 svenbarth
|
8287773b16
Reverted revision 33036. This feature is too controverse to be left in.
|
9 anni fa |
|
svenbarth
|
ed94ca4b24
Add support for IfThen() instrinsic that works like the if-statement in that it evaluates only the expression that is indeed executed.
|
9 anni fa |
 Jonas Maebe
|
8917633199
* implement sizeof/typeof completely at the node level, based on
|
9 anni fa |
|
svenbarth
|
74c79803f7
Extend tlocalvarsym with the ability to not register it if needed.
|
9 anni fa |
 florian
|
70b8789262
o basic extended pascal support:
|
9 anni fa |
|
Jonas Maebe
|
2778bf55c7
* add "doregister" parameter to torddef.create, and don't register temporary
|
9 anni fa |
|
florian
|
f00a55ae08
+ tinlinenode.createintern
|
10 anni fa |
|
Jonas Maebe
|
67b8aceaee
* synchronized with privatetrunk till r30095
|
10 anni fa |
|
Jonas Maebe
|
68303b8df1
* add explicit check that argument of sizeof(x) is a simple load node (this
|
10 anni fa |
|
florian
|
0dfbba86f3
* properly firstpass nodes when inc(...) is converted into ...:=...+x, resolves #27471
|
10 anni fa |
|
florian
|
c9ba9665dd
* set nf_internal for if nodes generated by assert, prevent warning for assert(true); resolves #26467
|
10 anni fa |
 nickysn
|
a9e1419ee3
* extracted the pass 1 handling of inc/dec to a virtual method first_IncDec
|
10 anni fa |
 pierre
|
a77735d7af
Fix bug in in_const_swap_qword handling
|
10 anni fa |
|
Jonas Maebe
|
e8026ad8d3
* fixed the resultdef of length() intrinsic: it's ossinttype rather than
|
11 anni fa |
 sergei
|
04d8e8a5dc
* On 64-bit targets, handle abs(int64) internally, using the same code as for abs(longint), i.e. without branching. Both generic and x86-specific pass 2 code is already suitable for different operand sizes, only type checking needs removal of excessive conversions to 32 bits.
|
11 anni fa |
|
svenbarth
|
34394d6925
Fix for Mantis #26180. Accept undefineddef as first parameter type of an Assert if it is used inside a generic.
|
11 anni fa |
|
sergei
|
274c2fc82f
* handle_ln_const: if/else branches were swapped, correct values are: ln(0)=-Inf and ln(-#)=NaN.
|
11 anni fa |
|
sergei
|
743326dc70
* abs(): when translated to generic fpc_abs_real() helper, typecast result back to original precision, similar to how it is done for sqr and sqrt. Fixes compilation on m68k after r27808 (m86k looks like the only target still needing this generic handling).
|
11 anni fa |
|
sergei
|
37625211f5
* Fixed sqrt procedure selection on arm-wince after r27809.
|
11 anni fa |
|
sergei
|
22e099d000
* sqrt(real): for targets with emulated floating point, invoke float64_sqrt or float32_sqrt from softfpu code. Testing on ARM CPU shows that float64_sqrt executes about twice faster and offers better accuracy than fpc_sqrt_real from genmath.inc.
|
11 anni fa |
|
sergei
|
fc5f45f65c
* sqr(real) and sqrt(real): remove typeconv node inserted by initial call processing (see explanation in comments), allowing these functions to be evaluated using precision of argument. In particular, sqrt(single) and sqrt(double) now emit 'sqrtss' and 'sqrtsd' instructions on x86 targets with -Cfsse3. Non-x86 targets already have the necessary support in code generators.
|
11 anni fa |
|
florian
|
d88d644925
+ support for FMA intrinsic: if there is no hardware support, the compiler throws an error.
|
11 anni fa |
|
svenbarth
|
256c852631
Fix for Mantis #23776.
|
11 anni fa |
|
Jonas Maebe
|
d452686c39
* moved pbestrealtype from symdef to symcpu
|
11 anni fa |
|
Jonas Maebe
|
b57c95043f
+ support overriding tdef/tsym methods with target-specific functionality:
|
11 anni fa |
|
Jonas Maebe
|
0edb7b2288
* fixed r27320 for darwin, as its mangled names get prefixed with '_'
|
11 anni fa |
|
svenbarth
|
a08e8ff982
Fix for Mantis #25929
|
11 anni fa |
|
florian
|
ab3f5744e0
* skip dec/inc type checking in unspecialized generic methods, resolves #25603
|
11 anni fa |
|
florian
|
e841027a48
* fixes handling of typed files in iso mode
|
11 anni fa |
|
nickysn
|
e1332304ef
* when converting Hi() to a shr node, convert the right shr const parameter to
|
11 anni fa |
