 J. Gareth "Curious Kit" Moreton
|
6f24c8b4ef
* x86: Code generation fixes where FLAGS
|
3 lat temu |
 florian
|
637976e83f
* patch by Marģers to unify internal error numbers, resolves #37888
|
4 lat temu |
 yury
|
e5a923eb7a
* 8086: Proper implementation of: If a far procvar is called, it must be in a memory location. This fixes a hack to emulate CALL reg1:reg2.
|
5 lat temu |
|
yury
|
04baeb2bcf
* 8086: Reverted r46557. It must be implemented differently.
|
5 lat temu |
|
yury
|
6a3971c9b6
* 8086: If a far procvar is called, it must be in a memory location. This fixes a hack to emulate CALL reg1:reg2.
|
5 lat temu |
|
yury
|
97eb32e6fd
* 8086: Also fixed loading a register to a 32-bit ref in case the source register is used as the ref base or index.
|
5 lat temu |
|
yury
|
4873ec73a7
* 8086: Fixed loading a 32-bit ref to a register in case the target register is used as the ref base or index.
|
5 lat temu |
|
yury
|
0000682043
* i8086: When a ref points to the 32-bit data, preload the ref base to a temp register to reduce spilling.
|
5 lat temu |
|
yury
|
e70c898503
* Removed the i8086 specific hack. Since r46199 the register allocator is able to handle the problem code.
|
5 lat temu |
 nickysn
|
92043c6e20
* call inherited in tcg64f8086.a_op64_reg_ref for OP_NEG and OP_NOT, because it is handled differently after r45302
|
5 lat temu |
|
nickysn
|
4f4d8191f0
+ generate more optimal code for OP_NOT and OP_NEG in tcg64f8086.a_op64_ref
|
5 lat temu |
|
nickysn
|
52572e10e9
* i8086 compilation fixed after r45302
|
5 lat temu |
|
nickysn
|
0f6ab0de17
* handle LOC_(C)SUBSETREG/REF in second_NegNot_assign
|
5 lat temu |
|
florian
|
11319353fd
* loading a got depends only on the fact if a got is needed and not if pic shall be generated
|
5 lat temu |
 Jonas Maebe
|
281b3ad276
* fix case completeness and unreachable code warnings in compiler that would
|
6 lat temu |
 pierre
|
9c90f593ab
Add global range check disable for i8086 cgcpu and x86 nx86add units
|
6 lat temu |
|
florian
|
d4c65cdac4
* better register de-allocation after CWD/CWB
|
7 lat temu |
|
Jonas Maebe
|
d69ad8fa41
* removed temppos field again from parameter locations: they're not allocated
|
7 lat temu |
|
Jonas Maebe
|
4686f61002
* keep track of the temp position separately from the offset in references,
|
7 lat temu |
|
nickysn
|
bd3d35f2da
+ add support for passing 32-bit values in a pair of registers in
|
7 lat temu |
|
nickysn
|
4fc2fa7899
* alloc/dealloc NR_DEFAULTFLAGS in the a_cmp_* methods in the i8086 code
|
7 lat temu |
|
nickysn
|
db09759763
* also integrated the getnextreg() implementation for 8-bit and 16-bit alus from
|
8 lat temu |
|
nickysn
|
cf28b202eb
* integrated the getintregister() implementation for 8-bit and 16-bit alus from
|
8 lat temu |
|
nickysn
|
ddba821561
* GetNextReg(), used by 16-bit and 8-bit code generators (i8086 and avr) moved
|
8 lat temu |
|
nickysn
|
4e4e5d6d07
+ allocate and free the flags register (when necessary), when generating code
|
8 lat temu |
|
nickysn
|
b8c4dd9e18
+ implemented 64-bit OP_SHR,OP_SHL and OP_SAR in a_op64_reg_reg for i8086 and
|
8 lat temu |
|
nickysn
|
a82c89d894
+ implemented OP_SHR,OP_SHL and OP_SAR in a_op64_const_reg for i8086. The shlshr
|
8 lat temu |
|
nickysn
|
65977f9f27
+ implemented OP_ROL and OP_ROR in tcg8086.a_op_reg_reg and .a_op_reg_ref
|
8 lat temu |
|
nickysn
|
8c200fcfba
+ implemented OP_SHR,OP_SHL,OP_SAR,OP_ROL and OP_ROR in tcg8086.a_op_const_ref
|
8 lat temu |
|
nickysn
|
50b1c9c088
+ implemented 32-bit OP_SHR,OP_SHL and OP_SAR in tcg8086.a_op_reg_ref
|
8 lat temu |
