profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/davidbolvansky/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.
Dávid Bolvanský davidbolvansky Pribylina, SK / Brno, CZ

llvm/llvm-project 9669

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies. Note: the repository does not accept github pull requests at this moment. Please submit your patches at http://reviews.llvm.org.

nesfit/fitcrack 26

A hashcat-based distributed password cracking system

davidbolvansky/TINZbierkaPrikladov 10

Študentská zbierka príkladov pre predmet Teoretická informatika

davidbolvansky/FLP-simplify-bkg-tester 2

Based on https://github.com/xhrebi04/flp_simplify_bkg_test. Added many new tests.

davidbolvansky/FITVUTHistogramySkusok 0

Histogramy skúšok na FIT VUT

davidbolvansky/hashcat 0

World's fastest and most advanced password recovery utility

davidbolvansky/ICP-Solitaire 0

Solitaire in C++

davidbolvansky/KNBH 0

Dormitory Girl Hunter A.K.A KNBH

startedahrm/sioyek

started time in 4 days

startedtiangolo/fastapi

started time in 18 days

CommitCommentEvent

issue openedfacebook/zstd

Significant compression speed diffs between Clang and GCC in long mode

As noted in https://www.phoronix.com/scan.php?page=article&item=clang12-gcc11-icelake&num=2, there is a huge difference in compression speed between clang and gcc in long mode.

image

Standard mode looks kinda fine.

image

Missed vectorization? Any chance to improve it?

created time in 2 months

startedalvinwan/tex2py

started time in 2 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 62fcda93788a93726f7bc7cb5c0699595e009ceb

Revert "[InstSimplify] Added tests for PR50173, NFC" This reverts commit 4e7a4c73dab6605f4fcc7bf09c2ee85e7925f6d7. Not needed, pattern is handled by instcombine already.

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 4e7a4c73dab6605f4fcc7bf09c2ee85e7925f6d7

[InstSimplify] Added tests for PR50173, NFC

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 80b897e21bf0ac56b04d415cf9bf671f81a84416

[InstCombine] ctpop(X) ^ ctpop(Y) & 1 --> ctpop(X^Y) & 1 (PR50094) Original pattern: (__builtin_parity(x) ^ __builtin_parity(y)) LLVM rewrites it as: (__builtin_popcount(x) ^ __builtin_popcount(y)) & 1 Optimized form: __builtin_popcount(X^Y) & 1 Alive proof: https://alive2.llvm.org/ce/z/-GdWFr Reviewed By: RKSimon Differential Revision: https://reviews.llvm.org/D101802

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 88ca010cc1854059c9e4a668750cd981e44238bc

[InstCombine] Added tests for PR50094, NFC

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 08c08577f94fd7efd2a670d0e5c3c6f72882fa0a

[InstCombine] cttz(sext(x)) -> cttz(zext(x)) ``` ---------------------------------------- define i32 @src(i16 %x, i1 %b) { %0: %z = sext i16 %x to i32 %p = cttz i32 %z, %b ret i32 %p } => define i32 @tgt(i16 %x, i1 %b) { %0: %z = zext i16 %x to i32 %p = cttz i32 %z, %b ret i32 %p } Transformation seems to be correct! ``` https://alive2.llvm.org/ce/z/evomeg Reviewed By: spatel Differential Revision: https://reviews.llvm.org/D101764

view details

push time in 3 months

startedsharkdp/hyperfine

started time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 27b651ca473ba33999888cc302efb5f856ec7cea

[InstCombine] cttz(zext(x)) -> zext(cttz(x)) if the 'ZeroIsUndef' parameter is 'true' (PR50172) Zext doesn't change the number of trailing zeros, so narrow cttz(zext(x)) -> zext(cttz(x)) if the 'ZeroIsUndef' parameter is 'true'. Proofs: https://alive2.llvm.org/ce/z/o2dnjY Solves https://bugs.llvm.org/show_bug.cgi?id=50172 Reviewed By: spatel Differential Revision: https://reviews.llvm.org/D101582

view details

push time in 3 months

startedintel/yarpgen

started time in 3 months

startedinducer/cgen

started time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 2af95a527580333940333c532679356d99b88e3d

[X86] Promote 16-bit CTTZ_ZERO_UNDEF to 32-bit variant Related to PR50172. Protects us against regressions after we will start doing cttz(zext(x)) -> zext(cttz(x)) transformation in the middle-end. Reviewed By: craig.topper Differential Revision: https://reviews.llvm.org/D101662

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha ad12590fdd1459ffb8db185ea1b45748c20b538e

[InstCombine] Added tests for PR50172, NFC

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha e20b32ff3b029f50c7237f49e5e5c97bd47ea540

[BuildLibCalls] Remove inaccessiblememonly inference for calloc Solves regression mentioned in PR50143. As noted in D101440, proper modelling for calloc would require new attribute inaccessible_or_returned_memonly.

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha e81819377ececef53d3e90743eeb44e6a089b65a

[DSE] Eliminate zero memset after calloc Solves PR11896 As noted, this can be improved futher (calloc -> malloc) in some cases. But for know, this is the first step. Reviewed By: nikic Differential Revision: https://reviews.llvm.org/D101391

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 87fc97169e7901dfe56ca0da0d92da0c02d2ef48

[DSE] Added testcases for 11896, NFC

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 942d2e19e11d6d65bf147ecab8513b7e8bd5d7e4

Revert "Only ignore -Wdeprecated-copy if the used compiler supports the warning" This reverts commit b125c034af89e32d5258d593080ff0c4810776a0. Some bots say " missing binary operator before token "(" "

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha b125c034af89e32d5258d593080ff0c4810776a0

Only ignore -Wdeprecated-copy if the used compiler supports the warning This is needed after https://reviews.llvm.org/rG9658d045926545e62cc3f963fe611d7c5d0c9d98 which introduced code that at least didn't compile clean with clang 8. Reviewed By: xbolva00 Differential Revision: https://reviews.llvm.org/D101288

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 691badc3d6a76ceb98530d58f80c11565337d767

[InstCombine] C - ctpop(a) - > ctpop(~a)) if C is bitwidth (PR50104) Proof: https://alive2.llvm.org/ce/z/mncA9K Solves https://bugs.llvm.org/show_bug.cgi?id=50104 Reviewed By: spatel Differential Revision: https://reviews.llvm.org/D101257

view details

push time in 3 months

push eventllvm/llvm-project

Dávid Bolvanský

commit sha 819d35e6e916ef6c7cc267291aa9307c79a3da81

[InstCombine] Added testcases for PR50104, NFC

view details

push time in 3 months

startedpinetr2e/napkin

started time in 3 months

pull request commentprotocolbuffers/upb

Fast table-driven parsing for upb (2+GB/s)

Related to your article and "jne followed by jmp" issue - LLVM does it intentionally this way: https://llvm.org/doxygen/BranchFolding_8cpp_source.html

line 1521

haberman

comment created time in 3 months