-
Notifications
You must be signed in to change notification settings - Fork 775
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
v0.8.2 #873
Commits on Dec 23, 2021
-
Introduce XXH_SIZE_OPT and XXH_NO_STREAM
- XXH_SIZE_OPT is a value from 0-2 which indicates how much xxHash should care about code size, default 1 for -Os/-Oz, default 0 otherwise - XXH_NO_STREAM disables the streaming API. - These two interact: if XXH_SIZE_OPT == 2 and XXH_NO_STREAM is not defined, XXH32 and XXH64 use the streaming API for single shot - TODO: apply this to XXH3 as well
Configuration menu - View commit details
-
Copy full SHA for 6f97097 - Browse repository at this point
Copy the full SHA 6f97097View commit details
Commits on Dec 26, 2021
-
do no longer depend on
<assert.h>
for XXH_STATIC_ASSERTsince some versions are buggy. Use `_Static_assert` instead, which is part of the C11 language.
Configuration menu - View commit details
-
Copy full SHA for 6189ecd - Browse repository at this point
Copy the full SHA 6189ecdView commit details
Commits on Dec 29, 2021
-
Merge pull request #670 from Cyan4973/fix664
do no longer depend on `<assert.h>` for XXH_STATIC_ASSERT
Configuration menu - View commit details
-
Copy full SHA for d878e1d - Browse repository at this point
Copy the full SHA d878e1dView commit details -
[sizeopt] Reroll midrange on size opt, disable prefetch
It seems that using the streaming API is not beneficial on GCC so it isn't worth the complexity.
Configuration menu - View commit details
-
Copy full SHA for fdf28f6 - Browse repository at this point
Copy the full SHA fdf28f6View commit details
Commits on Dec 30, 2021
-
Properly fix altivec namespace collisions
Uses `#pragma push_macro`/`#pragma pop_macro` instead of assuming that `__APPLE_ALTIVEC__` can be defined safely. Safer and less hacky fix for #668.
Configuration menu - View commit details
-
Copy full SHA for ec4030d - Browse repository at this point
Copy the full SHA ec4030dView commit details -
Configuration menu - View commit details
-
Copy full SHA for 4690bbe - Browse repository at this point
Copy the full SHA 4690bbeView commit details
Commits on Jan 6, 2022
-
Merge pull request #672 from easyaspi314/vsx_hotfix
Properly fix altivec namespace collisions
Configuration menu - View commit details
-
Copy full SHA for 83b2235 - Browse repository at this point
Copy the full SHA 83b2235View commit details -
Merge pull request #667 from easyaspi314/sizeopt
Introduce XXH_SIZE_OPT and XXH_NO_STREAM
Configuration menu - View commit details
-
Copy full SHA for b1a61df - Browse repository at this point
Copy the full SHA b1a61dfView commit details
Commits on Jan 18, 2022
-
Configuration menu - View commit details
-
Copy full SHA for ab6699a - Browse repository at this point
Copy the full SHA ab6699aView commit details
Commits on Jan 19, 2022
-
Merge pull request #676 from mterron/patch-1
Remove duplicated definition of XXH3_128bits
Configuration menu - View commit details
-
Copy full SHA for d8aff14 - Browse repository at this point
Copy the full SHA d8aff14View commit details
Commits on Mar 9, 2022
-
3
Configuration menu - View commit details
-
Copy full SHA for 81972bf - Browse repository at this point
Copy the full SHA 81972bfView commit details
Commits on Mar 10, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 7b3bb44 - Browse repository at this point
Copy the full SHA 7b3bb44View commit details -
Configuration menu - View commit details
-
Copy full SHA for a0f750f - Browse repository at this point
Copy the full SHA a0f750fView commit details -
Configuration menu - View commit details
-
Copy full SHA for 0c0ea60 - Browse repository at this point
Copy the full SHA 0c0ea60View commit details -
Merge pull request #690 from t-mat/fix-issue-689
Removed windows-2016 from ci.yml
Configuration menu - View commit details
-
Copy full SHA for 2ae055b - Browse repository at this point
Copy the full SHA 2ae055bView commit details -
Configuration menu - View commit details
-
Copy full SHA for 4c223a5 - Browse repository at this point
Copy the full SHA 4c223a5View commit details -
Configuration menu - View commit details
-
Copy full SHA for aa7ac8b - Browse repository at this point
Copy the full SHA aa7ac8bView commit details
Commits on Mar 29, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 8b86185 - Browse repository at this point
Copy the full SHA 8b86185View commit details
Commits on Apr 4, 2022
-
Introduce md5sum style filename escape convention
For details, see issue #695. If filename contains special character (\n, 0x0a, LF), - Put '\' (0x5c) at the beginning of the line. - Escape special character by '\'.
Configuration menu - View commit details
-
Copy full SHA for bd4ded9 - Browse repository at this point
Copy the full SHA bd4ded9View commit details -
Configuration menu - View commit details
-
Copy full SHA for f1f2ff9 - Browse repository at this point
Copy the full SHA f1f2ff9View commit details -
Configuration menu - View commit details
-
Copy full SHA for b51ffce - Browse repository at this point
Copy the full SHA b51ffceView commit details
Commits on Apr 6, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 9244d6e - Browse repository at this point
Copy the full SHA 9244d6eView commit details -
Configuration menu - View commit details
-
Copy full SHA for bcc4c6a - Browse repository at this point
Copy the full SHA bcc4c6aView commit details
Commits on May 4, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 4866da2 - Browse repository at this point
Copy the full SHA 4866da2View commit details -
Configuration menu - View commit details
-
Copy full SHA for 93dcc6f - Browse repository at this point
Copy the full SHA 93dcc6fView commit details
Commits on May 6, 2022
-
Merge pull request #708 from MaxiBoether/dev
Use inline assembler for Power/IBM XL Compiler
Configuration menu - View commit details
-
Copy full SHA for cc8f0d0 - Browse repository at this point
Copy the full SHA cc8f0d0View commit details -
Configuration menu - View commit details
-
Copy full SHA for d278c2e - Browse repository at this point
Copy the full SHA d278c2eView commit details -
Merge pull request #710 from Cyan4973/test_escape
test filename-escape
Configuration menu - View commit details
-
Copy full SHA for 35ff826 - Browse repository at this point
Copy the full SHA 35ff826View commit details -
Configuration menu - View commit details
-
Copy full SHA for 4692d75 - Browse repository at this point
Copy the full SHA 4692d75View commit details
Commits on May 7, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 0504556 - Browse repository at this point
Copy the full SHA 0504556View commit details -
Configuration menu - View commit details
-
Copy full SHA for bfd20ae - Browse repository at this point
Copy the full SHA bfd20aeView commit details -
Merge pull request #711 from Cyan4973/fix682
avoid add_compile_definitions for cmake < v3.12
Configuration menu - View commit details
-
Copy full SHA for f852875 - Browse repository at this point
Copy the full SHA f852875View commit details
Commits on May 24, 2022
-
added XXH_RESTRICT to XXH3_hashLong_64b_withSeed()
for closer correspondence with XXH3_hashLong64_f definition
Configuration menu - View commit details
-
Copy full SHA for 6078dd6 - Browse repository at this point
Copy the full SHA 6078dd6View commit details
Commits on Jun 7, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 7e9ada3 - Browse repository at this point
Copy the full SHA 7e9ada3View commit details
Commits on Jun 8, 2022
-
Signed-off-by: Ghasem Ramezani (Work) <gh.ramazani@arad.co.net>
Ghasem Ramezani committedJun 8, 2022 Configuration menu - View commit details
-
Copy full SHA for ba96407 - Browse repository at this point
Copy the full SHA ba96407View commit details -
Configuration menu - View commit details
-
Copy full SHA for cf73b43 - Browse repository at this point
Copy the full SHA cf73b43View commit details -
Merge pull request #721 from Cyan4973/moreCmake2812tests
just more cmake v2.8.12 tests
Configuration menu - View commit details
-
Copy full SHA for b966a6f - Browse repository at this point
Copy the full SHA b966a6fView commit details -
Configuration menu - View commit details
-
Copy full SHA for 6932421 - Browse repository at this point
Copy the full SHA 6932421View commit details -
Configuration menu - View commit details
-
Copy full SHA for 6a1bda6 - Browse repository at this point
Copy the full SHA 6a1bda6View commit details
Commits on Jul 9, 2022
-
Remove stream loads and slightly improve avx512 seed generation
Streamed loads (`vmovntdqa`) don't provide any benefit over normal aligned loads (`vmovdqa{8|16|32|64}`) but prevent GCC/Clang from micro-fusing the loads with the ALU wasting code size. Replace lane cross `vpbroadcastq` with `vpsubq`
Configuration menu - View commit details
-
Copy full SHA for acfcfa9 - Browse repository at this point
Copy the full SHA acfcfa9View commit details
Commits on Jul 10, 2022
-
Merge pull request #726 from goldsteinn/dev
Remove stream loads and slightly improve avx512 seed generation
Configuration menu - View commit details
-
Copy full SHA for f7ab847 - Browse repository at this point
Copy the full SHA f7ab847View commit details
Commits on Jul 29, 2022
-
Since GNU make invokes `shell` by default, it doesn't support brace expansion. This changeset expands the braces to fix the issue.
Configuration menu - View commit details
-
Copy full SHA for 051276a - Browse repository at this point
Copy the full SHA 051276aView commit details -
Introduce new enum AlgoBitmask to fix #724.
main() introduces new variable algoBitmask which represents the algorithm(s) which will be accepted by the CLI. main() also sets suitable algoBitmask for exename and passes it to XSUM_checkFiles(). The default algoBitmask is algo_bitmask_all which accepts all xxh algorithms. XSUM_checkFiles() passes algoBitmask to XSUM_checkFile(). XSUM_checkFile() sets algoBitmask to XSUM_parseFileArg->algoBitmask and invokes XSUM_parseFile1(). XSUM_parseFile1() passes algoBitmask to XSUM_parseLine(). XSUM_parseLine() passes algoBitmask to XSUM_parseLine1(). XSUM_parseLine1() calles XSUM_algoBitmask_Accepts() to check actual xxh algorithm of the line and algoBitmask matches or not.
Configuration menu - View commit details
-
Copy full SHA for 02dc2ab - Browse repository at this point
Copy the full SHA 02dc2abView commit details -
Configuration menu - View commit details
-
Copy full SHA for 06061c9 - Browse repository at this point
Copy the full SHA 06061c9View commit details -
Configuration menu - View commit details
-
Copy full SHA for 7a9ff01 - Browse repository at this point
Copy the full SHA 7a9ff01View commit details -
Merge pull request #729 from t-mat/fix-makefile-glob
Fix: brace expansion
Configuration menu - View commit details
-
Copy full SHA for 1dc8eb9 - Browse repository at this point
Copy the full SHA 1dc8eb9View commit details -
Configuration menu - View commit details
-
Copy full SHA for c4359b1 - Browse repository at this point
Copy the full SHA c4359b1View commit details
Commits on Aug 31, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 620facc - Browse repository at this point
Copy the full SHA 620faccView commit details
Commits on Sep 2, 2022
-
update README introduction for improved clarity
list available algorithms
2Configuration menu - View commit details
-
Copy full SHA for b35df4a - Browse repository at this point
Copy the full SHA b35df4aView commit details -
Configuration menu - View commit details
-
Copy full SHA for c547540 - Browse repository at this point
Copy the full SHA c547540View commit details -
Configuration menu - View commit details
-
Copy full SHA for b56b0e7 - Browse repository at this point
Copy the full SHA b56b0e7View commit details -
Merge pull request #736 from t-mat/gh-actions-remove-macos-10.15
Remove macOS-10.15 from ci.yml
Configuration menu - View commit details
-
Copy full SHA for 780c304 - Browse repository at this point
Copy the full SHA 780c304View commit details -
Configuration menu - View commit details
-
Copy full SHA for 66248f4 - Browse repository at this point
Copy the full SHA 66248f4View commit details -
Configuration menu - View commit details
-
Copy full SHA for 1309c27 - Browse repository at this point
Copy the full SHA 1309c27View commit details
Commits on Sep 7, 2022
-
Merge pull request #718 from Cyan4973/icc_fallthrough
blind fix for fallthrough on icc
Configuration menu - View commit details
-
Copy full SHA for 521e6d3 - Browse repository at this point
Copy the full SHA 521e6d3View commit details -
Merge pull request #734 from dougallj/neon-opt
Optimize XXH3_accumulate_512_neon
Configuration menu - View commit details
-
Copy full SHA for c420b59 - Browse repository at this point
Copy the full SHA c420b59View commit details
Commits on Sep 8, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 82b600a - Browse repository at this point
Copy the full SHA 82b600aView commit details -
Merge pull request #739 from DimitriPapadopoulos/codespell
Fix typos found by codespell
Configuration menu - View commit details
-
Copy full SHA for 43ea6fd - Browse repository at this point
Copy the full SHA 43ea6fdView commit details
Commits on Oct 19, 2022
-
fix issues reported by cppcheck
Run command "make clean cppcheck". ---- static analyzer - cppcheck ---- cppcheck . --force --enable=warning,portability,performance,style --error-exitcode=1 > /dev/null cli/xsum_os_specific.c:220:9: style: The scope of the variable 'ret' can be reduced. [variableScope] int ret; ^ cli/xsum_sanity_check.c:571:19: style: Condition 'len>0' is always true [knownConditionTrueFalse] } else if (len>0) { ^ cli/xsum_sanity_check.c:569:13: note: Assuming that condition 'len==0' is not redundant if (len == 0) { ^ cli/xsum_sanity_check.c:571:19: note: Condition 'len>0' is always true } else if (len>0) { ^ tests/bench/benchHash.c:112:55: error: Uninitialized variables: result.internal_never_ever_use_directly, result.error_result_never_ever_use_directly, result.error_tag_never_ever_use_directly [uninitvar] BMK_runTime_t const runTime = BMK_extract_runTime(result); ^ tests/bench/benchHash.c:107:12: note: Assuming condition is false while (!BMK_isCompleted_TimedFn(txf)) { ^ tests/bench/benchHash.c:107:12: note: Assuming condition is false while (!BMK_isCompleted_TimedFn(txf)) { ^ tests/bench/benchHash.c:107:12: note: Assuming condition is false while (!BMK_isCompleted_TimedFn(txf)) { ^ tests/bench/benchHash.c:112:55: note: Uninitialized variables: result.internal_never_ever_use_directly, result.error_result_never_ever_use_directly, result.error_tag_never_ever_use_directly BMK_runTime_t const runTime = BMK_extract_runTime(result); ^ tests/bench/benchfn.c:95:12: error: Uninitialized variable: outcome.error_result_never_ever_use_directly [uninitvar] return outcome; ^ tests/collisions/main.c:209:14: style: Same expression on both sides of '<='. [duplicateExpression] assert(m <= m); ^ make: *** [Makefile:387: cppcheck] Error 1 Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org>
Configuration menu - View commit details
-
Copy full SHA for dde858b - Browse repository at this point
Copy the full SHA dde858bView commit details -
ci: fix tipi build error on github CI workflow
[err] Dynamic exception type: std::runtime_error std::exception::what: can not create git repository [-3] - config value 'safe.directory' was not found [err] Something went wrong, for help on your issue, we could enjoy getting your report at https://tipi.build/#contact Error: Process completed with exit code 1. Apply the workaround from Yannic (#744) on github CI workflow Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org> Cc: Yannic Staudt <yannic@tipi.build>
Configuration menu - View commit details
-
Copy full SHA for 058a465 - Browse repository at this point
Copy the full SHA 058a465View commit details -
Merge pull request #749 from hzhuang1/fix_tipi
ci: fix tipi build error on github CI workflow
Configuration menu - View commit details
-
Copy full SHA for 2de0fd6 - Browse repository at this point
Copy the full SHA 2de0fd6View commit details
Commits on Oct 24, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 930b33d - Browse repository at this point
Copy the full SHA 930b33dView commit details -
Configuration menu - View commit details
-
Copy full SHA for 1e02e9a - Browse repository at this point
Copy the full SHA 1e02e9aView commit details -
Configuration menu - View commit details
-
Copy full SHA for 664c4b3 - Browse repository at this point
Copy the full SHA 664c4b3View commit details -
Configuration menu - View commit details
-
Copy full SHA for 9759aa1 - Browse repository at this point
Copy the full SHA 9759aa1View commit details -
Configuration menu - View commit details
-
Copy full SHA for 0e1d320 - Browse repository at this point
Copy the full SHA 0e1d320View commit details
Commits on Oct 31, 2022
-
Use
vpsrlq
instead ofvpshufd
in avx2/avx512 accumulate.`vpshufd` was just to essentially implement `dst >> 32` which can (and should) be done with `vpsrlq`. `XXH3_{32|64|128}bits` throughput seems to be limitted by a port_5 bottleneck and under-uses port_0. Hashing 1mb in a loop 100000 times with `vpshufd`: x86-64-v4: 5,363,095,005 cycles 5,195,740,817 port_0 17,104,516 port_1 5,358,811,641 port_5 x86-64-v3: 7,954,143,671 cycles 6,610,562,251 port_0 7,405,810,291 port_1 7,474,008,658 port_5 `vpshufd` on x86-64-v3 only runs on port_5 (and for x86-64-v4 only runs on port_5 when used with `zmm` registers). `vpsrlq` can run on port_{0|1} (port_0 only for `zmm`) which helps to alleviate this bottleneck. Hashing 1mb in a loop 100000 times with `vpsrlq`: x86-64-v4: 5,305,723,460 cycles 5,246,052,997 port_0 17,759,291 port_1 5,303,090,711 port_5 x86-64-v3: 7,852,139,442 cycles 6,950,368,526 port_0 7,301,134,590 port_1 7,215,987,871 port_5 which in general gets a better port distribution and results in slightly better times. Note the sse2 accumulate also would benefit from shifting uops from port5 -> port0 x86-64-v2 with `pshufd`: 13,128,826,786 port_0 15,014,228,537 port_1 14,801,410,216 port_5 x86-64-v2 with `psrlq`: 13,899,468,407 port_0 14,774,790,828 port_1 14,555,300,082 port_5 But since `pshufd` has a 3-term encoding and `psrlq` only has a 2-term encoding, using `psrlq` causes an extra `movdqa` in codegen which is a net-negative (at least on Icelake benchmarks). Benchmarking on Icelake this commit results in the following improvements: NB: The avx2 changes where also benchmarked on Icelake which is not really ideal. Although note that Icelake (unlike native x86-64-v3 hardware has a faster `vpshufd` (port_{1|5} vs just port_5) so hopefully anything skewed by incorrect hardware was infavor of the original implementation).
Configuration menu - View commit details
-
Copy full SHA for f534e5b - Browse repository at this point
Copy the full SHA f534e5bView commit details -
Avoid signed truncation in hash 17to128 and 129to240
Signed truncation of len gets worse codegen so keep it unsigned. Signed codegen for loop bound on x86-64-v4: ``` mov %esi,%ecx lea 0xe(%rsi),%r10d dec %ecx cmovns %ecx,%r10d sar $0x4,%r10d ``` Unsigned codegen for loop bound on x86-64-v4: ``` lea -0x1(%rsi),%ecx mov %ecx,%r9d shr $0x4,%r9d ```
Configuration menu - View commit details
-
Copy full SHA for 1456dc5 - Browse repository at this point
Copy the full SHA 1456dc5View commit details -
Replace 2x vpxor with vpternlogd in avx512 scramble
Neither Clang nor GCC <= 11.2 get this so seems worth it to do by hand so save the instruction. Not really on the critical path so only minor perf changes. Hashing 8kb On Icelake with 2x vpxor: 5,451,267,137 cycles 5,371,545,012 p0 301,488,579 p1 5,420,594,430 p5 450,583,083 p6 Hashing 8kb On Icelake with 2x vpternlogd: 5,400,583,791 cycles 5,329,872,021 p0 305,543,626 p1 5,372,500,814 p5 446,853,381 p6
Configuration menu - View commit details
-
Copy full SHA for b2929c4 - Browse repository at this point
Copy the full SHA b2929c4View commit details
Commits on Nov 2, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 2e69ede - Browse repository at this point
Copy the full SHA 2e69edeView commit details -
Configuration menu - View commit details
-
Copy full SHA for 38f1de2 - Browse repository at this point
Copy the full SHA 38f1de2View commit details -
Merge pull request #742 from DimitriPapadopoulos/actions
Update GitHub Actions
Configuration menu - View commit details
-
Copy full SHA for 267ba94 - Browse repository at this point
Copy the full SHA 267ba94View commit details
Commits on Nov 8, 2022
-
xxhash: add ARM SVE intrinsic implementation
Implement intrinsic code for ARM SVE. In this patch, it seems that there is no improvement. Further optimization will be contained in the later patch set. SCALAR implementation (default) === benchmarking 4 hash functions === benchmarking large inputs : from 512 bytes (log9) to 128 MB (log27) xxh3 , 2718, 3093, 3151, 3187, 3217, 3232, 3238, 3223, 3190, 3240, 3241, 3222, 3238, 3049, 2253, 2113, 2102, 2117, 2132 XXH32 , 1499, 1520, 1538, 1543, 1543, 1545, 1547, 1535, 1486, 1487, 1484, 1468, 1484, 1435, 1231, 1183, 1184, 1185, 1187 XXH64 , 2632, 2874, 3017, 3092, 3132, 3143, 3155, 3122, 2990, 2994, 2998, 3000, 3006, 2842, 2121, 1838, 1833, 1811, 1839 XXH128 , 2435, 2872, 3024, 3121, 3187, 3219, 3234, 3222, 3221, 3225, 3225, 3227, 3227, 3034, 1962, 1789, 1866, 1871, 1874 Intrinsic SVE512 implementation === benchmarking 4 hash functions === benchmarking large inputs : from 512 bytes (log9) to 128 MB (log27) xxh3 , 1903, 2313, 2469, 2568, 2631, 2663, 2683, 2662, 2673, 2678, 2679, 2678, 2676, 2584, 2236, 2173, 2173, 2163, 2183 XXH32 , 1326, 1436, 1495, 1523, 1535, 1543, 1547, 1536, 1505, 1506, 1504, 1507, 1508, 1446, 1246, 1193, 1194, 1194, 1167 XXH64 , 2510, 2802, 2977, 3071, 3121, 3136, 3156, 3126, 3041, 3047, 3050, 3050, 3048, 2890, 2043, 1944, 1944, 1945, 1955 XXH128 , 1867, 2295, 2462, 2542, 2627, 2663, 2679, 2656, 2661, 2675, 2673, 2647, 2671, 2574, 2218, 2144, 2169, 2175, 2181 Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org> Signed-off-by: Devin Hussey <easyaspi314@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for c6205ae - Browse repository at this point
Copy the full SHA c6205aeView commit details -
Since the vector size of ARM64 SVE ranges from 128-bit to 2048-bit, configure CI to run with each vector size. Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org>
Configuration menu - View commit details
-
Copy full SHA for dd19080 - Browse repository at this point
Copy the full SHA dd19080View commit details
Commits on Nov 9, 2022
-
Speedup size hash for 17to128 by using independent accumulators.
Faster (better IPC) to make computation slightly more independent. Results are from geometric mean of N = 6 runs. Machine: ``` 11th Gen Intel(R) Core(TM) i7-11850H @ 2.50GHz ``` CC: ``` gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0 ``` Bench Command: ``` // Change is only for 17-128 range. $> ./benchHash xxh3 --mins=1 --maxs=256 --minl=0 --maxl=0 ``` Aggregated Results for [17, 128] size range: Times reported as geometric mean of all speedups. Latency for small inputs of fixed size : - 1.073 Throughput small inputs of fixed size (from 1 to 256 bytes): - 1.173 Latency for small inputs of random size [1-N] : - 1.051 benchmarking random size inputs [1-N] : - 1.134 So roughly 5-17% improvement.
15Configuration menu - View commit details
-
Copy full SHA for 7221be4 - Browse repository at this point
Copy the full SHA 7221be4View commit details -
Merge pull request #752 from hzhuang1/sve_01
xxhash: support SVE by intrinsic code
Configuration menu - View commit details
-
Copy full SHA for dbacfec - Browse repository at this point
Copy the full SHA dbacfecView commit details
Commits on Nov 10, 2022
-
Merge pull request #746 from hzhuang1/cppcheck_01
fix issues reported by cppcheck
Configuration menu - View commit details
-
Copy full SHA for d0abdaa - Browse repository at this point
Copy the full SHA d0abdaaView commit details -
Missing space caused that github action doesn't work. Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org>
Configuration menu - View commit details
-
Copy full SHA for 63356da - Browse repository at this point
Copy the full SHA 63356daView commit details -
Configuration menu - View commit details
-
Copy full SHA for 8e94343 - Browse repository at this point
Copy the full SHA 8e94343View commit details -
Configuration menu - View commit details
-
Copy full SHA for e445e5b - Browse repository at this point
Copy the full SHA e445e5bView commit details -
Merge pull request #760 from tipi-deps/dev
Fixing tipi-build / Build as dependency CI step
Configuration menu - View commit details
-
Copy full SHA for 3078dc6 - Browse repository at this point
Copy the full SHA 3078dc6View commit details -
switch to full accmulating loop
XXH3_accumulate() handle the whole accumulating loop and architecture optimized code is in the mini loop of 512 bytes. But it also causes accessing memory frequently for the large block data. Now make XXH3_accumulate() as architecture optimized code. Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org> Signed-off-by: Devin Hussey <easyaspi314@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for 91788f1 - Browse repository at this point
Copy the full SHA 91788f1View commit details -
customize full accumulating loop for ARM64 SVE
With optimized full accumulating loop, the performance is improved at least 2 times. The ACC result needn't to save to stack in the full loop. And instructions of prefetching data for SVE are also used. Without this patch, the performance result is in below. === benchmarking 4 hash functions === benchmarking large inputs : from 512 bytes (log9) to 128 MB (log27) xxh3 , 1904, 2315, 2468, 2580, 2640, 2670, 2682, 2673, 2677, 2663, 2683, 2688, 2686, 2591, 2241, 2181, 2191, 2048, 2048 XXH32 , 1326, 1440, 1493, 1523, 1534, 1543, 1547, 1532, 1504, 1507, 1507, 1505, 1506, 1446, 1218, 1150, 1151, 1153, 1135 XXH64 , 2511, 2795, 2975, 3068, 3120, 3125, 3154, 3128, 3034, 3045, 3052, 3053, 3053, 2842, 2050, 1853, 1848, 1853, 1853 XXH128 , 1867, 2294, 2465, 2569, 2622, 2662, 2676, 2667, 2677, 2682, 2684, 2677, 2683, 2570, 2093, 2013, 2045, 2046, 2046 With this patch, the performance result is in below. === benchmarking 4 hash functions === benchmarking large inputs : from 512 bytes (log9) to 128 MB (log27) xxh3 , 3681, 6007, 7803, 8954, 9875, 10411, 10703, 10505, 10670, 10794, 10812, 10804, 10205, 9923, 6279, 5927, 5967, 6022, 6062 XXH32 , 1281, 1434, 1494, 1523, 1534, 1543, 1547, 1535, 1500, 1502, 1502, 1502, 1501, 1443, 1242, 1169, 1193, 1196, 1195 XXH64 , 2497, 2801, 2961, 3074, 3092, 3136, 3155, 3123, 3031, 3037, 3040, 3037, 3033, 2847, 2102, 1955, 1967, 1974, 1971 XXH128 , 3419, 5798, 7488, 8854, 9787, 10357, 10673, 10468, 10647, 10748, 10785, 10751, 10805, 9698, 6011, 5677, 5999, 6065, 6074 Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org> Signed-off-by: Devin Hussey <easyaspi314@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for cfbf0b7 - Browse repository at this point
Copy the full SHA cfbf0b7View commit details
Commits on Nov 24, 2022
-
Merge pull request #756 from hzhuang1/sve_02
Customize full accumulating loop for SVE
Configuration menu - View commit details
-
Copy full SHA for 30d6a3e - Browse repository at this point
Copy the full SHA 30d6a3eView commit details
Commits on Nov 30, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 3d01088 - Browse repository at this point
Copy the full SHA 3d01088View commit details -
Merge pull request #765 from Cyan4973/ci-macos12
added macos-12 test to GH CI
Configuration menu - View commit details
-
Copy full SHA for 081c1b6 - Browse repository at this point
Copy the full SHA 081c1b6View commit details
Commits on Dec 2, 2022
-
Fix '-Wsign-conversion' warning in XXH3_initCustomSecret_avx512
No change to generated code, just suppresses the warning when using stricter compiler checks.
Configuration menu - View commit details
-
Copy full SHA for dced34e - Browse repository at this point
Copy the full SHA dced34eView commit details -
Merge pull request #754 from goldsteinn/dev
Small improvement to x86 vectorized hashes and medium-sizes hash.
Configuration menu - View commit details
-
Copy full SHA for 86e1f8e - Browse repository at this point
Copy the full SHA 86e1f8eView commit details
Commits on Dec 13, 2022
-
dispatch: Use __attribute__((constructor)) on XXH_setDispatch
Minimal change but eliminates the NULL check if the attribute is supported.
Configuration menu - View commit details
-
Copy full SHA for b8c3176 - Browse repository at this point
Copy the full SHA b8c3176View commit details -
Merge pull request #773 from goldsteinn/dev
dispatch: Use __attribute__((constructor)) on XXH_setDispatch
Configuration menu - View commit details
-
Copy full SHA for 7f9cec7 - Browse repository at this point
Copy the full SHA 7f9cec7View commit details
Commits on Dec 27, 2022
-
Makefile: add install_libxxhash.includes target
Add install_libxxhash.includes to be able to install headers when building and installing the static library Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for b7c91f7 - Browse repository at this point
Copy the full SHA b7c91f7View commit details
Commits on Dec 29, 2022
-
Configuration menu - View commit details
-
Copy full SHA for 842a149 - Browse repository at this point
Copy the full SHA 842a149View commit details -
Configuration menu - View commit details
-
Copy full SHA for bdd92de - Browse repository at this point
Copy the full SHA bdd92deView commit details
Commits on Dec 30, 2022
-
Merge pull request #777 from DimitriPapadopoulos/codespell
Fix typo found by codespell
Configuration menu - View commit details
-
Copy full SHA for 65515bb - Browse repository at this point
Copy the full SHA 65515bbView commit details -
Merge pull request #778 from DimitriPapadopoulos/http
http:// → https://
Configuration menu - View commit details
-
Copy full SHA for 4ebd833 - Browse repository at this point
Copy the full SHA 4ebd833View commit details
Commits on Jan 9, 2023
-
s390x: add memory barrier to avoid fault issue
The issue link is in below. #766 This issue occurs randomly. If we downgrade gcc optimization from "O3" to "O2", this issue is gone. Now use memory barrier to implement the same effect. Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org>
Configuration menu - View commit details
-
Copy full SHA for d8c2dfe - Browse repository at this point
Copy the full SHA d8c2dfeView commit details
Commits on Jan 12, 2023
-
Add __attribute__((noescape)) on pointers in public API
LLVM only at the moment. Without LTO its the only way to inform the compiler that the pointer does not become globally visible due to the function call.
Configuration menu - View commit details
-
Copy full SHA for 2765879 - Browse repository at this point
Copy the full SHA 2765879View commit details -
Small improvements to 129to240 hash ranges
1. Don't compute `nbRounds` and use len as upper bound in 129to240 Codegen was previously using the divide to truncate the lower-bits but was shifting back for the actual loop. Previous Codegen: ``` // XXH128_129to240: shr $0x5,%r15d lea -0x5(%r15),%r13d shl $0x5,%r13 ... ``` which is totally unnecessary. On aarch64 we are likewise able to eliminate some `lsl` instructions. Similiar changes "should" make sense in the 64bit version but seemed to always cause minor regressions for M1 so scrapped. 2. Add second accumulator in `XXH3_len_129to240_64b` to get a bit more out of OOE. Note, for aarch64 we have to add a `XXH_COMPILER_GUARD(acc)` in the final loop of `XXH3_len_129to240_64b`. Without it clang and gcc where unrolling the second loop and interleaving it with the first. This caused a LOT of unnecessary constant materialization (unique for each unrolled iteration).
Configuration menu - View commit details
-
Copy full SHA for 755445d - Browse repository at this point
Copy the full SHA 755445dView commit details
Commits on Jan 13, 2023
-
Merge pull request #782 from goldsteinn/dev
Add __attribute__((noescape)) on pointers in public API
Configuration menu - View commit details
-
Copy full SHA for fe98d2b - Browse repository at this point
Copy the full SHA fe98d2bView commit details -
Annotate xxh_x86dispatch with XXH_NOESCAPE
The pointers don't escape.
3Configuration menu - View commit details
-
Copy full SHA for f4e2de6 - Browse repository at this point
Copy the full SHA f4e2de6View commit details
Commits on Jan 16, 2023
-
Merge pull request #786 from goldsteinn/dev
Annotate xxh_x86dispatch with XXH_NOESCAPE
Configuration menu - View commit details
-
Copy full SHA for d7197dd - Browse repository at this point
Copy the full SHA d7197ddView commit details
Commits on Jan 22, 2023
-
- Remove VZIP hack - ARMv7a is going to be using the 4 lane path, so there is no benefit to having the increased complexity. - Reorder the 4 lane path to clarify the paired operations. This also seems to slightly improve performance on Clang (not tested on GCC). - Rename variables to match SSE2 and be more consistent - Document how the VUZP trick works
Configuration menu - View commit details
-
Copy full SHA for 0e73b84 - Browse repository at this point
Copy the full SHA 0e73b84View commit details -
Configuration menu - View commit details
-
Copy full SHA for 2f54d1b - Browse repository at this point
Copy the full SHA 2f54d1bView commit details
Commits on Jan 23, 2023
-
[NEON] Simplify, only compiler guard on Clang
- Remove acc_vec variable in favor of directly indexing - Removing the compiler guard for GCC 11 allows it to get to 23 GB/s from 20 GB/s - Slight cosmetic tweaks for the 4 lane loop allowing it to be commented out.
Configuration menu - View commit details
-
Copy full SHA for 0230310 - Browse repository at this point
Copy the full SHA 0230310View commit details
Commits on Jan 25, 2023
-
Define XXH_RESTRICT for some C++ compilers (gcc/clang/msvc/icc)
There are a LOT more compilers that recognize `__restrict` but this this hits the common ones.
Configuration menu - View commit details
-
Copy full SHA for 2ddf958 - Browse repository at this point
Copy the full SHA 2ddf958View commit details -
Merge pull request #787 from easyaspi314/neon_refactor
[NEON] Refactor NEON code
Configuration menu - View commit details
-
Copy full SHA for 22a9afe - Browse repository at this point
Copy the full SHA 22a9afeView commit details -
Merge pull request #776 from ffontaine/dev
Makefile: add install_libxxhash.includes target
Configuration menu - View commit details
-
Copy full SHA for e0fd060 - Browse repository at this point
Copy the full SHA e0fd060View commit details -
Merge pull request #785 from hzhuang1/barrier_01
s390x: add memory barrier to avoid fault issue
Configuration menu - View commit details
-
Copy full SHA for 1dc3a5c - Browse repository at this point
Copy the full SHA 1dc3a5cView commit details -
Merge pull request #788 from goldsteinn/dev
Define XXH_RESTRICT for some C++ compilers (gcc/clang/msvc/icc)
Configuration menu - View commit details
-
Copy full SHA for 6dba9ab - Browse repository at this point
Copy the full SHA 6dba9abView commit details
Commits on Jan 29, 2023
-
Add dispatch mode with option DISPATCH. Build with DISPATCH: $cmake ../cmake_unofficial -DDISPATCH=ON $cmake --build ./ Signed-off-by: Haojian Zhuang <haojian.zhuang@linaro.org>
Configuration menu - View commit details
-
Copy full SHA for c171483 - Browse repository at this point
Copy the full SHA c171483View commit details
Commits on Feb 1, 2023
-
Fix
XXH_UNREACHABLE
feature test for C++23 and remove for C23 #791gcc11/12 don't support `std::unreachable()` in there `-std=c++23` build. A more reliable check is: ``` ``` so use that. A similiar issue may occur for C23 `unreachable()` so just remove it on the basis that if a compiler doesn't currently support some builtin for unreachable/assume its very likely not using it in analysis so we are probably not missing much (if anything) leaving it out.
Configuration menu - View commit details
-
Copy full SHA for 35031d3 - Browse repository at this point
Copy the full SHA 35031d3View commit details
Commits on Feb 12, 2023
-
Configuration menu - View commit details
-
Copy full SHA for adbfd2a - Browse repository at this point
Copy the full SHA adbfd2aView commit details
Commits on Feb 23, 2023
-
Configuration menu - View commit details
-
Copy full SHA for e169bed - Browse repository at this point
Copy the full SHA e169bedView commit details -
Merge pull request #790 from hzhuang1/cmake_01
cmake: support dispatch mode
Configuration menu - View commit details
-
Copy full SHA for dba3a29 - Browse repository at this point
Copy the full SHA dba3a29View commit details -
Merge pull request #792 from goldsteinn/dev
Fix `XXH_UNREACHABLE` feature test for C++23 and remove for C23 #791
Configuration menu - View commit details
-
Copy full SHA for 3f5c75c - Browse repository at this point
Copy the full SHA 3f5c75cView commit details
Commits on Feb 25, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 4ff3a1b - Browse repository at this point
Copy the full SHA 4ff3a1bView commit details
Commits on Feb 27, 2023
-
The issue actually appears to be GCC disliking that a function pointer is SOMETIMES forceinline, and withSecret was marked force inline instead of noinline (despite what the comments suggested)
Configuration menu - View commit details
-
Copy full SHA for a2f7a0f - Browse repository at this point
Copy the full SHA a2f7a0fView commit details -
- XXH3_consumeStripes can now process multiple stripes, greatly simplifying the logic - Tail merged the last XXH3_accumulate_512 in XXH3_digestLong - XXH3_128bits_update now just tail calls XXH3_64bits_update to reduce code size
Configuration menu - View commit details
-
Copy full SHA for 44bb94e - Browse repository at this point
Copy the full SHA 44bb94eView commit details -
Configuration menu - View commit details
-
Copy full SHA for 268fde4 - Browse repository at this point
Copy the full SHA 268fde4View commit details
Commits on Feb 28, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 86c3c05 - Browse repository at this point
Copy the full SHA 86c3c05View commit details -
Merge pull request #805 from easyaspi314/xxh3_update_rewrite
Rewrite XXH3_update
Configuration menu - View commit details
-
Copy full SHA for 2b328a1 - Browse repository at this point
Copy the full SHA 2b328a1View commit details
Commits on Mar 1, 2023
-
Configuration menu - View commit details
-
Copy full SHA for ace22bd - Browse repository at this point
Copy the full SHA ace22bdView commit details -
Configuration menu - View commit details
-
Copy full SHA for edb63d3 - Browse repository at this point
Copy the full SHA edb63d3View commit details
Commits on Mar 2, 2023
-
- Use inline assembly to force umaddl in the scalarRound code - Prevents the slow 64-bit `MADD` which has +4c latency and a 2c stall on little cores. - Also prevents a mask - Improves scalar speed even on big cores, should also improve mixed NEON speed on those little cores, although I haven't tested it. - Use inline assembly for `vmlal_u32` - Fixes NEON performance being garbage on GCC < 11 because it can't fold `vget_low/high` into `vmlal` (arm_neon.h uses inline assembly) - Documents the presumed reason why umlal must come before add. - Doubles as Clang's asm guard. (The old asm guard is still used on the single vector path) - Much cleaner in the code itself - Fix header detection - Include arm_neon.h when `__ARM_FEATURE_SVE` is defined - Don't include arm_neon.h if only `__aarch64__` is defined (e.g. -march=armv8-a+nofp) - Don't include arm_neon.h on < ARMv7 MSVC (which only would happen on old WinCE toolchains) - Fix the generateSecret MOVK hack - Add it for GCC, it does that too now - Remove the assert because the change to use `unreachable()` undoes it
Configuration menu - View commit details
-
Copy full SHA for e701df3 - Browse repository at this point
Copy the full SHA e701df3View commit details
Commits on Mar 4, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 62aa2f0 - Browse repository at this point
Copy the full SHA 62aa2f0View commit details
Commits on Mar 5, 2023
-
Remove volatile from asm guards, return Clang guards
The volatile in the asm guards was causing a reorder barrier which was preventing interleaving of XXH32 and XXH3 NEON. Additionally, with the better asm guard, Clang does better with intrinsics.
Configuration menu - View commit details
-
Copy full SHA for e6732e9 - Browse repository at this point
Copy the full SHA e6732e9View commit details -
[NEON/VSX] Fix GCC strict aliasing issues
This fixes GCC 12 on macOS ARM64, as well as the existing s390x zVector bug that was worked around with a `__sync_synchronize()` by using `__attribute__((may_alias))` which excludes a type from strict aliasing optimizations. `vldq`/`vst1q` are still avoided due to older versions having issues with load-store optimizations. This is not needed on x86 because the SSE types are impossible to not alias.
Configuration menu - View commit details
-
Copy full SHA for ad63946 - Browse repository at this point
Copy the full SHA ad63946View commit details
Commits on Mar 6, 2023
-
Configuration menu - View commit details
-
Copy full SHA for a09fd87 - Browse repository at this point
Copy the full SHA a09fd87View commit details
Commits on Mar 7, 2023
-
fix: GH-Actions - removed ubuntu-18.04
Due to deprecation, we remove ubuntu-18.04 and the following compilers. x86_64 - Removed - gcc-{ 4.8, 5, 6 } - clang-{ 3.9, 4, 5 } - Moved from ubuntu-20.04 to ubuntu-22.04 - gcc-{ 9, 10, 11 } - clang-{ 11, 12 } QEMU - Removed {arm, aarch64, mips, ppc64le, ppc64, s390x}-gcc-{ 6, 7 } - Removed mips-gcc-8 - Ubuntu-20.04 didn't have mips-gcc-8 for some reason.
Takayuki Matsuoka committedMar 7, 2023 Configuration menu - View commit details
-
Copy full SHA for eaea300 - Browse repository at this point
Copy the full SHA eaea300View commit details -
Merge pull request #817 from t-mat/remove-ubuntu-18.04
fix: GH-Actions - remove ubuntu-18.04
Configuration menu - View commit details
-
Copy full SHA for ca91743 - Browse repository at this point
Copy the full SHA ca91743View commit details -
Fix off-by-one in XXH3_consumeStripes() (Fixes #816)
Silly error by me that caused a scramble to be skipped.
Configuration menu - View commit details
-
Copy full SHA for 65094db - Browse repository at this point
Copy the full SHA 65094dbView commit details
Commits on Mar 8, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 3dd5a3a - Browse repository at this point
Copy the full SHA 3dd5a3aView commit details -
Configuration menu - View commit details
-
Copy full SHA for 09b6fc7 - Browse repository at this point
Copy the full SHA 09b6fc7View commit details -
Configuration menu - View commit details
-
Copy full SHA for 5be457e - Browse repository at this point
Copy the full SHA 5be457eView commit details
Commits on Mar 9, 2023
-
Merge pull request #814 from t-mat/dev
Fix: add gcc -Og check and fix its false positive warnings
Configuration menu - View commit details
-
Copy full SHA for a568bee - Browse repository at this point
Copy the full SHA a568beeView commit details -
Merge pull request #818 from easyaspi314/fix-816
Fix off-by-one in XXH3_consumeStripes() (Fixes #816)
Configuration menu - View commit details
-
Copy full SHA for 5b2032f - Browse repository at this point
Copy the full SHA 5b2032fView commit details -
Merge pull request #807 from Cyan4973/gh_read
added read-all permission status to Github Actions Workflow
Configuration menu - View commit details
-
Copy full SHA for 74c416c - Browse repository at this point
Copy the full SHA 74c416cView commit details -
added one test length for xxh sanity checks
to detect off-by-one scrambling error like #816
Configuration menu - View commit details
-
Copy full SHA for 7dd426a - Browse repository at this point
Copy the full SHA 7dd426aView commit details -
Merge pull request #803 from devnexen/clang__assume
XXH_ASSUME macro using `__builtin_assume` if supported (clang only fo…
Configuration menu - View commit details
-
Copy full SHA for adc4fc8 - Browse repository at this point
Copy the full SHA adc4fc8View commit details -
Configuration menu - View commit details
-
Copy full SHA for 449372c - Browse repository at this point
Copy the full SHA 449372cView commit details
Commits on Mar 10, 2023
-
benchmark : adjust initial speed target to 10 MB/s
Additionally, the initial speed target can be selected at compilation time, by setting new optional build macro XXH_1ST_SPEED_TARGET. Solves @easyaspi314 comment in #821 (comment), in order to support "1.5 GHz Cortex-A53 tablet with s390x in QEMU with -O0 -mzvector" which is very slow.
Configuration menu - View commit details
-
Copy full SHA for bea35c6 - Browse repository at this point
Copy the full SHA bea35c6View commit details -
Configuration menu - View commit details
-
Copy full SHA for 62d8bd1 - Browse repository at this point
Copy the full SHA 62d8bd1View commit details
Commits on Mar 11, 2023
-
Add sanity test vector generator
sanity_test_vectors_generator.c generates sanity_test_vectors.h
Configuration menu - View commit details
-
Copy full SHA for 94ddfec - Browse repository at this point
Copy the full SHA 94ddfecView commit details -
Configuration menu - View commit details
-
Copy full SHA for 464645e - Browse repository at this point
Copy the full SHA 464645eView commit details -
Note: sanity_test.c depends sanity_test_vectors.h which has generated by sanity_test_vectors_generator.c.
Configuration menu - View commit details
-
Copy full SHA for 9c377f5 - Browse repository at this point
Copy the full SHA 9c377f5View commit details -
Add Makefile entry and gitignore for sanity_test
make check Depends (invokes) test_sanity make test_sanity Invokes make -C tests test_sanity make -C tests test_sanity Create and execute tests/sanity_test make -C tests sanity_test_vectors.h Create and execute tests/sanity_test_vectors_generator. It generates tests/sanity_test_vectors.h.
Configuration menu - View commit details
-
Copy full SHA for 25dfa1a - Browse repository at this point
Copy the full SHA 25dfa1aView commit details -
Configuration menu - View commit details
-
Copy full SHA for 9d34bcc - Browse repository at this point
Copy the full SHA 9d34bccView commit details -
Configuration menu - View commit details
-
Copy full SHA for b618de4 - Browse repository at this point
Copy the full SHA b618de4View commit details -
Temporarily comment out 'AARCH64, gcc-8' for incompatibility.
As for ["AARCH64, gcc-8" error](https://github.com/Cyan4973/xxHash/actions/runs/4392604424/jobs/7692471884), It seems aarch64-linux-gnu-gcc-8 has linker issue. When I remove `LDFLAGS="-static"` linker works. But when I don't give `LDFLAGS="-static"`, `qemu-aarch64-static` fails to read `/lib/ld-linux-aarch64.so.1`. We can run it with `qemu-aarch64` (no `-static`) though. ``` CPPFLAGS="-DXXH_VECTOR=XXH_SCALAR" CC=aarch64-linux-gnu-gcc-8 RUN_ENV="qemu-aarch64 -L /usr/aarch64-linux-gnu" make clean check ```
Configuration menu - View commit details
-
Copy full SHA for 7190dbe - Browse repository at this point
Copy the full SHA 7190dbeView commit details -
Configuration menu - View commit details
-
Copy full SHA for 1b3dde6 - Browse repository at this point
Copy the full SHA 1b3dde6View commit details -
Merge pull request #822 from Cyan4973/bench10
benchmark : adjust initial speed target to 10 MB/s
Configuration menu - View commit details
-
Copy full SHA for 6ff46e7 - Browse repository at this point
Copy the full SHA 6ff46e7View commit details -
Merge pull request #820 from Cyan4973/test816
added one test length for xxh sanity checks
Configuration menu - View commit details
-
Copy full SHA for c9df1e2 - Browse repository at this point
Copy the full SHA c9df1e2View commit details -
Merge pull request #809 from easyaspi314/aarch64-hacks
Various AArch64 speed hacks
2Configuration menu - View commit details
-
Copy full SHA for a60bd05 - Browse repository at this point
Copy the full SHA a60bd05View commit details
Commits on Mar 15, 2023
-
Add some nodejs workarounds so make check works
- Link in `nodefs` and `noderawfs` - Use Node's `tty.isatty()` via inline JS instead of the broken libc `isatty()` - Used with `make NODE_JS=1`
Configuration menu - View commit details
-
Copy full SHA for 207e2ab - Browse repository at this point
Copy the full SHA 207e2abView commit details -
Add node.js wasm/asm.js tests for Github CI
Currently only one version of EMCC, testing node 16, 17, and 18. Cache is used because the emsdk has to cache each library which takes a bit.
Configuration menu - View commit details
-
Copy full SHA for 125f89f - Browse repository at this point
Copy the full SHA 125f89fView commit details -
[WASM] Support SIMD128 via XXH_NEON (from EMSDK's SIMDe polyfill)
The Emscripten SDK includes arm_neon.h from SIMDeverywhere to port NEON projects. Since SIMD128 is very similar to NEON without half vectors, the double NEON path maps perfectly to WASM SIMD128 and can reach full speed with little modification. (As a matter of fact I was able to make it slightly faster). Note that if XXH3_NEON_LANES is not a multiple of 4, SIMDe will scalarize the single width NEON paths, so 8 is strongly recommended. Also, I found an optimization to scrambleAcc which should have been obvious.
Configuration menu - View commit details
-
Copy full SHA for c9a7470 - Browse repository at this point
Copy the full SHA c9a7470View commit details -
Compiler guard XXH32 on WASM128
It JITs to the same thing that slowed down SSE4 and NEON.
Configuration menu - View commit details
-
Copy full SHA for ab50132 - Browse repository at this point
Copy the full SHA ab50132View commit details -
Configuration menu - View commit details
-
Copy full SHA for 5a9427c - Browse repository at this point
Copy the full SHA 5a9427cView commit details -
Configuration menu - View commit details
-
Copy full SHA for fd63843 - Browse repository at this point
Copy the full SHA fd63843View commit details
Commits on Mar 16, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 082a9a8 - Browse repository at this point
Copy the full SHA 082a9a8View commit details -
Merge pull request #823 from t-mat/dedicated-sanity-test-program
Dedicated sanity test program
Configuration menu - View commit details
-
Copy full SHA for 554e98d - Browse repository at this point
Copy the full SHA 554e98dView commit details
Commits on Apr 3, 2023
-
Signed-off-by: Gabriela Gutierrez <gabigutierrez@google.com>
Configuration menu - View commit details
-
Copy full SHA for 772eb38 - Browse repository at this point
Copy the full SHA 772eb38View commit details -
Configuration menu - View commit details
-
Copy full SHA for f4bef92 - Browse repository at this point
Copy the full SHA f4bef92View commit details
Commits on May 12, 2023
-
Ref actions by commit SHA in ci.yml
It's important to make sure the SHA's are from the original repositories and not forks. For reference: https://github.com/actions/checkout/releases/tag/v3.5.2 actions/checkout@8e5e7e5 https://github.com/msys2/setup-msys2/releases/tag/v2 msys2/setup-msys2@7efe20b Signed-off-by: Gabriela Gutierrez <gabigutierrez@google.com>
Configuration menu - View commit details
-
Copy full SHA for 295ee07 - Browse repository at this point
Copy the full SHA 295ee07View commit details
Commits on May 30, 2023
-
Merge pull request #833 from gabibguti/dev
Reference actions by commit SHA
Configuration menu - View commit details
-
Copy full SHA for 5562a14 - Browse repository at this point
Copy the full SHA 5562a14View commit details -
Configuration menu - View commit details
-
Copy full SHA for 616ca44 - Browse repository at this point
Copy the full SHA 616ca44View commit details
Commits on Jun 12, 2023
-
Bump actions/checkout from 3.5.2 to 3.5.3
Bumps [actions/checkout](https://github.com/actions/checkout) from 3.5.2 to 3.5.3. - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](actions/checkout@8e5e7e5...c85c95e) --- updated-dependencies: - dependency-name: actions/checkout dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com>
Configuration menu - View commit details
-
Copy full SHA for 63d953a - Browse repository at this point
Copy the full SHA 63d953aView commit details -
Merge pull request #837 from Cyan4973/dependabot/github_actions/actio…
…ns/checkout-3.5.3 Bump actions/checkout from 3.5.2 to 3.5.3
Configuration menu - View commit details
-
Copy full SHA for f11745e - Browse repository at this point
Copy the full SHA f11745eView commit details
Commits on Jun 17, 2023
-
feature: support # comment lines in --check files
This change fixes issue #836. Basically, we just follow the following logic in the digest.c of GNU coreutils. https://github.com/coreutils/coreutils/blob/d53190ed46a55f599800ebb2d8ddfe38205dbd24/src/digest.c#L1180-L1182 Note that we only allow '#' at the beginning of the line.
Configuration menu - View commit details
-
Copy full SHA for 2f405e4 - Browse repository at this point
Copy the full SHA 2f405e4View commit details -
Configuration menu - View commit details
-
Copy full SHA for 0691980 - Browse repository at this point
Copy the full SHA 0691980View commit details -
Configuration menu - View commit details
-
Copy full SHA for e24f111 - Browse repository at this point
Copy the full SHA e24f111View commit details -
Add "case GetLine_comment:" to conform -Werror=switch-enum. ``` cli/xxhsum.c: In function ‘XSUM_parseFile1’: cli/xxhsum.c:814:17: error: enumeration value ‘GetLine_comment’ not handled in switch [-Werror=switch-enum] 814 | switch (XSUM_getLineResult) | ^~~~~~ ```
Configuration menu - View commit details
-
Copy full SHA for 4823f56 - Browse repository at this point
Copy the full SHA 4823f56View commit details
Commits on Jun 19, 2023
-
reworded the AVX error message when compiling xxh_x86dispatch.c
Suggest the XXH_X86DISPATCH_ALLOW_AVX solution upfront.
Configuration menu - View commit details
-
Copy full SHA for 3682673 - Browse repository at this point
Copy the full SHA 3682673View commit details -
Configuration menu - View commit details
-
Copy full SHA for bb52930 - Browse repository at this point
Copy the full SHA bb52930View commit details -
Configuration menu - View commit details
-
Copy full SHA for 848f034 - Browse repository at this point
Copy the full SHA 848f034View commit details -
Merge pull request #841 from Cyan4973/notipi
removed tipi test from CI suite because it's unstable
Configuration menu - View commit details
-
Copy full SHA for 7c7a821 - Browse repository at this point
Copy the full SHA 7c7a821View commit details -
Merge pull request #840 from Cyan4973/avx_warning
reworded the AVX error message when compiling xxh_x86dispatch.c
Configuration menu - View commit details
-
Copy full SHA for b3c34c8 - Browse repository at this point
Copy the full SHA b3c34c8View commit details -
Merge pull request #838 from t-mat/feature/issue-836
Feature: Support # comment lines in --check files
Configuration menu - View commit details
-
Copy full SHA for 34fdc93 - Browse repository at this point
Copy the full SHA 34fdc93View commit details
Commits on Jun 20, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 5b45185 - Browse repository at this point
Copy the full SHA 5b45185View commit details
Commits on Jun 25, 2023
-
Merge pull request #723 from ilya-fedin/fix-cmake-prefix
Fix double prefix when building with cmake
Configuration menu - View commit details
-
Copy full SHA for 17bbdee - Browse repository at this point
Copy the full SHA 17bbdeeView commit details -
added a cmake pkgconfig generation test
ensure proper path after advanced example provided by @ilya-fedin : #723 (comment) note : is expected to fail before integration of #723
Configuration menu - View commit details
-
Copy full SHA for e0f1c02 - Browse repository at this point
Copy the full SHA e0f1c02View commit details -
Merge pull request #845 from Cyan4973/pkgcmaketest
added a cmake pkgconfig generation test
Configuration menu - View commit details
-
Copy full SHA for 10fed7e - Browse repository at this point
Copy the full SHA 10fed7eView commit details -
Configuration menu - View commit details
-
Copy full SHA for 74fdef1 - Browse repository at this point
Copy the full SHA 74fdef1View commit details -
Configuration menu - View commit details
-
Copy full SHA for 1a9e4b9 - Browse repository at this point
Copy the full SHA 1a9e4b9View commit details -
Merge pull request #846 from Cyan4973/riscv
added RISC-V emulation to CI
Configuration menu - View commit details
-
Copy full SHA for 811212c - Browse repository at this point
Copy the full SHA 811212cView commit details
Commits on Jun 29, 2023
-
Merge pull request #750 from adrien1018/dev
Add algorithm specification of XXH3
Configuration menu - View commit details
-
Copy full SHA for 0656ed7 - Browse repository at this point
Copy the full SHA 0656ed7View commit details -
minor follow-ups for XXH3 specification (#750)
minor nit and clarifications, minor source modification (constant name correspond to spec), update version, attribute credit to @adrien1018.
Configuration menu - View commit details
-
Copy full SHA for d66a9cb - Browse repository at this point
Copy the full SHA d66a9cbView commit details -
Merge pull request #847 from Cyan4973/specxxh3followup
minor follow-ups for XXH3 specification (#750)
Configuration menu - View commit details
-
Copy full SHA for 4c9e996 - Browse repository at this point
Copy the full SHA 4c9e996View commit details
Commits on Jun 30, 2023
-
Configuration menu - View commit details
-
Copy full SHA for dfd7617 - Browse repository at this point
Copy the full SHA dfd7617View commit details -
Since we've merged #841 unfortunately, we can remove it for now.
Configuration menu - View commit details
-
Copy full SHA for 2c74110 - Browse repository at this point
Copy the full SHA 2c74110View commit details -
Configuration menu - View commit details
-
Copy full SHA for ff0fa82 - Browse repository at this point
Copy the full SHA ff0fa82View commit details -
Merge pull request #848 from t-mat/fix/remove-remaining-0.8.1s
Remove `0.8.1`s
Configuration menu - View commit details
-
Copy full SHA for 1d4ad8a - Browse repository at this point
Copy the full SHA 1d4ad8aView commit details
Commits on Jul 1, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 0d0e823 - Browse repository at this point
Copy the full SHA 0d0e823View commit details
Commits on Jul 2, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 0da2381 - Browse repository at this point
Copy the full SHA 0da2381View commit details
Commits on Jul 3, 2023
-
Note : SVE toolchain requires at least gcc version 10.1
Configuration menu - View commit details
-
Copy full SHA for 2d62c3e - Browse repository at this point
Copy the full SHA 2d62c3eView commit details -
Configuration menu - View commit details
-
Copy full SHA for 5bf31c2 - Browse repository at this point
Copy the full SHA 5bf31c2View commit details -
Configuration menu - View commit details
-
Copy full SHA for 1b8956d - Browse repository at this point
Copy the full SHA 1b8956dView commit details -
Merge pull request #849 from Cyan4973/sve_display
Detect and display SVE when active
Configuration menu - View commit details
-
Copy full SHA for 17d4aca - Browse repository at this point
Copy the full SHA 17d4acaView commit details
Commits on Jul 5, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 0578efd - Browse repository at this point
Copy the full SHA 0578efdView commit details -
Configuration menu - View commit details
-
Copy full SHA for 914458d - Browse repository at this point
Copy the full SHA 914458dView commit details -
Configuration menu - View commit details
-
Copy full SHA for b321462 - Browse repository at this point
Copy the full SHA b321462View commit details -
fix: add special #ifndef block for XXH_DOXYGEN
This special block prevents the following strange warning message: > warning: unable to resolve reference to 'XXH_INLINE_ALL' for \ref command Since Doxyfile explicitly includes xxh_x86dispatch.c, we need special care for multiple/inconsistent document definition.
Configuration menu - View commit details
-
Copy full SHA for e2cf439 - Browse repository at this point
Copy the full SHA e2cf439View commit details -
add: document for XXH3_*State()
Add document for XXH3_createState() and XXH3_freeState().
Configuration menu - View commit details
-
Copy full SHA for f7a6aca - Browse repository at this point
Copy the full SHA f7a6acaView commit details -
Configuration menu - View commit details
-
Copy full SHA for 28f40f3 - Browse repository at this point
Copy the full SHA 28f40f3View commit details -
Configuration menu - View commit details
-
Copy full SHA for 6aad58a - Browse repository at this point
Copy the full SHA 6aad58aView commit details -
fix: Remove ref for XXH_X86DISPATCH
Since XXH_X86DISPATCH is an internal macro, we can remove it from the document for now.
Configuration menu - View commit details
-
Copy full SHA for 20bf712 - Browse repository at this point
Copy the full SHA 20bf712View commit details
Commits on Jul 9, 2023
-
Add `--binary` to xxhsum. It's just ignored. Note that we don't have `--text`. See also #812
Configuration menu - View commit details
-
Copy full SHA for d7dce18 - Browse repository at this point
Copy the full SHA d7dce18View commit details -
Merge pull request #851 from t-mat/fix/doxygen-warnings-2023-0705-2
Fix doxygen warnings (2nd attempt)
Configuration menu - View commit details
-
Copy full SHA for a37d212 - Browse repository at this point
Copy the full SHA a37d212View commit details -
Merge pull request #852 from t-mat/feature/cli-option-binary
feature: CLI: --binary
Configuration menu - View commit details
-
Copy full SHA for 3f255fa - Browse repository at this point
Copy the full SHA 3f255faView commit details
Commits on Jul 10, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 4f092f2 - Browse repository at this point
Copy the full SHA 4f092f2View commit details -
Merge pull request #825 from easyaspi314/wasm128
Add WebAssembly SIMD128 implementation and Node.JS support
Configuration menu - View commit details
-
Copy full SHA for c234b94 - Browse repository at this point
Copy the full SHA c234b94View commit details -
Configuration menu - View commit details
-
Copy full SHA for 258351b - Browse repository at this point
Copy the full SHA 258351bView commit details -
- Add `--ignore-missing` option for `xxhsum`. - Add test script `tests/cli-ignore-missing.sh` and related `Makefile` entries. - Note that we still don't have CLI test entry in our CI. `xxhsum` returns failure exit code when - `--ignore-missing` is presented. - Checksum file is specified. - But `XSUM_checkFile()` doesn't read actual file (may) due to `--ignore-missing` If so, `xxhsum` reports `no file was verified` with failure exit code. <details><summary>`--ignore-missing` related code in `digest.c`</summary> - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L190-L191 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L362 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L380 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L506 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L978-L982 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L1236-L1240 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L1325-L1328 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L1429-L1431 - https://github.com/coreutils/coreutils/blob/2f1cffe07ab0f0b4135a52d95f1689d7fc7f26c9/src/digest.c#L1558-L1564 </details>
Configuration menu - View commit details
-
Copy full SHA for e592478 - Browse repository at this point
Copy the full SHA e592478View commit details -
Add missing doxygen style comments to XXH64 family
All comments are copied from correspond XXH32 functions.
Configuration menu - View commit details
-
Copy full SHA for 946d46d - Browse repository at this point
Copy the full SHA 946d46dView commit details -
Add doxygen comments to XXH3_64bits_ family
Add missing doxygen style comments to XXH3_64bits_ family.
Configuration menu - View commit details
-
Copy full SHA for 58d6f48 - Browse repository at this point
Copy the full SHA 58d6f48View commit details -
Configuration menu - View commit details
-
Copy full SHA for c4d049b - Browse repository at this point
Copy the full SHA c4d049bView commit details -
Configuration menu - View commit details
-
Copy full SHA for 17dfca6 - Browse repository at this point
Copy the full SHA 17dfca6View commit details -
Configuration menu - View commit details
-
Copy full SHA for bfe162c - Browse repository at this point
Copy the full SHA bfe162cView commit details -
Generate with the following commands: rm cli/xxhsum.1 make man man cli/xxhsum.1
Configuration menu - View commit details
-
Copy full SHA for e88af46 - Browse repository at this point
Copy the full SHA e88af46View commit details -
Configuration menu - View commit details
-
Copy full SHA for b6103a4 - Browse repository at this point
Copy the full SHA b6103a4View commit details -
Configuration menu - View commit details
-
Copy full SHA for c45304f - Browse repository at this point
Copy the full SHA c45304fView commit details -
Generate with the following commands: rm cli/xxhsum.1 make man man cli/xxhsum.1
Configuration menu - View commit details
-
Copy full SHA for 4a47319 - Browse repository at this point
Copy the full SHA 4a47319View commit details
Commits on Jul 12, 2023
-
Merge pull request #853 from t-mat/feature/cli-ignore-missing
Add --ignore-missing
Configuration menu - View commit details
-
Copy full SHA for 9391c17 - Browse repository at this point
Copy the full SHA 9391c17View commit details -
Merge pull request #855 from t-mat/fix/xxh64-doxygen-comments
Add missing doxygen style comments to XXH64 family
Configuration menu - View commit details
-
Copy full SHA for 5e95c08 - Browse repository at this point
Copy the full SHA 5e95c08View commit details -
Merge pull request #856 from t-mat/fix/xxh3-64bits-doxygen-documents
Add missing doxygen style comments to XXH3 family
Configuration menu - View commit details
-
Copy full SHA for bea4423 - Browse repository at this point
Copy the full SHA bea4423View commit details -
Merge pull request #857 from t-mat/fix/ignore-missing-manpages
Add --ignore-missing and --binary to --help and xxhsum.1
Configuration menu - View commit details
-
Copy full SHA for 0ac3a27 - Browse repository at this point
Copy the full SHA 0ac3a27View commit details -
added a compilation warning message for XXH_OLD_NAMES
as suggested by @t-mat, warning about the upcoming deprecation of this feature.
Configuration menu - View commit details
-
Copy full SHA for b203110 - Browse repository at this point
Copy the full SHA b203110View commit details
Commits on Jul 14, 2023
-
Merge pull request #860 from Cyan4973/old_names_warning
added a compilation warning message for XXH_OLD_NAMES
Configuration menu - View commit details
-
Copy full SHA for 4fd75d7 - Browse repository at this point
Copy the full SHA 4fd75d7View commit details
Commits on Jul 15, 2023
-
fix: __cpuidex pointer type warning
clang-cl warns the first argument of __cpuidex() because it's int[]. void __cpuidex( int cpuInfo[4], int function_id, int subfunction_id ); https://learn.microsoft.com/en-us/cpp/intrinsics/cpuid-cpuidex
Configuration menu - View commit details
-
Copy full SHA for 5953419 - Browse repository at this point
Copy the full SHA 5953419View commit details -
Merge pull request #864 from t-mat/fix/clang-cl-sign-warning
fix: __cpuidex pointer type warning
Configuration menu - View commit details
-
Copy full SHA for f5cbcb8 - Browse repository at this point
Copy the full SHA f5cbcb8View commit details -
clang-cl defines both __clang__ and _MSC_VER. Also it seems clang-cl needs extra headers to compile AVX2/512F intrinsic.
Configuration menu - View commit details
-
Copy full SHA for 1e5c0f5 - Browse repository at this point
Copy the full SHA 1e5c0f5View commit details -
fix: support clang-cl and dispatch
- Actual platform name of x64 for MSVC/clang-cl is "AMD64" - When we compile xxhsum with dispatch, we need xxh_x86dispatch.c
Configuration menu - View commit details
-
Copy full SHA for d82321b - Browse repository at this point
Copy the full SHA d82321bView commit details -
Merge pull request #865 from t-mat/fix/clang-cl-dispatch
Fix clang-cl dispatch
Configuration menu - View commit details
-
Copy full SHA for 1a53004 - Browse repository at this point
Copy the full SHA 1a53004View commit details -
add a test combining Windows + Visual + Clang-Cl + DISPATCH=1
test that #763 is fixed and remains fixed
Configuration menu - View commit details
-
Copy full SHA for a8ded18 - Browse repository at this point
Copy the full SHA a8ded18View commit details -
Merge pull request #863 from Cyan4973/visual_clang_dispatch_test
add a test combining Windows + Visual + Clang-Cl + DISPATCH=1
Configuration menu - View commit details
-
Copy full SHA for 0c8e930 - Browse repository at this point
Copy the full SHA 0c8e930View commit details -
Configuration menu - View commit details
-
Copy full SHA for 8b093f2 - Browse repository at this point
Copy the full SHA 8b093f2View commit details -
Merge pull request #866 from t-mat/fix/vc2022-clang-cl
fix: enable clang-cl test for VC++2022
Configuration menu - View commit details
-
Copy full SHA for a112385 - Browse repository at this point
Copy the full SHA a112385View commit details -
fix: Intel compiler's warning for XXH_ASSERT()
Intel compiler expects "unsigned char" as an argument of __builtin_assume().
Configuration menu - View commit details
-
Copy full SHA for ee51543 - Browse repository at this point
Copy the full SHA ee51543View commit details
Commits on Jul 16, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 97be68d - Browse repository at this point
Copy the full SHA 97be68dView commit details -
Configuration menu - View commit details
-
Copy full SHA for 0f31a5f - Browse repository at this point
Copy the full SHA 0f31a5fView commit details -
Merge pull request #867 from t-mat/fix/intel-compiler-assert
fix: Intel compiler's warning for XXH_ASSERT()
Configuration menu - View commit details
-
Copy full SHA for 9e1cf4e - Browse repository at this point
Copy the full SHA 9e1cf4eView commit details -
add: introduce "public" and "internal" Doxyfile
"Doxyfile" is configured to generate public document. Differences are: - INTERNAL_DOCS - EXTRACT_STATIC - EXTRACT_PRIVATE
Configuration menu - View commit details
-
Copy full SHA for 4b93d0a - Browse repository at this point
Copy the full SHA 4b93d0aView commit details -
fix: hide x86 dispatcher internals from document
Hide internals to simplify public document.
Configuration menu - View commit details
-
Copy full SHA for 7031f87 - Browse repository at this point
Copy the full SHA 7031f87View commit details -
Merge pull request #868 from t-mat/fix/hide-internal-macros
fix: hide internal macros from doxygen
Configuration menu - View commit details
-
Copy full SHA for 557799a - Browse repository at this point
Copy the full SHA 557799aView commit details -
Merge pull request #869 from t-mat/fix/hide-x86-dispatcher-internals
hide x86 dispatcher internals
Configuration menu - View commit details
-
Copy full SHA for d589159 - Browse repository at this point
Copy the full SHA d589159View commit details
Commits on Jul 19, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 2352ef1 - Browse repository at this point
Copy the full SHA 2352ef1View commit details -
revert XXH3's 17to128 variant to single accumulator
and re-employ original nest order to reproduce same performance as v0.8.1.
Configuration menu - View commit details
-
Copy full SHA for 2180439 - Browse repository at this point
Copy the full SHA 2180439View commit details -
Merge pull request #872 from t-mat/dev
CI: Add gcc-13 and clang-{13,14,15}
Configuration menu - View commit details
-
Copy full SHA for d507414 - Browse repository at this point
Copy the full SHA d507414View commit details -
Merge pull request #871 from Cyan4973/revert_17to128
revert XXH3's 17to128 variant to single accumulator
Configuration menu - View commit details
-
Copy full SHA for c8060b4 - Browse repository at this point
Copy the full SHA c8060b4View commit details
Commits on Jul 20, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 72e69d3 - Browse repository at this point
Copy the full SHA 72e69d3View commit details