 Jonas Maebe
|
3f736f6114
* handle the loading of VMT entries at the node level, so it's done in a
|
10 lat temu |
 florian
|
d598351664
* call also optimize_op_const in the generic a_op_const_reg_reg
|
10 lat temu |
|
Jonas Maebe
|
61e4a1b811
+ added tasmlist parameter to getintparaloc() (needed for llvm)
|
10 lat temu |
|
Jonas Maebe
|
bd203a5b57
* synchronised with trunk till r30240
|
10 lat temu |
|
Jonas Maebe
|
622852b8c9
* check that a_load_cgparaloc_anyreg() is not used to try to move an fpu
|
10 lat temu |
|
Jonas Maebe
|
6d02aedf70
* support multiple register paralocs in tcgobj.a_loadfpu_ref_cgpara()
|
10 lat temu |
|
Jonas Maebe
|
67b8aceaee
* synchronized with privatetrunk till r30095
|
10 lat temu |
|
Jonas Maebe
|
2ab7f5c35d
* moved x86-specific requirements from the generic bsr/bsf code to the
|
10 lat temu |
|
Jonas Maebe
|
504a9d1594
* fixed register sizes in a_load_ref_reg_unaligned()
|
10 lat temu |
|
Jonas Maebe
|
e6511bcb33
* support multiple paralocs in a_load_reg_cgpara() by taking a round trip
|
10 lat temu |
 sergei
|
8f4430e795
* tcg.translate_register: check that reg.allocator is not nil before calling its method (in case of "assembler nostackframe" function returning an x87 float compiler tries to translate NR_ST which is used for function result).
|
10 lat temu |
|
Jonas Maebe
|
b745dcc64c
* moved g_external_wrapper() to the hlcg, and also g_intf_wrapper() because
|
11 lat temu |
|
Jonas Maebe
|
a637fbe596
* moved all g_exception_*() methods to hlcgobj and cleaned them up (no more
|
11 lat temu |
|
Jonas Maebe
|
5c75b6dd6b
* synchronised with trunk up till r28402
|
11 lat temu |
|
Jonas Maebe
|
831cc96f48
* give an internalerror when calling cg.makeregsize() for a high level
|
11 lat temu |
|
sergei
|
e4fea2ebc8
* Dummy implementations of a_bit_scan_reg_reg and g_stackpointer_alloc in tcg, removes the need to override these methods in every descendant code generator solely to avoid "constructing a class with abstract method" warning.
|
11 lat temu |
|
Jonas Maebe
|
7949bebb8d
* synchronised with r28168 of trunk
|
11 lat temu |
|
sergei
|
cd27d64cd5
+ Support (as target-independent as possible) optimization of division by constants:
|
11 lat temu |
|
Jonas Maebe
|
bacd303208
* synchronized with trunk up to r27758
|
11 lat temu |
 nickysn
|
5dcbaa4b2d
- rm the cpu16bitalu and cpu64bitalu ifdefs from tcg.a_load_ref_cgpara as well
|
11 lat temu |
|
Jonas Maebe
|
dab8754bb6
* moved joinreg64 from cg64f32 to cgobj (joinreg128 is also declared
|
11 lat temu |
|
Jonas Maebe
|
e9268a0a14
* synchronised with trunk up till r26975
|
11 lat temu |
|
sergei
|
9c1f917e3a
* a_call_ref functionality cannot be implemented efficiently at code generator level, because references need specific preparations at earlier points. Moved this support to tcgcallnode and its x86 descendants, and got rid of all ifdef's around.
|
11 lat temu |
|
sergei
|
0d3f36eebf
- Remove references to global variable 'cg' from methods of tcg and some of its descendants.
|
11 lat temu |
|
nickysn
|
be85998313
* fixes in optimize_op_const:
|
11 lat temu |
|
nickysn
|
85dd9e5789
+ added a size parameter to optimize_op_const and do a sign extension of the 'a' parameter up from the specified size, so that things like (i and $ffffffff) get optimized away the same way as (i and -1)
|
11 lat temu |
|
sergei
|
5cd0684d45
* tcg.a_load_ref_reg_unaligned: if loading signed 16-bit value, treat upper half as signed 8-bit to ensure that result is properly extended. This routine is used only by SPARC target (other CPUs use specific implementations), and it generates pretty inefficient code, therefore it's probably worth to drop it altogether and implement SPARC-specific version. Mantis #25440.
|
11 lat temu |
|
Jonas Maebe
|
1df3039424
+ LLVM temp allocator based on new R_TEMPREGISTER register class. For every
|
11 lat temu |
|
nickysn
|
bf07fc077b
* specify AT_DATA in all references to the tobjectdef.vmt_mangledname symbol.
|
11 lat temu |
 svenbarth
|
c48d572996
Implement support for saving and restoring address registers.
|
11 lat temu |
