- made OP_NEW a builtin function instead of an opcode.
The code was present 3 times due to the JIt, and this is not something that benefits from being a real opcode, even in the interpreted case.
This commit is contained in:
parent
cc73449d29
commit
4126f8ce72
6 changed files with 64 additions and 113 deletions
|
|
@ -108,8 +108,6 @@ xx(SCOPE, scope, RPI8, NOP, 0, 0) // Scope check at runtime.
|
|||
xx(RESULT, result, __BCP, NOP, 0, 0) // Result should go in register encoded in BC (in caller, after CALL)
|
||||
xx(RET, ret, I8BCP, NOP, 0, 0) // Copy value from register encoded in BC to return value A, possibly returning
|
||||
xx(RETI, reti, I8I16, NOP, 0, 0) // Copy immediate from BC to return value A, possibly returning
|
||||
xx(NEW, new, RPRPI8, NOP, 0, 0)
|
||||
xx(NEW_K, new, RPKP, NOP, 0, 0)
|
||||
//xx(TRY, try, I24, NOP, 0, 0) // When an exception is thrown, start searching for a handler at pc + ABC
|
||||
//xx(UNTRY, untry, I8, NOP, 0, 0) // Pop A entries off the exception stack
|
||||
xx(THROW, throw, THROW, NOP, 0, 0) // A == 0: Throw exception object pB
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue