# New Ticket Created by Jason Gloudon # Please include the string: [perl #21508]
# in the subject line of all future correspondence about this issue. # <URL: http://rt.perl.org/rt2/Ticket/Display.html?id=21508 >
Thanks, applied.
Some of the new ops are commented out. They really need a constant pool to
implement them correctly. It would be nice if the constant pool was more
generic,
I don't know too much about this topic, but many processors seem to be lacking instructions with immediate constants > $SIZE, where $SIZE seems to be irregular on arm and 32 bit on 64 bit processors.
So the question is, why we don't put integer constants into the const_table too, if we know that this PBC will be run on this architecture with said limitation.
So the portable PBC still would have integer constants inlined, they can be loaded with some penalty and more instructions. And a PBC file compiled for a specific architecture could have constants > $SIZE in the const_table, coded as index with some bias.
To have no penalty on non JITed opcodes, the assembler (i.e. imcc) has to look at op_jit, if the instruction is JITed or not and emit either variant, that is best.
leo
