| .. |
|
aasmcpu.pas
|
f0e6d2c054
* MIPS: MULT[U] and DIV[U] don't modify first operand.
|
12 years ago |
|
aoptcpu.pas
|
0c8546f94c
* more MIPS code of David Zhang integrated
|
16 years ago |
|
aoptcpub.pas
|
93e0dd9c2f
* Patch from Fuxin Zhang: other mips and mipsel CPUs changes
|
13 years ago |
|
aoptcpud.pas
|
0c8546f94c
* more MIPS code of David Zhang integrated
|
16 years ago |
|
cgcpu.pas
|
fb88cc4257
* TCGMIPS.a_load_reg_reg: reduce code duplication, and don't generate same register move for OS_32->OS_S32 and vice versa. Such moves explode into at least 4 instructions if register needs spilling, after which they are no longer recognized and cannot be removed by reg.allocator. So it's much better not to generate them in first place.
|
12 years ago |
|
cpubase.pas
|
7e0ae2e984
* MIPS: fixed cgsize2subreg to return correct result for float registers.
|
12 years ago |
|
cpuelf.pas
|
cded05ccef
+ MIPS linker: support linking PIC object files with non-PIC ones, by routing absolute calls into PIC code through stubs that load R25.
|
12 years ago |
|
cpugas.pas
|
2868a30cce
+ Added mips32r2 opcodes needed for pic32.
|
12 years ago |
|
cpuinfo.pas
|
fe322f35d5
* MIPS: fixed passing CPU type specified with -CpXXX switch to assembler
|
12 years ago |
|
cpunode.pas
|
a3ef2b42a8
Remove more TABs in sources
|
13 years ago |
|
cpupara.pas
|
9938169d2c
* don't use the paracgsize in get_paraloc_def(), because it generally
|
12 years ago |
|
cpupi.pas
|
86637a9ff9
* MIPS: pi_needs_got is necessary when doing unsigned to float conversions (it uses a global constant) and also if procedure does any calls in PIC mode.
|
12 years ago |
|
cputarg.pas
|
32ffddaad8
+ ELF linker back-ends for ARM and MIPS.
|
12 years ago |
|
hlcgcpu.pas
|
823e3ea398
MIPS: reworked PIC/call code:
|
12 years ago |
|
itcpugas.pas
|
3d2a27c66c
* fix fpu register type
|
13 years ago |
|
mipsreg.dat
|
944d500d55
Change std reg names to allow use with GAS assembler
|
13 years ago |
|
ncpuadd.pas
|
2944fc8839
* MIPS improvements:
|
12 years ago |
|
ncpucall.pas
|
a3ef2b42a8
Remove more TABs in sources
|
13 years ago |
|
ncpucnv.pas
|
86637a9ff9
* MIPS: pi_needs_got is necessary when doing unsigned to float conversions (it uses a global constant) and also if procedure does any calls in PIC mode.
|
12 years ago |
|
ncpuinln.pas
|
f54365db94
* adapted more fpc-mips stuff to trunk
|
16 years ago |
|
ncpuld.pas
|
4b820a1ca5
- Removed tcgloadnode.generate_picvaraccess, it is never used and is not necessary because PIC stuff is handled at lower levels.
|
12 years ago |
|
ncpumat.pas
|
1f8bd4a2d1
+ MIPS: initial target-specific unary minus node.
|
12 years ago |
|
ncpuset.pas
|
b023627f6a
* converted tcgcasenode.pass_generate_code() to hlcgobj
|
14 years ago |
|
opcode.inc
|
2868a30cce
+ Added mips32r2 opcodes needed for pic32.
|
12 years ago |
|
racpugas.pas
|
2868a30cce
+ Added mips32r2 opcodes needed for pic32.
|
12 years ago |
|
rgcpu.pas
|
1421998752
* MIPS register spilling: relocations are not needed to load constant values, and offset limit is 32767, not 4095.
|
12 years ago |
|
rmipscon.inc
|
de4a96f96d
* fixes several register allocation related mips issues
|
13 years ago |
|
rmipsdwf.inc
|
f58fcdf401
+ basic mips stuff
|
20 years ago |
|
rmipsgas.inc
|
ae37b9f5b9
* fix floating point registers gas name
|
13 years ago |
|
rmipsgri.inc
|
ae37b9f5b9
* fix floating point registers gas name
|
13 years ago |
|
rmipsgss.inc
|
f58fcdf401
+ basic mips stuff
|
20 years ago |
|
rmipsnor.inc
|
f58fcdf401
+ basic mips stuff
|
20 years ago |
|
rmipsnum.inc
|
de4a96f96d
* fixes several register allocation related mips issues
|
13 years ago |
|
rmipsrni.inc
|
f58fcdf401
+ basic mips stuff
|
20 years ago |
|
rmipssri.inc
|
944d500d55
Change std reg names to allow use with GAS assembler
|
13 years ago |
|
rmipssta.inc
|
f58fcdf401
+ basic mips stuff
|
20 years ago |
|
rmipsstd.inc
|
944d500d55
Change std reg names to allow use with GAS assembler
|
13 years ago |
|
rmipssup.inc
|
de4a96f96d
* fixes several register allocation related mips issues
|
13 years ago |
|
strinst.inc
|
2868a30cce
+ Added mips32r2 opcodes needed for pic32.
|
12 years ago |
sergei
fb88cc4257
* TCGMIPS.a_load_reg_reg: reduce code duplication, and don't generate same register move for OS_32->OS_S32 and vice versa. Such moves explode into at least 4 instructions if register needs spilling, after which they are no longer recognized and cannot be removed by reg.allocator. So it's much better not to generate them in first place.