[hist] Used fixed-width SIMD type for interpreted code in TFormula #20769
Conversation
This file contains hidden or 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
Test Results 22 files 22 suites 3d 20h 22m 5s ⏱️ For more details on these failures, see this check. Results for commit 096f2df. ♻️ This comment has been updated with latest results. |
guitargeek
force-pushed
the
veccore_jira_issue
branch
from
d0d5941 to
d82921e
Compare
dpiparo
approved these changes
Dec 24, 2025
guitargeek
force-pushed
the
veccore_jira_issue
branch
from
d82921e to
8ee2709
Compare
In the interpreter, we must match the SIMD width used by compiled ROOT. `ROOT::Double_v` aliases the best available native SIMD type, which may differ between compiled and interpreted contexts (e.g. with `-march=native`). Therefore, we explicitly select the fixed-size SIMD type corresponding to the native SIMD width used in compiled ROOT. This fixes failues in the vectorized TFormula tests for platforms with possible mismatch between the SIMD types between interpretd and compiled context (e.g. vanilla Alma 10 or Alma 9 with `march=native` in the CI). Also, disable some subtests in the vectorized TFormula tests if they'd still fail becaues of missing Vc symbols.
This is to show that issues related to VecCore and Vc are fixed, and to provide much-needed test coverage for a gradual migration from VecCore and Vc to `std::experimental::simd`.
Using the subset of the Vc API that is also part of `std::experimental::simd` makes it possible to transparently use `std::experimental::simd` as a GenVector template parameter.
guitargeek
force-pushed
the
veccore_jira_issue
branch
from
8ee2709 to
096f2df
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
In the interpreter, we must match the SIMD width used by compiled ROOT.
ROOT::Double_valiases the best available native SIMD type, which may differ between compiled and interpreted contexts (e.g. with-march=native). Therefore, we explicitly select the fixed-size SIMD type corresponding to the native SIMD width used in compiled ROOT.This fixes failues in the vectorized TFormula tests for platforms with possible mismatch between the SIMD types between interpretd and compiled context (e.g. vanilla Alma 10 or Alma 9 with
march=nativein the CI).A separate commit also enables VecCore and Vc for the
-march=nativebuild in the CI, so we get test coverage.Note that this fix is not specific to VecCore or Vc: the same needs to be done if we were using
std::experimental::simdorstd::simd.