CE Performance Automation : step 1 of 2 (build) DONE.

This comment was automatically generated given a benchmark was triggered.
Started building at 2024-11-08 06:59:38.746304 and took 64 seconds.
You can check each build/benchmark progress in grafana:

• git hash: 986c4e4
• git branch: Nugine:hll-simd
• commit date and time: n/a
• commit summary: n/a
• test filters:
  • command priority lower limit: 0
  • command priority upper limit: 10000
  • test name regex: .*
  • command group regex: .*

You can check a comparison in detail via the grafana link

CE Performance Automation : step 2 of 2 (benchmark) FINISHED.

This comment was automatically generated given a benchmark was triggered.

Started benchmark suite at 2024-11-13 19:19:28.887445 and took 48079.879437 seconds to finish.
Status: [################################################################################] 100.0% completed.

In total will run 141 benchmarks.
- 0 pending.
- 141 completed:
- 141 successful.
- 0 failed.
You can check a the status in detail via the grafana link

Automated performance analysis summary

This comment was automatically generated given there is performance data available.

Using platform named: intel64-ubuntu22.04-redis-icx1 to do the comparison.

In summary:

• Detected a total of 264 stable tests between versions.
• Detected a total of 8 highly unstable benchmarks.
• Detected a total of 1 improvements above the improvement water line.
• Detected a total of 4 regressions bellow the regression water line 10.0.
  • Median/Common-Case regression was -14.0% and ranged from [-15.0%,-12.0%].

You can check a comparison in detail via the grafana link

Comparison between unstable and Nugine:hll-simd.

Time Period from 5 months ago. (environment used: oss-standalone)

Unstable Table

Unstable test regexp names: memtier_benchmark-10Kkeys-load-hash-50-fields-with-10000B-values|memtier_benchmark-1Mkeys-generic-ttl-pipeline-10|memtier_benchmark-1Mkeys-load-zset-listpack-with-100-elements-double-score|memtier_benchmark-1Mkeys-string-incrbyfloat-pipeline-10|memtier_benchmark-1Mkeys-string-mget-1KiB|memtier_benchmark-1key-zset-1K-elements-zrange-all-elements|memtier_benchmark-2keys-set-10-100-elements-sunion|memtier_benchmark-3Mkeys-string-mixed-20-80-with-512B-values-pipeline-10-5200_conns

Regressions Table

Regressions test regexp names: memtier_benchmark-1key-list-10K-elements-linsert-lrem-integer|memtier_benchmark-1key-list-10K-elements-linsert-lrem-string|memtier_benchmark-1key-list-10K-elements-lpos-integer|memtier_benchmark-1key-list-10K-elements-lpos-string

Improvements Table

Improvements test regexp names: memtier_benchmark-1Mkeys-10B-expire-use-case

WARNING: There were 13 benchmarks with NO datapoints for both baseline and comparison.

NO DATAPOINTS test regexp names: m|m|m|m|m|m|m|m|m|r|r|r|r

currently this PR looks good to me, please also have a look @sundb @fcostaoliveira @YaacovHazan

fully CI on arm64: https://github.com/redis/redis-extra-ci/actions/runs/11586127029
fully CI: https://github.com/sundb/redis/actions/runs/11586131243

@Nugine merged, thank you

Hi @ShooterIT @sundb
I'd like to submit this performance improvement to valkey. But I'm not sure what to do with the co-authored parts considering the Redis licenses. https://github.com/orgs/valkey-io/discussions/1286
What do you think?

Hi @ShooterIT @sundb
I'd like to submit this performance improvement to valkey. But I'm not sure what to do with the co-authored parts considering the Redis licenses. https://github.com/orgs/valkey-io/discussions/1286
What do you think?

@Nugine you have the right to apply this PR to valkey.

Note that after merging this PR, building the code with clang 18 gives the following error -

In file included from /usr/lib/llvm-18/lib/clang/18/include/immintrin.h:26:
In file included from /usr/lib/llvm-18/lib/clang/18/include/xmmintrin.h:31:
/usr/lib/llvm-18/lib/clang/18/include/mm_malloc.h:35:12: error: 'malloc' is deprecated [-Werror,-Wdeprecated-declarations]
   35 |     return malloc(__size);
      |            ^
./server.h:3873:43: note: 'malloc' has been explicitly marked deprecated here
 3873 | void *malloc(size_t size) __attribute__ ((deprecated));

@alonre24 Have you seen it on github action? This complaint is not caused by this, but because ubuntu-latest is using 24.04 in the ci of some users, I want to wait until this repository ubuntu is upgraded before fixing it.

you can try locally with ubuntu 24.04:
CC=clang make

@sundb not only in github actions, I reproduced it locally with ubuntu 22. From what I see, it happens when using clang-18 (older clang are ok).
I believe that #include <immintrin.h> might causing this, since it assumes GCC compiler. Perhaps we can use the following logic instead:

#include <x86intrin.h>
#elif defined(__clang__)
#include <xmmintrin.h>
#elif defined(_MSC_VER)
#include <intrin.h>
#include <stdexcept>
#endif

?

cool, @alonre24 could you make a PR for this issue?

@alonre24 yeah, you're right.
I found that the complaint was caused by #include <mm_malloc.h> in <xmmintrin.h>, which caused the compiler to mistakenly assume that we were using malloc and free, which were set to deprecated in server.h.

one of the alternative is forbiding attribute deprecated for clang.

#if defined(__GNUC__) && !defined(__clang__)
void *calloc(size_t count, size_t size) __attribute__ ((deprecated));
void free(void *ptr) __attribute__ ((deprecated));
void *malloc(size_t size) __attribute__ ((deprecated));
void *realloc(void *ptr, size_t size) __attribute__ ((deprecated));
#endif