 sergei
|
847adff94d
* Load double-precision locations from integer registers directly into cgpara if sizes match and destination has a single location.
|
11 rokov pred |
 nickysn
|
3e308e0ad7
* 16/8-bit alu fix in thlcg2ll.location_force_reg for the 32->64 conversion of LOC_CONSTANT sources
|
11 rokov pred |
 florian
|
6f14dbdebe
* avoid internal error if an empty record is assigned to a function result, resolves #25895
|
11 rokov pred |
 Jonas Maebe
|
4065483a50
* completed thlcgobj.location_force_fpureg(), use it everywhere and removed
|
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 |
|
nickysn
|
f69a76b5b1
* generate shorter code on 8-bit and 16-bit CPUs in thlcg2ll.location_force_reg,
|
11 rokov pred |
|
nickysn
|
9ec84d1259
* generate shorter code on 8 and 16-bit CPUs in thlcg2ll.location_force_reg,
|
11 rokov pred |
|
nickysn
|
4e795eca99
* fixes for 16 and 8-bit ALUs in thlcg2ll.location_force_reg for the cases when
|
11 rokov pred |
|
Jonas Maebe
|
1b6e4e572f
- removed duplicate thlcg2ll a_label method (it's declared the same in
|
12 rokov pred |
|
Jonas Maebe
|
7d9f98e516
+ def_cgmmsize() that returns a vector tcgsize for arraydefs (as generated
|
12 rokov pred |
|
Jonas Maebe
|
978b5530ec
* renamed getmmcgsize to getintmmcgsize + comment
|
12 rokov pred |
|
Jonas Maebe
|
2dd75e707e
* renamed thlcgobj.tcgsize2orddef to defutil.cgsize_orddef
|
12 rokov pred |
|
Jonas Maebe
|
3456f20546
* changed checks for support of opmm* calls in hlcg2ll from mms_movescalar
|
12 rokov pred |
|
Jonas Maebe
|
1551c773e0
- removed no longer necessary thlcg2ll.initialize_regvars override and
|
12 rokov pred |
|
Jonas Maebe
|
5051453806
+ support for LOC_(C)MMREGISTER in hlcg
|
12 rokov pred |
|
nickysn
|
cbf3b14e20
* yet another 8/16-bit ALU fix in thlcg2ll.location_force_reg
|
12 rokov pred |
|
nickysn
|
a03f0d84ea
high level code generator: const types changed from aint to tcgint to avoid problems on < 32-bit cpus; ppcjvm should also be updated, but isn't yet
|
12 rokov pred |
|
nickysn
|
27e4048ed6
16-bit and 8-bit alu fix in thlcg2ll.location_force_reg
|
12 rokov pred |
|
florian
|
d93cee995b
* fix register method pointer for ppc64 and x86_64-linux
|
13 rokov pred |
|
florian
|
7361e19799
+ support for handling OS_128/OS_S128 on 64 Bit CPUs as far as needed for method pointers in registers
|
13 rokov pred |
|
florian
|
d6b548ce13
* don't force parameters into memory when loading them if both have the size OS_64/OS_S64
|
13 rokov pred |
|
florian
|
d8161c185c
+ track usage of flags by using a new register RS_/NR_DEFAULTFLAGS
|
13 rokov pred |
|
Jonas Maebe
|
7717e43929
* moved setting the call result to a separate method, so it can still be
|
13 rokov pred |
|
Jonas Maebe
|
be8f8fec76
* converted tcg.g_releasevaluepara_openarray() to thlcg
|
13 rokov pred |
|
Jonas Maebe
|
9197ce100e
* converted tcg.g_copyvaluepara_openarray() to thlcg
|
13 rokov pred |
|
Jonas Maebe
|
1955255dda
* let thlcg.a_call_name() return the tcgpara representing the function
|
13 rokov pred |
|
Jonas Maebe
|
aba6923187
* moved g_test_self() from cgobj to hlcgobj
|
13 rokov pred |
|
Jonas Maebe
|
7c21cba1e2
* migrated g_copyshortstring, g_incrrefcount and g_array_rtti_helper to thlcg
|
13 rokov pred |
|
Jonas Maebe
|
7a0ae38700
+ also specify the parameter def when allocating a parameter via
|
13 rokov pred |
|
Jonas Maebe
|
325c60866f
* use tcgpara.def instead of extra size parameters
|
13 rokov pred |
