aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAgeFilesLines
...
* switch to whitelist based .gitignorelemon2026-03-161-14/+29
|
* Update READMElemon2026-03-161-3/+1
|
* fix Makefilelemon2026-03-161-3/+3
|
* regalloc: improve spilling and cleanuplemon2026-03-164-354/+401
| | | | | | | | Instead of spilling current interval when running out of registers, spill the longest-lived active interval with a lower spill cost than current. Spill costs are estimated based on multiplicative loop depth. Also cleanup regalloc.c somewhat. Update todo.txt too.
* ir: blk loop index for spill costlemon2026-03-163-4/+39
|
* Makefile: add clean-toollemon2026-03-161-2/+5
|
* eval: fix errata with ELOGIOR, also rename lhs & rhslemon2026-03-151-78/+77
|
* c: default enum to unsigned int backing typelemon2026-03-151-1/+3
| | | | Matches GCC behavior.
* build system: use custom depgen instead of gcclemon2026-03-154-5/+170
|
* c-testsuite.shlemon2026-03-141-2/+4
|
* eval & fold: handle division overflow for MIN/-1lemon2026-03-142-2/+9
|
* cpp: better handling for #include file not foundlemon2026-03-141-12/+20
| | | | | | Instead of attempting to keep going (since things will probably break due to missing declarations), just report any other #include errors and abort.
* make clean rm --lemon2026-03-141-1/+1
|
* c: static eval refactoringlemon2026-03-148-279/+426
| | | | | Explicit node for static symbol (addresses). Should not break with edge cases like the previous ad-hoc approach. And some other bugfixes
* x86_64/isel: fix edge case with branch on float add resultlemon2026-03-112-1/+6
| | | | | | | | In `if (x + 1)` the implicit `!= 0` can be omitted for integers, because the x86 `add` instruction sets the zero flag accordingly. But for floats it doesn't, so applying that optimization there was wrong. Luckily it was caught by a nullptr dereference later in `emit::Xjcc()` for the (missing) unordered branch target, instead of miscompiling.
* c: relax constexpr constraints, fix alignoflemon2026-03-093-30/+62
| | | | | | | | | | | | - Allow short-circuiting of constant logical expressions where the unevaluated operand is not a constant expression (`1 || 0/0`) - Allow constant integer expressions that evaluate to zero to be used as null pointer constants (GNU extension). + According to the standard, `int *x = 5*0;` should be rejected. But compilers evaluate `5*0 -> 0` and allow it as if a null pointer literal.
* 0.1.10lemon2026-03-081-2/+2
|
* fix some more cases of signed integer overflowlemon2026-03-082-3/+3
|
* x86_64/emit: fix float store zero, unsafe range check for mul immediatelemon2026-03-081-1/+5
|
* ir/fold: fix float conversion overflowslemon2026-03-081-4/+14
|
* ir: fix inlining getting stuck on complex recursive call sequenceslemon2026-03-082-8/+30
| | | | | By maintaining a proper stack of inline expansions and refusing to expand callers already present in the stack.
* x86_64/isel: cleanup and fix edge caseslemon2026-03-082-79/+76
|
* ir: increase per-function stack size limitlemon2026-03-061-1/+1
|
* cpp: #include MACRO, free tkslemon2026-03-061-0/+1
|
* c: recognize another static initializer edge caselemon2026-03-051-3/+2
| | | | | | I think globsym & expr2reloc should be rewritten to be more robust to this kind of thing. Accumulate offsets whilst recursing into the child node that might contain a direct global symbol reference kind of thing.
* cpp: get rid of unnecessary ugly global variable for lexing header namelemon2026-03-051-31/+23
| | | | Make it a parameter.
* cpp: implement #include MACRO_NAMElemon2026-03-051-55/+98
|
* cpp: #include".." shouldn't look in working directorylemon2026-03-051-6/+8
|
* c: do not try to eval conditional expression's not taken branchlemon2026-03-052-2/+2
|
* ccopt.dbg.any -> uintlemon2026-03-041-1/+1
| | | | So it encompasses every bitfield.
* ir/dump: fix dumping data in .textlemon2026-03-041-4/+16
|
* regalloc: fix minor regression with storing of regs into spilled stack slotslemon2026-03-041-1/+1
|
* driver: fix -pthread, -Ulemon2026-03-022-2/+6
|
* add bswap16/32/64lemon2026-03-0210-37/+125
| | | | | - frontend: __builtin_bswapX intrinsics - backend: ObswapX instructions
* cpp: fix token spacing edge case when empty macros are involvedlemon2026-03-012-3/+8
|
* cpp: fix token spacing (#2) (hopefully)lemon2026-03-013-35/+104
|
* 0.1.9lemon2026-02-281-2/+2
|
* ir: make address ref hash table resizablelemon2026-02-287-27/+49
| | | | Would hit the limit on very large functions (thanks csmith).
* Update README.mdlsof2026-02-281-1/+1
|
* c: static intializer bugfix introduced in6c3f1593e9lemon2026-02-281-0/+1
|
* aarch64/isel: handle more instructionslemon2026-02-281-13/+18
|
* backend: fix edgecases where imm32 immediate overflowslemon2026-02-282-3/+10
|
* regalloc: fix defn of spilled var with immediatelemon2026-02-282-2/+16
| | | | | | With `i64 %a = copy 0xabcdef12345`, if `%a` was spilled, this would turn into a `storei64` instruction with an unencondable immediate operand (larger than 32 bits in x86, non-zero in aarch64).
* cpp: allow comments between function-like macro name and argumentslemon2026-02-262-9/+29
|
* 0.1.8lemon2026-02-251-2/+2
|
* c: allow casting of symbol address to intptr_t in static initializerslemon2026-02-252-2/+4
|
* preprocessor: use shared static buffers for replacement listslemon2026-02-251-150/+178
| | | | | | | One for macro definitions' bodies, another one for function-like macros with substituted arguments in the macro expansion stack. Cuts down on a lot of individual allocations. #undef'd macros' replacement lists are "leaked" but that is very minor
* fix io fa8c39d6df : make buffer shared (for nested note())lemon2026-02-241-1/+1
|
* fix various warningslemon2026-02-246-14/+11
|
* inline: fix undefined value returnslemon2026-02-241-1/+2
| | | | | | Previously if an inlined function has a return statement with no value (control flow reaching the closing brace of the function), would use an invalid null reference in the inlined body. Turn it into undef.
generated by cgit v1.2.3 (git 2.39.1) at 2026-05-02 07:38:50 +0000