Implement code generation for conditional expressions and regexp...

Implement code generation for conditional expressions and regexp
literals.  Regexp literals will not really work before we can
bootstrap, but the right code is being called with the right
arguments.

Finish implementation of object literals.  We can now call through the
StoreIC_Miss stub into the runtime system to do stores.

Restore comment that I accidentally removed in my last commit.

Committed: http://code.google.com/p/v8/source/detail?r=2244

[William Hesse, 2009-06-22 19:14:23]

LGTM.

http://codereview.chromium.org/140068/diff/1007/14
File src/x64/ic-x64.cc (right):

http://codereview.chromium.org/140068/diff/1007/14#newcode188
Line 188: //  -- rsp[4] : receiver
rsp[8]?

http://codereview.chromium.org/140068/diff/1007/15
File src/x64/register-allocator-x64.cc (right):

http://codereview.chromium.org/140068/diff/1007/15#newcode50
Line 50: RelocInfo::NONE);
Could you say
CodeGeneratorScope::Current()->masm()->Move(fresh.reg(), handle()), or is there
no way to access the masm?

http://codereview.chromium.org/140068/diff/1007/13
File src/x64/virtual-frame-x64.cc (right):

http://codereview.chromium.org/140068/diff/1007/13#newcode862
Line 862: // expects name in ecx, value in eax, and receiver on the stack.  It
rcx, value in rax,

http://codereview.chromium.org/140068/diff/1007/13#newcode874
Line 874: // Register eax is free for value, which frees ecx for name.
Register rax ... rcx

http://codereview.chromium.org/140068/diff/1007/13#newcode879
Line 879: // Register ecx is free for name, which guarantees eax is free for
rcx ... rax

[Mads Ager (chromium), 2009-06-22 19:25:00]

http://codereview.chromium.org/140068/diff/1007/14
File src/x64/ic-x64.cc (right):

http://codereview.chromium.org/140068/diff/1007/14#newcode188
Line 188: //  -- rsp[4] : receiver
On 2009/06/22 19:14:23, William Hesse wrote:
> rsp[8]?

Done.

http://codereview.chromium.org/140068/diff/1007/15
File src/x64/register-allocator-x64.cc (right):

http://codereview.chromium.org/140068/diff/1007/15#newcode50
Line 50: RelocInfo::NONE);
On 2009/06/22 19:14:23, William Hesse wrote:
> Could you say
> CodeGeneratorScope::Current()->masm()->Move(fresh.reg(), handle()), or is
there
> no way to access the masm?
> 
> 

Done.  And also done for the same piece of code below.

http://codereview.chromium.org/140068/diff/1007/13
File src/x64/virtual-frame-x64.cc (right):

http://codereview.chromium.org/140068/diff/1007/13#newcode862
Line 862: // expects name in ecx, value in eax, and receiver on the stack.  It
On 2009/06/22 19:14:23, William Hesse wrote:
> rcx, value in rax, 

Done.

http://codereview.chromium.org/140068/diff/1007/13#newcode874
Line 874: // Register eax is free for value, which frees ecx for name.
On 2009/06/22 19:14:23, William Hesse wrote:
> Register rax ... rcx

Done.

http://codereview.chromium.org/140068/diff/1007/13#newcode879
Line 879: // Register ecx is free for name, which guarantees eax is free for
On 2009/06/22 19:14:23, William Hesse wrote:
> rcx ... rax

Done.