 Jonas Maebe
|
0b246f3dbd
* converted Boolean8 to an internal type, and mapped Boolean to the
|
há 6 anos atrás |
 florian
|
e1546303f8
+ enable use of vfma and friends on arm when doing fastmath optimizations
|
há 9 anos atrás |
|
florian
|
1c067e96bf
* fix VFPv4 support
|
há 9 anos atrás |
 Jeppe Johansen
|
297c17d4f0
Add missing prefix for VCMP for FPv4_S16.
|
há 10 anos atrás |
|
Jonas Maebe
|
0fc1fd6ac1
* replaced current_procinfo.currtrue/falselabel with storing the true/false
|
há 10 anos atrás |
|
Jeppe Johansen
|
9e5979e8be
Implemented UAL syntax support in the ARM assembler reader. Can be toggled with a field for now, but not implemented yet. Still using pre-UAL syntax for now.
|
há 10 anos atrás |
|
florian
|
2fa7171a45
* generate AND for small set comparisons also when only set vars are involved using the cg class, so it works for arm thumb as well
|
há 11 anos atrás |
|
florian
|
97fc823e33
* generate AND for small set comparions using the cg class, so it works for arm thumb as well
|
há 11 anos atrás |
 Károly Balogh
|
09608a1c28
* fix warnings when compiling the compiler with DFA optimizer enabled on ARM
|
há 11 anos atrás |
|
Jonas Maebe
|
e21d31dc99
* fixed compilation with range checking enabled
|
há 11 anos atrás |
 masta
|
96915b3f0c
16bit Thumb is not able to use tst with an immediate value
|
há 11 anos atrás |
|
masta
|
b898b169d4
Fixed 0-cmp optimization in tarmaddnode.second_cmp64bit
|
há 11 anos atrás |
|
masta
|
c88fdb6a4a
Add minor optimization for int64 < 0 on arm
|
há 11 anos atrás |
 sergei
|
819f6de01a
* ARM: fixed floating point comparison flags for VFP hardfloat (part of Mantis #9362).
|
há 11 anos atrás |
|
sergei
|
4c3a4646ec
* Fixed floating-point comparisons for softfloat targets (part of Mantis #9362).
|
há 11 anos atrás |
|
Jonas Maebe
|
4065483a50
* completed thlcgobj.location_force_fpureg(), use it everywhere and removed
|
há 11 anos atrás |
|
florian
|
9e4c130bc4
* do not allocate unnecessarily an extra pair of registers when multiplying int64s on arm
|
há 11 anos atrás |
|
sergei
|
a3efd9e1df
+ Added method taddnode.use_generic_mul64bit, allowing it to inline full 64-bit multiplications, and fixed ARM to comply (it was not checking for possible 32x32 to 64 optimization after detecting a 64-bit operand, so recently added code for 32x32 to 64 bit optimization was inactive).
|
há 11 anos atrás |
|
florian
|
1fc6316cbc
* use generic 32*32 -> 64 bit multiplication on arm thumb
|
há 11 anos atrás |
|
masta
|
bcec29f1b4
Disable generation of UMULL when generating thumb1 code
|
há 11 anos atrás |
|
masta
|
e5b3d89a5d
Add CPUARM_HAS_UMULL flag
|
há 11 anos atrás |
|
masta
|
d32abdf047
Little cleanup in tarmaddnode.second_add64bit
|
há 11 anos atrás |
|
masta
|
cec633e86e
Optimized support for 32x32 => 64bit multiplications on ARM
|
há 11 anos atrás |
|
Károly Balogh
|
25619d4991
removed the const-in-the-left internalerror
|
há 12 anos atrás |
|
florian
|
d4968e054b
+ arm: tsettings.instructionset
|
há 12 anos atrás |
|
Károly Balogh
|
6591a24bd3
removed some code which must never be called, added an internalerror() instead
|
há 12 anos atrás |
|
florian
|
bb73a2891b
* prevent generation of mul/mla statements with illegal registers
|
há 12 anos atrás |
|
Károly Balogh
|
614afca755
ARM: try to inline 64bit multiplication for the most common cases instead of calling FPC_MUL_ helpers (results in 5-10x speedup)
|
há 12 anos atrás |
|
florian
|
3a393f839e
* do not reuse register locations on arm fpa/vfp
|
há 12 anos atrás |
|
Jonas Maebe
|
5051453806
+ support for LOC_(C)MMREGISTER in hlcg
|
há 12 anos atrás |
