X64: Implement InstanceofStub. Prohibit zero-size code objects.

X64: Implement InstanceofStub.  Prohibit zero-size code objects.

Committed: http://code.google.com/p/v8/source/detail?r=2330

[Lasse Reichstein, 2009-07-01 10:56:15]

LGTM

http://codereview.chromium.org/151142/diff/1/5
File src/x64/codegen-x64.cc (right):

http://codereview.chromium.org/151142/diff/1/5#newcode5765
Line 5765: __ movq(rax, Operand(rsp, 2 * kPointerSize));  // 2 ~ return address,
function
Is there no constant that can be used here?
The comment is not immediately understandable. Perhaps add a section at the top:
/* 
 * Implements "value instanceof function" operator.
 * Expected input state:
 * rsp[0] : return address
 * rsp[1] : function pointer
 * rsp[2] : value
 */

http://codereview.chromium.org/151142/diff/1/5#newcode5795
Line 5795: __ Cmp(rcx, Factory::null_value());
Could you store Factory::null_value() in kScratchRegister outside the loop?
Should save an instruction.

http://codereview.chromium.org/151142/diff/1/5#newcode5802
Line 5802: __ Set(rax, 0);
use __ xor_(rax, rax) instead? It affects some flags, which is why it's not done
automatically by Set, but that doesn't matter here.

http://codereview.chromium.org/151142/diff/1/2
File src/x64/macro-assembler-x64.cc (right):

http://codereview.chromium.org/151142/diff/1/2#newcode416
Line 416: testb(kScratchRegister, Immediate(1 <<
Map::kHasNonInstancePrototype));
Why not use testb(Operand,Immediate).
Should be: F6 /0 ib   TEST r/m8, imm8

http://codereview.chromium.org/151142/diff/1/2#newcode421
Line 421: FieldOperand(function, JSFunction::kPrototypeOrInitialMapOffset));
Indentation.

http://codereview.chromium.org/151142/diff/1/4
File src/x64/macro-assembler-x64.h (right):

http://codereview.chromium.org/151142/diff/1/4#newcode179
Line 179: // They may be the same register.
And map may be kScratchRegister, since it isn't used by the macro otherwise.

http://codereview.chromium.org/151142/diff/1/4#newcode388
Line 388: reinterpret_cast<byte*>(FUNCTION_ADDR(LogGeneratedCodeCoverage)); \
Could you align the backslashes?

[William Hesse, 2009-07-01 12:14:35]

All comments adressed.