 sergei
|
ae627a4ba8
* tcgx86.a_op_const_reg_reg: optimize trivial cases before considering more complex ones.
|
11 rokov pred |
 nickysn
|
daf71e6d88
* fixed multiplication by 64-bit constants on x86_64, mantis #26230
|
11 rokov pred |
|
nickysn
|
6e7dd647fd
* Tcgx86.g_concatcopy: optimization for the case where source.segment is
|
11 rokov pred |
|
nickysn
|
1eb24f9d98
+ remove the segment prefix in make_simple_ref on i8086, in case it is the same
|
11 rokov pred |
|
sergei
|
07e90aaa24
+ Implemented IEEE 754-compliant checking for unordered results of floating-point compares on x86 targets. Mantis #9362.
|
11 rokov pred |
 florian
|
18233f2c0c
* fix compilation of i8086 compiler
|
11 rokov pred |
|
florian
|
d88d644925
+ support for FMA intrinsic: if there is no hardware support, the compiler throws an error.
|
11 rokov pred |
|
nickysn
|
afcb8c3f10
* initialize DS from CS in the interrupt procedure entry code in tiny model,
|
11 rokov pred |
|
nickysn
|
7a873f495d
* use the stack segment for SP-relative addresses in Tcgx86.g_concatcopy in
|
11 rokov pred |
|
nickysn
|
ee20ecfcf2
* tcgx86.make_simple_ref: in i8086 memory models with DS<>SS, add an SS: segment
|
11 rokov pred |
|
nickysn
|
b36a98ce51
+ Tcgx86.g_concatcopy: added support for memory models with DS<>SS
|
11 rokov pred |
|
nickysn
|
e9e4cc8bcb
* Tcgx86.g_concatcopy: use boolean variables saved_ds and saved_es to keep track
|
11 rokov pred |
|
nickysn
|
096bba90da
* fixed Tcgx86.g_concatcopy not using the correct push instruction size when
|
11 rokov pred |
 pierre
|
044891fd94
Avoid range check error in a_op_const_reg_reg
|
11 rokov pred |
|
nickysn
|
229ebf33a8
+ support getting the address of far references on i8086 in
|
11 rokov pred |
|
florian
|
37b34edbc6
* do not reuse registers when calculating addresses in pic mode
|
11 rokov pred |
|
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 rokov pred |
|
sergei
|
0d3f36eebf
- Remove references to global variable 'cg' from methods of tcg and some of its descendants.
|
11 rokov pred |
|
florian
|
28a8dc42c6
* fix tcgx86.a_op_reg_reg on x86-64 and make use of it
|
11 rokov pred |
|
florian
|
7d7bf1d877
+ make use of SHLX/SHRX on CPUs supporting BMI2
|
11 rokov pred |
|
florian
|
48ae2d215a
+ concatcopy variants using sse and avx, only activated if optimization for size is done because at least on an i7-4770 it has shown no benefit
|
11 rokov pred |
|
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 rokov pred |
|
sergei
|
6b1f021fcf
- Removed remaining optimizations that duplicate functionality of tcg.optimize_op_const.
|
11 rokov pred |
|
nickysn
|
1f7b14f38e
* fixed the conversion of powerof2 mul/imuls to lea in tcgx86.a_op_const_reg_reg
|
11 rokov pred |
|
nickysn
|
cc1faad3d2
+ make use of the 3-operand form of IMUL on i386 and x86_64
|
11 rokov pred |
|
nickysn
|
5d75bf4f92
* always use IMUL instead of MUL on i386, when doing a 32->32 multiplication
|
11 rokov pred |
|
nickysn
|
820b667e1b
- rm the OP_IMUL to SHL optimization from tcgx86.a_op_const_reg and tcgx86.a_op_const_ref, because it's already done in optimize_op_const, which is called in the beginning of both methods.
|
11 rokov pred |
|
florian
|
e210d5f30e
+ cpu_capabilites for x86_64 and i386
|
11 rokov pred |
|
sergei
|
f2096de53a
* tcg386.g_proc_exit: instead of recalculating stack size, use current_procinfo.final_localsize which was calculated in g_proc_entry.
|
11 rokov pred |
|
sergei
|
179586f589
+ SEH support for Win32. Enable by cycling with OPT=-dTEST_WIN32_SEH.
|
11 rokov pred |
