 J. Gareth "Curious Kit" Moreton
|
67a1d52806
JccAdd2SetccAdd modified to make use of GetIntRegisterBetween
|
3 år sedan |
|
J. Gareth "Curious Kit" Moreton
|
30166f8eb7
Procedure-saved registers are now recorded for peephole optimizers to use
|
3 år sedan |
 florian
|
dca4bde3a2
* unified internal error
|
3 år sedan |
|
florian
|
c0d75c1c69
* patch Christo Crause: Use LDS for 8 bit references, resolves #38173
|
4 år sedan |
|
florian
|
eb098a3d11
* do not reuse a loaded reference for avrtiny in a_op_const*, resolves #38142
|
4 år sedan |
|
florian
|
637976e83f
* patch by Marģers to unify internal error numbers, resolves #37888
|
4 år sedan |
|
florian
|
d7fe9914a7
+ introduce tcgobj.a_loadfpu_reg_intreg
|
4 år sedan |
 pierre
|
fdb1cd38ab
Disable limitation of handling of negative shift values, can be enalbed again by setting LIMIT_NEG_SHIFTVALUES macro
|
5 år sedan |
|
pierre
|
02fd6f6e54
* tcg.a_load_cgparaloc_ref: Always enable SHR instruction for mips/mipsel CPUs
|
5 år sedan |
 Jonas Maebe
|
e7d1a77f9a
* rename the ARM/AArch64-Darwin targets to ARM/AArch64-iOS
|
5 år sedan |
 nickysn
|
0f6ab0de17
* handle LOC_(C)SUBSETREG/REF in second_NegNot_assign
|
5 år sedan |
|
Jonas Maebe
|
3f6ad30b69
* don't convert the fpu parameters size from tcgsize -> int -> float_tcgsize
|
5 år sedan |
|
Jonas Maebe
|
722ad1ff7b
* support floating point parameters split over multiple locations, including
|
5 år sedan |
|
florian
|
9e0337f248
* do not use an extra register in tcginlinenode.second_IncDec if not needed
|
5 år sedan |
|
nickysn
|
70e90175e5
* fix warnings in cgobj for 8-bit alu cpus
|
5 år sedan |
|
nickysn
|
bf8d560cc6
* treat all Z80 registers as 8-bit
|
5 år sedan |
|
nickysn
|
4b281dd6c9
* changed the ifndef avr to ifdef avr in GetNextReg
|
5 år sedan |
|
nickysn
|
71cadc0a3e
* moved the AVR-specific comment next to the AVR specific code
|
5 år sedan |
|
nickysn
|
54811831b5
- disable the check for R_SUBWHOLE in GetNextReg for Z80
|
5 år sedan |
|
pierre
|
41e554067a
Fix handling of parameters with size below the size of a full register
|
5 år sedan |
|
florian
|
b7c6e01b03
* cleaning up tcgsize: it makes no sense to declare every combination and type
|
5 år sedan |
 svenbarth
|
6f584333e5
* symbols called by g_call might need to be imported from dynamic packages as well
|
5 år sedan |
|
florian
|
d275e7a7e6
* removed accidently committed debug statement
|
5 år sedan |
|
florian
|
c8f746b881
* do not allocate an extra register for some integer operations if not needed
|
5 år sedan |
|
florian
|
b3ed34592f
+ software handling of exceptions on arm
|
6 år sedan |
|
Jonas Maebe
|
281b3ad276
* fix case completeness and unreachable code warnings in compiler that would
|
6 år sedan |
|
Jonas Maebe
|
d99d1f1f30
* let the ARM code generator use the generic tcg.a_load_ref_cgpara() instead
|
6 år sedan |
|
florian
|
9f16c34329
+ initial work for tls-based threadvar support on arm-linux
|
6 år sedan |
 yury
|
2b459b30d2
* Optimized generic implementations of tcg.a_op_const_ref() and tcg.a_op_reg_ref() for PIC and CPUs which need 2 or more instructions to load a symbol's value by reference.
|
6 år sedan |
|
florian
|
3c69f9a066
+ basic infrastructure to generate code for floating point exception
|
7 år sedan |
