Skip to content

Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store"#40033

Merged
rschu1ze merged 3 commits intomasterfrom
reenable-avx512-vbmi-columnvector-filter
Aug 15, 2022
Merged

Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store"#40033
rschu1ze merged 3 commits intomasterfrom
reenable-avx512-vbmi-columnvector-filter

Conversation

@rschu1ze
Copy link
Copy Markdown
Member

@rschu1ze rschu1ze commented Aug 9, 2022

This PR re-enables #39633 after issue #39895 has been detected and the PR was reverted (#39963). It contains the revert's revert plus a fix for issue #39895.

Changelog category (leave one):

  • Performance Improvement

Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):

For systems with AVX512 VBMI2, this PR improves performance by ca. 6% for SSB benchmark queries queries 3.1, 3.2 and 3.3 (SF=100). Tested on Intel Icelake Xeon 8380 * 2 socket.

Courtesy to @guowangy

@robot-clickhouse robot-clickhouse added the pr-performance Pull request with some performance improvements label Aug 9, 2022
@rschu1ze rschu1ze force-pushed the reenable-avx512-vbmi-columnvector-filter branch from d996662 to 8a7df6f Compare August 9, 2022 17:24
@rschu1ze
Copy link
Copy Markdown
Member Author

Test failures are unrelated.

@rschu1ze rschu1ze changed the title ColumnVector: optimize filter with AVX512 VBMI2 compress store Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store" Aug 11, 2022
@nickitat nickitat self-assigned this Aug 15, 2022
@rschu1ze rschu1ze merged commit 70f63d2 into master Aug 15, 2022
@rschu1ze rschu1ze deleted the reenable-avx512-vbmi-columnvector-filter branch August 15, 2022 16:39
alexey-milovidov added a commit that referenced this pull request Sep 25, 2022
…i-columnvector-filter"

This reverts commit 70f63d2, reversing
changes made to a0693c3.
alexey-milovidov added a commit that referenced this pull request Sep 26, 2022
Revert of "Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store" #40033"
robot-clickhouse pushed a commit that referenced this pull request Sep 26, 2022
…r: optimize filter with AVX512 VBMI2 compress store" #40033"
robot-clickhouse pushed a commit that referenced this pull request Sep 26, 2022
…r: optimize filter with AVX512 VBMI2 compress store" #40033"
alexey-milovidov added a commit that referenced this pull request Sep 27, 2022
Backport #41752 to 22.8: Revert of "Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store" #40033"
alexey-milovidov added a commit that referenced this pull request Sep 27, 2022
Backport #41752 to 22.9: Revert of "Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store" #40033"
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

pr-performance Pull request with some performance improvements

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants