diff options
| author | Elliott Hughes <enh@google.com> | 2010-02-24 16:36:18 -0800 |
|---|---|---|
| committer | Elliott Hughes <enh@google.com> | 2010-02-24 16:36:18 -0800 |
| commit | b4c05977c28c38d2f81b48d0cb15559dc3d05564 (patch) | |
| tree | b40cc2cc302a088b931eb97bdc0fd6997d96f41f /vm/compiler/codegen/arm/Assemble.c | |
| parent | 88a0f970e47dc0091d2c9965aa9bd06667e5f4b7 (diff) | |
Optimize more easy multiplications by constants.
Rather than make these changes in the libraries (*10 being a common case),
let's do them once and for all in the JIT.
The 2^n-1 case could be better if we generated RSB instructions, but the
current "fake" RSB is still better than a full multiply.
Thumb doesn't support reg/reg/reg/shift instructions, so we can't optimize
the "population count <= 2" cases (such as *10) there.
Tested on sholes, passion, and passion-running-sapphire (and visually
inspected to check we weren't trying to generate Thumb2 instructions there).
Also tested with the self-verifier.
Diffstat (limited to 'vm/compiler/codegen/arm/Assemble.c')
| -rw-r--r-- | vm/compiler/codegen/arm/Assemble.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/vm/compiler/codegen/arm/Assemble.c b/vm/compiler/codegen/arm/Assemble.c index 09783f232..7f8ad9edf 100644 --- a/vm/compiler/codegen/arm/Assemble.c +++ b/vm/compiler/codegen/arm/Assemble.c @@ -36,7 +36,7 @@ * s2e: src2 end bit position * operands: number of operands (for sanity check purposes) * name: mnemonic name - * fmt: for pretty-prining + * fmt: for pretty-printing */ #define ENCODING_MAP(opcode, skeleton, k0, ds, de, k1, s1s, s1e, k2, s2s, s2e, \ k3, k3s, k3e, flags, name, fmt, size) \ |