8 + x2 % 8 < 15`
|
Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.
Generated by tvm-bot |
Hzfengsy
commented
Aug 3, 2023
| @@ -64,7 +64,8 @@ def test_add_pipeline(): | |||
| s[C].pragma(xo1, "parallel_launch_point") | |||
| s[C].pragma(xo2, "parallel_stride_pattern") | |||
| s[C].pragma(xo2, "parallel_barrier_when_finish") | |||
| s[C].vectorize(xi) | |||
| # FIXME(tvm-team): vector operators are not supported for codegen to C yet | |||
There was a problem hiding this comment.
This case is because of the poor support of vectorization for c_codegen.
Before this PR, TVM failed to simplify the branch predicate and skip the vectorization. This PR enhances the arith, and makes it TRUE vectoring, but failed on codegen stage.
In short:
- It's not related to this PR, as it's codegen issue
- It's not a regression, vectorized step is skipped before this PR.
tqchen
approved these changes
Aug 8, 2023
MasterJH5574
added a commit
to MasterJH5574/tvm
that referenced
this pull request
Mar 12, 2024
PR apache#15471 enhances the simplification for LE, while missed a case where the upper bound `kPosInf` is divisible by a factor. Therefore, prior to this PR, when simplifying `x * 1024 + y < z * 7168`, it will fails with the error message ``` InternalError: Check failed: value < 1LL << (dtype.bits() - 1) (8589934591 vs. 2147483648) : ValueError: Literal value 8589934591 exceeds maximum of int32 ``` This is just because the upper bound 7 here divides `kPosInf` the maximum value of int64, which passes an "if" condition in apache#15471 unexpectedly. This PR fixes the issue.
MasterJH5574
added a commit
to MasterJH5574/tvm
that referenced
this pull request
Mar 12, 2024
PR apache#15471 enhances the simplification for LE, while missed a case where the upper bound `kPosInf` is divisible by a factor. Therefore, prior to this PR, when simplifying `x * 1024 + y < z * 7168`, it will fails with the error message ``` InternalError: Check failed: value < 1LL << (dtype.bits() - 1) (8589934591 vs. 2147483648) : ValueError: Literal value 8589934591 exceeds maximum of int32 ``` This is just because the upper bound 7 here divides `kPosInf` the maximum value of int64, which passes an "if" condition in apache#15471 unexpectedly. This PR fixes the issue.
MasterJH5574
added a commit
to MasterJH5574/tvm
that referenced
this pull request
Mar 12, 2024
PR apache#15471 enhances the simplification for LE, while missed a case where the upper bound `kPosInf` is divisible by a factor. Therefore, prior to this PR, when simplifying `x * 1024 + y < z * 7168`, it will fails with the error message ``` InternalError: Check failed: value < 1LL << (dtype.bits() - 1) (8589934591 vs. 2147483648) : ValueError: Literal value 8589934591 exceeds maximum of int32 ``` This is just because the upper bound 7 here divides `kPosInf` the maximum value of int64, which passes an "if" condition in apache#15471 unexpectedly. This PR fixes the issue.
tqchen
pushed a commit
that referenced
this pull request
Mar 12, 2024
PR #15471 enhances the simplification for LE, while missed a case where the upper bound `kPosInf` is divisible by a factor. Therefore, prior to this PR, when simplifying `x * 1024 + y < z * 7168`, it will fails with the error message ``` InternalError: Check failed: value < 1LL << (dtype.bits() - 1) (8589934591 vs. 2147483648) : ValueError: Literal value 8589934591 exceeds maximum of int32 ``` This is just because the upper bound 7 here divides `kPosInf` the maximum value of int64, which passes an "if" condition in #15471 unexpectedly. This PR fixes the issue.
thaisacs
pushed a commit
to thaisacs/tvm
that referenced
this pull request
Apr 3, 2024
PR apache#15471 enhances the simplification for LE, while missed a case where the upper bound `kPosInf` is divisible by a factor. Therefore, prior to this PR, when simplifying `x * 1024 + y < z * 7168`, it will fails with the error message ``` InternalError: Check failed: value < 1LL << (dtype.bits() - 1) (8589934591 vs. 2147483648) : ValueError: Literal value 8589934591 exceeds maximum of int32 ``` This is just because the upper bound 7 here divides `kPosInf` the maximum value of int64, which passes an "if" condition in apache#15471 unexpectedly. This PR fixes the issue.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR enhances the canonical simplifier to support the following patterns:
x0 * s0 + x1 * s1 + ... + xn + c < 0, let d = gcd(s0, s1, ..., s{n-1}, c)x0 * (s0/d) + x1 * (s1/d) + ... + x{n-1} * (s{n-1}/d) < c/d, e.g.x * 8 + y < 16wherey\in [0, 8), we can simplify it tox < 2x1 * 64 + (x2 * 8 + x3) % 64 < 120,x3\in [0, 8), we can simplify it tox1 * 8 + (x2 * 8 + x3) // 8 % 8 < 15==>x1 * 8 + x2 % 8 < 15cc @tqchen