 Yuriy Sydorov
|
07dcd22b1c
* sizeof(array) and length(array) must return sizeUint in case the array
|
4 years ago |
 svenbarth
|
ca7c775e36
* fix for Mantis #38642: for enumerations with jumps Delphi behaves as follows:
|
4 years ago |
 Jonas Maebe
|
3e047d3691
* generate more efficient code for trunc(currency), trunc(comp), and
|
4 years ago |
 florian
|
6e4997265c
+ min/max intrinsics for internal use by the compiler
|
4 years ago |
|
florian
|
28d248c9d4
* cleanup merge overlefts
|
4 years ago |
|
florian
|
bcdfd07bdd
* if a variant is read/written, include the variants unit
|
4 years ago |
|
florian
|
637976e83f
* patch by Marģers to unify internal error numbers, resolves #37888
|
4 years ago |
|
svenbarth
|
cce2dad1bf
* fix for Mantis #37806: allow undefineddefs for Include() and Exclude()
|
4 years ago |
|
svenbarth
|
ff4e1345ff
* fix for Mantis #37779: use init RTTI instead of full RTTI when copying an array
|
4 years ago |
|
svenbarth
|
d3e18ccb5e
* fix for Mantis #37221: apply adjusted patch by Ondrej Pokorny to allow the use of Copy() on open array parameters (the result will be a dynamic array)
|
4 years ago |
|
florian
|
60bd9c5ba0
* write a proper error message if types are read/written which are not supported in iso mode, resolves #37763
|
4 years ago |
|
svenbarth
|
8770af84bd
* fix for Mantis #37187: inside generics the constant code in pexpr does not handle all cases and thus current_procinfo needs to be checked as well
|
5 years ago |
|
florian
|
f542ee7b92
* do not throw an internal error if an errorneous type is passed to str, resolves #37462
|
5 years ago |
|
florian
|
18877dc96e
* check if current_procinfo is assigned when creating a default sym to avoid a crash, resolves #37303
|
5 years ago |
|
svenbarth
|
d401639b24
* fix for Mantis #37251: apply patches by Bi0T1N to implement the IsConstValue() intrinsic
|
5 years ago |
|
florian
|
d2d2904d78
* make more use of is_typeparam
|
5 years ago |
|
svenbarth
|
3d2945726a
* handle generic parameters in Ord()
|
5 years ago |
 nickysn
|
b4ff7388df
+ retinparam fixes for exp, frac, int, cos, sin, arctan, ln, round and trunc
|
5 years ago |
|
nickysn
|
661db99581
+ ret in param fix in pass_typecheck for exp,frac,int,cos,sin,arctan,ln
|
5 years ago |
|
nickysn
|
5200ff5e49
+ ret in param fix for tinlinenode.first_sqrt_real
|
5 years ago |
|
nickysn
|
0b19d4235f
+ ret in param fix for tinlinenode.first_sqr_real
|
5 years ago |
|
nickysn
|
a1ccab5e23
+ handle properly ret in param in tinlinenode.first_abs_real
|
5 years ago |
|
nickysn
|
0e605eb30e
* fix for the handling of abs(real), sqr(real) and sqrt(real) when the result is returned in param
|
5 years ago |
 Jeppe Johansen
|
c20b27ede9
Add most SSE instructions as intrinsics.
|
5 years ago |
|
florian
|
3947922fea
* set nf_* correctly when an inc/dec node is converted into an add/sub node
|
5 years ago |
|
Jonas Maebe
|
b625afe800
* never inline routines that call get_frame (otherwise get_frame would return
|
5 years ago |
|
florian
|
70a836c4a2
* first part of merging parts of Jeppe's intrinsics patch, mainly r31135
|
5 years ago |
|
florian
|
f7a052f884
+ inline high(<dyn. array>), resolves #28536
|
5 years ago |
|
Jonas Maebe
|
d23c7980aa
* properly perform (or not) range/overflow checking when performing constant
|
5 years ago |
|
svenbarth
|
b0ac0ddcc9
+ add support for the IsManagedType() intrinsics which reports for a type (or variable) whether it needs initialization/finalization
|
5 years ago |
