 florian
|
e46b426ca1
+ tx86addnode.second_opvector makes use of AVX instructions
|
4 years ago |
|
florian
|
637976e83f
* patch by Marģers to unify internal error numbers, resolves #37888
|
4 years ago |
|
florian
|
043c3df292
* properly handle boolean64 xor boolean64 on i386
|
5 years ago |
|
florian
|
b2ed667a86
* swap fpu stack correctly when moving data into xmm registers
|
5 years ago |
|
florian
|
5f67d830e7
* better rules if a 3 op mul operation is tried
|
5 years ago |
|
florian
|
b4de0b2e9c
* insert fwait after fcompp, resolves #37136
|
5 years ago |
|
florian
|
97efb03ae6
* call x86 addordinal for non short boolean evaluated boolean nodes
|
5 years ago |
|
florian
|
fca4d3315b
* let the operation do the type conversion for x87 floating point operations
|
5 years ago |
|
florian
|
3c16324f80
* x86: avoid to put variables into registers if they are handled by the x87-FPU
|
5 years ago |
|
florian
|
91f3f022fe
* MM registers which might not be changed can be still used in 3 op mathematical operations as source
|
5 years ago |
|
florian
|
5f3050961f
* load left properly into a register in tx86addnode.second_addordinal, resolves #36287
|
5 years ago |
|
florian
|
c8f746b881
* do not allocate an extra register for some integer operations if not needed
|
5 years ago |
 Jonas Maebe
|
ce598c15ec
* factored out the conditions under which add nodes need to perform
|
6 years ago |
|
Jonas Maebe
|
281b3ad276
* fix case completeness and unreachable code warnings in compiler that would
|
6 years ago |
 pierre
|
9c90f593ab
Add global range check disable for i8086 cgcpu and x86 nx86add units
|
6 years ago |
|
Jonas Maebe
|
4686f61002
* keep track of the temp position separately from the offset in references,
|
7 years ago |
|
florian
|
088af633ba
Fix casing, bug ID #38660
|
7 years ago |
|
florian
|
9b18e39c81
* enable Lea2AddBase and Lea2AddIndex in TX86AsmOptimizer.PostPeepholeOptLea as we have flag tracking now
|
7 years ago |
|
florian
|
d86ffb9bfb
+ tcg.a_op_loc_reg
|
7 years ago |
 nickysn
|
efc5e339d0
* use an enum instead of integer constants to represent inline numbers
|
8 years ago |
|
florian
|
b1dff29cbf
* removed unused units
|
8 years ago |
|
Jonas Maebe
|
a25ebbba3e
+ added volatility information to all memory references
|
8 years ago |
|
Jonas Maebe
|
da696057ab
* converted register_maybe_adjust_setbase() to the high level code generator
|
9 years ago |
|
nickysn
|
457107d165
* check UseIncDec before generating an INC instruction in tx86addnode.emit_generic_code
|
9 years ago |
 yury
|
e04a3b4ac6
* Removed unused vars for x86 compiler.
|
10 years ago |
|
florian
|
c9b1df2c91
* the test $-1,%reg optimization shall be done not only if the cpu type is i386, but for the i386 compiler
|
10 years ago |
 Károly Balogh
|
249a60b28b
x86: fix a variable op not initialized warning. This hopefully fixes our x86 testsuite run.
|
11 years ago |
|
florian
|
499dd078e3
+ automatically insert fma inlines into floating point code if possible and fastmath is activated
|
11 years ago |
 sergei
|
2ee0c8de45
* i386: For integer comparisons with zero, emit "test $-1,%reg" instead of "test %reg,%reg". It is more spilling-friendly, because it transforms into "test $-1,spilltemp" and does not require a register.
|
11 years ago |
|
sergei
|
07e90aaa24
+ Implemented IEEE 754-compliant checking for unordered results of floating-point compares on x86 targets. Mantis #9362.
|
11 years ago |
