To avoid GCC warnings when building with `-Os`:
warning: inlining failed in call to 'table_walk': call is
unlikely and code size would grow [-Winline]
I don't know if this causes a performance regression when optimizing for
speed instead of size, but perhaps there's a different way we can help
the compiler make this decision in such cases.
Where we'd end up losing the upper bits.
GCC catches this and emits a warning such as:
warning: result of '7 << 31' requires 35 bits to represent, but
'long int' only has 32 bits [-Wshift-overflow=]
Where size_t is only 32 bits wide, and we end up losing the upper bits.
GCC catches this and emits a warning such as:
warning: left shift count >= width of type [-Wshift-count-overflow]
The AMD64 instructions VPERMIL2PS and VPERMIL2PD (currently not
supported) encode a fifth immediate operand in the lower bits of the
re-purposed immediate. Expose this value in any case so that no
information gets lost during decoding.
This changes the instruction description format:
- Use Intel/AMD terminology for describing operands (where applicable)
- Group instructions by ISA extension
- Indicate read/written status flags
As the formatter no longer demands a null-terminated string, mnemonics
can arbitarily overlap and therefore save space.
This is the shortest superstring problem, which is NP-hard. This is
currently approximated with a greedy heuristic.
These instructions have plenty of corner cases and some instructions
have a different usage of the memory operand. Given that MPX is already
deprecated by Intel, it seems that the better option is to decode these
(rarely occuring) instructions as NOPs.
The new trie implementation is more flexible and allows omitting
prefixes even with a ModRM specifier in the opcode. Use this flexibility
to simplify instruction descriptions.
