Add near calls (32-bit displacement) to Code objects on X64 platform.

Add near calls (32-bit displacement) to Code objects on X64 platform.

Committed: http://code.google.com/p/v8/source/detail?r=3021

[William Hesse, 2009-09-23 12:39:27]

Please review.  Erik just needs to review (de)serialization.

[Erik Corry, 2009-09-24 08:36:42]

That part of it looks good to me.

http://codereview.chromium.org/200095/diff/10001/10012
File src/serialize.cc (right):

http://codereview.chromium.org/200095/diff/10001/10012#newcode940
Line 940: Code* target =
Code::GetCodeFromTargetAddress(rinfo->target_address());
This is ugly, but lets just put it in anyway because I plan to remove all this
code real soon now.

http://codereview.chromium.org/200095/diff/10001/10012#newcode943
Line 943: intptr_t int_target = reinterpret_cast<intptr_t>(encoded_target);
Please assert this fits in low 32 bits.

http://codereview.chromium.org/200095/diff/10001/10012#newcode944
Line 944: int_target = (int_target & 0xFFFFFFFFU) |
(Memory::uint64_at(rinfo->pc() + 4) << 32);
This won't lint!  And looks like it doesn't work in 32 bit.

[Lasse Reichstein, 2009-09-24 10:37:20]

LGTM if it works (I don't have complete understanding of how the relocation
works).

http://codereview.chromium.org/200095/diff/10001/10010
File src/arm/assembler-arm-inl.h (right):

http://codereview.chromium.org/200095/diff/10001/10010#newcode84
Line 84: return Memory::Object_at(Assembler::target_address_address_at(pc_));
Can't we find a better name than "address_address"?

http://codereview.chromium.org/200095/diff/10001/10012
File src/serialize.cc (right):

http://codereview.chromium.org/200095/diff/10001/10012#newcode943
Line 943: intptr_t int_target = reinterpret_cast<intptr_t>(encoded_target);
int_target is a bad name for something that isn't an int.
How about "scalar_target", or "target_as_scalar", or just "intptr_target".

http://codereview.chromium.org/200095/diff/10001/10012#newcode944
Line 944: int_target = (int_target & 0xFFFFFFFFU) |
(Memory::uint64_at(rinfo->pc() + 4) << 32);
This needs a comment!
I.e., I have no idea what is going on.

Can't you get rid of the uint64_at thing in 32-bit mode?

http://codereview.chromium.org/200095/diff/10001/10012#newcode1670
Line 1670: encoded = reinterpret_cast<Address>(0xFFFFFFFFU &
reinterpret_cast<intptr_t>(encoded));
The content of "encoded" isn't really an Address (it's not a pointer to a byte
in memory if it's always at most 32 bit). Maybe it should have a different type.

http://codereview.chromium.org/200095/diff/10001/10008
File src/x64/assembler-x64-inl.h (right):

http://codereview.chromium.org/200095/diff/10001/10008#newcode284
Line 284: pc_ + Assembler::kRealPatchReturnSequenceAddressOffset);
"Real" is not a good prefix for a variable. It doesn't explain how
kPatchReturnSequenceAddressOffset isn't real.

http://codereview.chromium.org/200095/diff/10001/10006
File src/x64/assembler-x64.cc (right):

http://codereview.chromium.org/200095/diff/10001/10006#newcode267
Line 267: Assembler::Assembler(void* buffer, int buffer_size):
Move ":" to next line or initializer to this line.

http://codereview.chromium.org/200095/diff/10001/10007
File src/x64/assembler-x64.h (right):

http://codereview.chromium.org/200095/diff/10001/10007#newcode940
Line 940: void jmp(Handle<Code> target, RelocInfo::Mode rmode);
On a side note, I would like to move all Handle operations to the
Macro-assembler, so that the assembler doesn't need to know about the domain
model. I.e., all handle and smi operations should be in the macro assembler, and
reduce to assembler operations on simple values.