Daniil Tatianin

Invocation of _GTS was permanently removed from ACPICA here: https://github.com/acpica/acpica/commit/33b1aae263e80c47c687e9f40a5d1a18bfc75155 And it has pretty good reasoning as for why, I think LAI shouldn't try to do it either.

LAI misunderstanding of the Preserve field property(?)

2

As indicated by LAI's own field test here: ``` OperationRegion (IDXF, SystemIO, 0xC00, 0x2) Field (IDXF, ByteAcc, NoLock, Preserve) { IDX, 8, DAT, 8 } IndexField (IDX, DAT, ByteAcc, NoLock,...

Output ``` amldebug: (string "---------- ERROR : SHL0") amldebug: (string "TITLE : Integer arithmetic") amldebug: (string "COLLECTION : functional") amldebug: (string "TEST CASE : arithmetic") amldebug: (string "TEST : SHL0")...

I've already opened a similar issue in ACPICA: https://github.com/acpica/acpica/issues/908 I think LAI is mimicking ACPICA here, but that is not correct behavior. The operand of `IncrementOp` is a `SuperName` therefore...

Steps to reproduce: 1. Compile this code ``` DefinitionBlock ("", "DSDT", 2, "TTEST", "TESTTABL", 0xF0F0F0F0) { Scope(_SB) { Method(_INI, 0, NotSerialized) { Local0 = 123 Debug = Concatenate("Test", RefOf(Local0)) }...

- [ ] ACPI Global Lock - [ ] Customizable _OSI - [ ] An ability to override tables - [ ] PCI API - [ ] Atomicity of initialization...

A list of things that must be done to consider uACPI fully thread-safe. Most of these are guarding against edge cases that basically don't exist in the real world but...