implement new
@bitCast semantics, change LLVM backend integer lowering
@linus wrote in ziglang/zig#35711 (comment):
math.big.int: add benchmark and performance improvement
cd5dc27e6a
math.big.benchmark: remove stat file and python graphs
6683b2e4fa
math.big: only use the larger llmulLimb on ReleaseFast
a8111198ad
math.big: only use the larger llmulLimb on ReleaseFast
b32d9a11b4
math.big: better codegen for llmulLimb
94b5c8fd28
math.big: replace builtin @add/subWithOverflow
7ec8a5aac5
math.big: replace ll<add/sub>carry for better codegen
1857310764
math.big: better codegen for llmulLimb
2a2f547c24
math.big: replace builtin @add/subWithOverflow
e60f91babd
math.big: replace ll<add/sub>carry for better codegen
88b9a1c57c
math.big: better codegen for llmulLimb
d5a1345a73
math.big: replace builtin @add/subWithOverflow
c275239555
math.big: replace ll<add/sub>carry for better codegen
9869afd686
math.big: finalize benchmark
cebf5ba3d3
math.big: better codegen for llmulLimb
aa5610df81
math.big: replace builtin @add/subWithOverflow
9b870d04c9
math.big: better codegen for llmulLimb
8b0807c966
math.big: replace builtin @add/subWithOverflow
c0af8b7f5d
math.big: replace ll<add/sub>carry for better codegen
fad5615198
math.big: finalize benchmark
0dedd4dbd1
math.big benchmark: add visualization script
550a768200
math.big: fix minor mistake
9cc53749af
math.big: add benchmark script
edf1c4330b
math.big: soften assert
d8e7f7a5dc
math.big: move the most basic functions to raw.zig
125379d178
math.big: better codegen for llmulLimb
dac11020e7
math.big: better codegen for llmulLimb
09f29e8dc6
math.big: replace builtin
fa060bc70d
math.big: replace ll<add/sub>carry for better codegen
math.big.int: Improvements to division in the n by 1 case
Indeed, thanks
0036861f8f
math.big.int: better division and toString implementations
8b2544a008
math.big.int: use
math.Order for llcmp
facb1ad013
math.big: better codegen for llmulLimb
55264aacf3
math.big: replace builtin
a82ca3cee8
math.big: replace ll<add/sub>carry for better codegen