 Pierre Muller
|
eae74aaee0
Follow sdas documentation for .ascii limitations by using only printable chars and .byte for others
|
1 anno fa |
|
Pierre Muller
|
53bf5dc6ef
Avoid wrong typecast by adding check that p1 is an instuction before casting it to taicpu
|
1 anno fa |
 J. Gareth "Curious Kit" Moreton
|
271f69db12
* arm/a64: "OptPass2AND" and "OptPass2CMP" adapted to remove vestigial CMP and TST instructions
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
6b2e4fa413
* arm: "OpCmp2OpS" moved to Pass 2 so it doesn't conflict with AND; CMP -> TST optimisation
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
bf29f2051c
* arm/a64: Added new TST post-peephole optimisation to replace previous AND/CMP/B(c) optimisation
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
b18c10d0d8
* arm/a64: New "OptPass2TST" routine to catch "TST; B.c; AND -> ANDS; B.c" optimisation
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
9f19f582c4
* arm/a64: New AND/CMP -> TST or ANDS optimisation
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
29916bc6f6
* arm: Fixed "RegInInstruction" and "RegModifiedByInstruction" not handling the flags properly
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
38d2f3d58c
* a64: Renamed OptPostCMP/And to PostPeepholeOptCMP/AND for internal consistency
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
d7d4b446c1
* FindRegAllocBackward will now return nil if it hits a dealloc for the register first
|
2 anni fa |
|
J. Gareth "Curious Kit" Moreton
|
1fafcbc125
* New "TryRemoveRegAlloc" optimisation utility
|
1 anno fa |
 florian
|
4f82fade82
* extend LeaCallLeaRet2Jmp optimization to work on windows
|
1 anno fa |
|
florian
|
e7600ee245
* patch by Rika: Small improvements to (a|u)strings.inc, resolves #40364
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
62495c964a
* x86: New "aoc_DoPass2JccOpts" option to catch branch and
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
2055ef4e88
* x86: STC/CLC optimisations can now handle multiple Jcc instructions
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
de03a1b894
* x86: Additional STC/CLC; J(N)C deterministic jump optimisation
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
698d432cce
* x86: Separated out Jcc; STC/CLC optimisation and added additional peephole optimisations
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
3e06242fd8
* x86: New "STC/CLC; MOV" peephole optimisation
|
1 anno fa |
|
J. Gareth "Curious Kit" Moreton
|
d06dd2b0ea
* x86: New "Jcc; CLC/STC; .Lbl" optimisation
|
1 anno fa |
 Rika Ichinose
|
c29dd86bb2
Remove runtime ABI adapter in x86_64.inc:IndexByte/Word, and save two jumps in the common case.
|
1 anno fa |
|
florian
|
2575cbc439
* complain about turned off support of c style operators only in the parser,
|
1 anno fa |
|
Rika Ichinose
|
7bf502ad40
Change Mov*DQ to Mov*PS; they are always equivalent because no operations but the memory transfers are performed, and 1 byte shorter each.
|
1 anno fa |
|
Rika Ichinose
|
12f18177ae
Simplify x86_64.inc:Move non-temporal loops, and adjust thresholds for move distances considered too short for NT.
|
1 anno fa |
|
Rika Ichinose
|
0b5998ee8b
Write two last values after 2× loops unconditionally instead of an extra check.
|
1 anno fa |
|
Rika Ichinose
|
e395166cb7
Check for Move overlaps in more obvious way (that also does no jumps in forward case).
|
1 anno fa |
|
florian
|
8a48d1bbbc
* clean up of arrayconstructor_to_set, the stringdef case didn't work for years and is not tested, so it probably is not supposed to be there at all
|
1 anno fa |
 Michaël Van Canneyt
|
cbbcc4356a
* Make color distinct from cardinal
|
1 anno fa |
 marcoonthegit
|
78eb340a4f
* change elfid to 1400097 aka FreeBSD 14.0
|
1 anno fa |
|
florian
|
bc0d2293a1
* stricter type checking for set constants, resolves #40631
|
1 anno fa |
 Sven/Sarah Barth
|
43721f21c4
* fix #40621: when checking for visibility of members then always use the genericdefs for specializations for correct scoping
|
1 anno fa |
