 sergei
|
d613ab8578
* x86: improve x87 qword to float conversion, using single-precision constants saves space and removes need in separate load on FPU stack. No precision loss occurs because 2**64 is representable exactly even in single precision.
|
пре 11 година |
 florian
|
e81d2d1f3b
* basic avx support for floating point operations (use -Cfavx to activate)
|
пре 12 година |
 nickysn
|
f3ad9ce9ce
* i8086 fixes in tx86typeconvnode.second_int_to_bool; this fixes tcnvint2.pp
|
пре 12 година |
|
nickysn
|
b43c4c4aca
* i8086 fixes in tx86typeconvnode.second_int_to_real
|
пре 12 година |
 Jonas Maebe
|
edd42aa42a
* moved subsetref/reg and bit_set/test support from cgobj to hlcgobj for
|
пре 13 година |
|
florian
|
111d05c68f
o patch by Alexander Shishkin, resolves #20409
|
пре 14 година |
|
sergei
|
7d99f95c45
* Always create a section before emitting data to current_asmdata.asmlists[al_typedconsts]. Without it, such data ends up in sections created elsewhere, creating very non-obvious dependencies on other parts of compiler.
|
пре 14 година |
|
Jonas Maebe
|
3aea9b134a
* make sure that all references used in the inline int->real conversion code
|
пре 15 година |
|
Jonas Maebe
|
d1538ab023
o added ARM VPFv2/VFPv3 support:
|
пре 15 година |
 yury
|
f726e1691b
* Fixed warnings and notes.
|
пре 16 година |
|
Jonas Maebe
|
dbeb3e060f
* fixed internalerror(2007120903) when using -Cfsse2 in case an unsigned
|
пре 16 година |
|
Jonas Maebe
|
7d459cf12a
* the compiler now explicitly keeps track of the minimally guaranteed
|
пре 16 година |
|
Jonas Maebe
|
718694d1d6
* always specify an explicit alignment for tgobj.gettemp (so e.g.
|
пре 17 година |
|
yury
|
b70bf05ad5
* I hope it's proper fix for r10688.
|
пре 17 година |
|
yury
|
e71d631c24
* Fix for tw10233 for all CPUs.
|
пре 17 година |
|
yury
|
771479e65c
* Improved fix for bug #10233 for better Delphi compatibility and efficiency:
|
пре 17 година |
|
yury
|
cf235145cb
* Fixed a_load_reg_reg for arm.
|
пре 17 година |
|
Jonas Maebe
|
62e66d2a56
* explicit typecasting of any ordinal type to a boolean of the same
|
пре 17 година |
|
Jonas Maebe
|
f36e5411af
* split cpu64bit compiler define into
|
пре 17 година |
|
Jonas Maebe
|
8349cde7db
* changed byte/word/longbool to be Delphi-compatible (+ similar changes
|
пре 17 година |
|
Jonas Maebe
|
974c240266
* also use sse2 instructions to convert int to real on i386
|
пре 17 година |
|
florian
|
1cb5d2a603
* some missing make_simple_ref calls added, resolves #8948
|
пре 18 година |
 peter
|
ad944fd198
* move x86_64 specific int_to_real conversion to nx64cnv
|
пре 18 година |
|
peter
|
6b8aed593f
* remove registers{int/mmx/fpu} from firstpass
|
пре 18 година |
|
florian
|
9a0f769b2f
* cleaned up set conversion
|
пре 18 година |
|
florian
|
05e0a804cc
* fixes cardinal in reg -> real on x86-64, resolves #8225
|
пре 18 година |
|
Jonas Maebe
|
05a07a7dd2
* better fix for previous revision
|
пре 18 година |
|
peter
|
658c46b903
* remove tdictionary and tindexarray
|
пре 19 година |
|
peter
|
3078a1927f
* remove ttype
|
пре 19 година |
|
Jonas Maebe
|
e2872c9aea
* fixed more missing subsetreg/subsetref support
|
пре 19 година |
