 pierre
|
1cbc526501
Reduce NumIterations for I8086 cpu to avoid timeout
|
9 years ago |
 Jonas Maebe
|
8df1d1f9b8
* fixed compilation of tw15391 with range checking enabled after r34034:
|
9 years ago |
 nickysn
|
ac5658470e
+ use the 16-bit movsx and movzx instructions on 386+ in tcg8086.a_load_ref_reg
|
9 years ago |
|
pierre
|
16477c598d
Fix huge memory model failure by adding nostackframe modifier needed to avoid DS reloading at proc entry
|
9 years ago |
|
pierre
|
ef6a8009cf
Add {$goto on} for FPC to avoid compile time error
|
9 years ago |
|
pierre
|
2ca06538d6
Limit alignment requirement for I8086 CPU to 2 even if pointer is 4 byte long
|
9 years ago |
|
nickysn
|
a5f1ae97d4
+ make use of the 16-bit movsx and movzx instructions in tcg8086.a_load_reg_reg
|
9 years ago |
|
pierre
|
fd4e42abaa
Regenerate after 34044: Use createlst.mak also to generate tbs/tbf/webtbs/webtbf directory lsitings to avoid hitting parameter length limitation
|
9 years ago |
|
pierre
|
7dcd7393e8
Use createlst.mak also to generate tbs/tbf/webtbs/webtbf directory lsitings to avoid hitting parameter length limitation
|
9 years ago |
 svenbarth
|
72be688a8c
A unit's threadvar list needs to be indirectly referenced by the THREADVARLIST as well.
|
9 years ago |
 marco
|
c335a3ea6e
* change TDIR fields to longint. Should have been abandonned long ago.
|
9 years ago |
|
marco
|
5038b944f1
* fix for bug #30331
|
9 years ago |
 michael
|
912c024f01
* Implement pascal class conversion test, from Mohamed Elemam
|
9 years ago |
|
michael
|
ffe626fe6e
* Patch from Graeme Geldenhuys to remove unneeded tools
|
9 years ago |
|
Jonas Maebe
|
95f28e5e0c
* avoid a crash when processing synthetic methods of specialised classes:
|
9 years ago |
|
Jonas Maebe
|
ec5dea1092
* don't perform the "/ const" to "* (1/const)" transformation for comp-typed
|
9 years ago |
|
Jonas Maebe
|
996e325175
* converted range checking for open arrays/array of const from the code
|
9 years ago |
|
Jonas Maebe
|
33b44443b8
+ get_unsigned_inttype(def) to get an unsigned integer type of the same
|
9 years ago |
|
Jonas Maebe
|
8e2607d96d
* data for bitpacked arrays is part of an aggregate
|
9 years ago |
|
Jonas Maebe
|
b5cb7a5d4f
* factored out loading of integers from memory into structs in registers or
|
9 years ago |
|
Jonas Maebe
|
214aea8853
* a_loadfpu_ref_cgpara: use the paralocation instead of the cgpara size
|
9 years ago |
|
Jonas Maebe
|
eeae34ea7d
* simplified handling of loading a reference into a register of a
|
9 years ago |
|
Jonas Maebe
|
f2c1e4234f
* use g_ptrtypecast_ref() in a_load_ref_reg() when we have to typecast
|
9 years ago |
|
Jonas Maebe
|
c421171461
* fixed handling of fpu/mm loads from "named registers" in case they involve
|
9 years ago |
|
Jonas Maebe
|
0ce68d223f
- removed unused tllvmshadowsymtable.recordalignmin field
|
9 years ago |
|
Jonas Maebe
|
542b6b604e
* missing closing comment
|
9 years ago |
|
pierre
|
fb50d0860b
Use BX register instead of AX to reload DS at proc entry if register convention is used
|
9 years ago |
|
pierre
|
8058f6b301
Use SizeInt type for DoMem variable argument, as it might be different from PtrInt, for instance for some msdos memory models
|
9 years ago |
|
pierre
|
d2113f5ead
Use SizeInt type for DoMem variable argument, as it might be different from PtrInt, for instance for some msdos memory models
|
9 years ago |
|
pierre
|
5ff51591cf
* Fix "-CX" use for i8086 with generic functions (test/tgenfuc*.pp tests),
|
9 years ago |
