Cleaned up calling-related methods in the ARM assembler.

Cleaned up calling-related methods in the ARM assembler.

Generalized Call macro instruction to handle AST IDs, too, making things more
similar to the Intel assemblers. Removed unused methods. Minor signature
cleanup. Simplified assertions.
Committed: http://code.google.com/p/v8/source/detail?r=8482

[Kevin Millikin (Chromium), 2011-06-30 08:55:02]

LGTM if small comments below are addressed.

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h
File src/arm/assembler-arm.h (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h#newcode...
src/arm/assembler-arm.h:1169: int SizeOfCodeGeneratedSince(Label* l) { return
pc_offset() - l->pos(); }
I'm not a fan of single character identifiers.  'l' is particularly bad.

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.cc
File src/arm/macro-assembler-arm.cc (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.cc#...
src/arm/macro-assembler-arm.cc:138: int32_t immediate =
reinterpret_cast<int32_t>(target);
I think this should really be intptr_t ('5.2.10.4 A pointer can be explicitly
converted to any integral type large enough to hold it.').  If that requires
changing a ton of signatures, keep it as int32_t.

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.h
File src/arm/macro-assembler-arm.h (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.h#n...
src/arm/macro-assembler-arm.h:100: RelocInfo::Mode rmode=
RelocInfo::CODE_TARGET,
Missing a space before '='.

[Sven Panne, 2011-06-30 09:29:48]

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h
File src/arm/assembler-arm.h (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h#newcode...
src/arm/assembler-arm.h:1169: int SizeOfCodeGeneratedSince(Label* l) { return
pc_offset() - l->pos(); }
On 2011/06/30 08:55:02, Kevin Millikin wrote:
> I'm not a fan of single character identifiers.  'l' is particularly bad.

Me neither, but I'll have to admit being guilty of copy-n-paste. ;-) I'll change
this and the corresponding identifiers on the other platforms.

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.cc
File src/arm/macro-assembler-arm.cc (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.cc#...
src/arm/macro-assembler-arm.cc:138: int32_t immediate =
reinterpret_cast<int32_t>(target);
On 2011/06/30 08:55:02, Kevin Millikin wrote:
> I think this should really be intptr_t ('5.2.10.4 A pointer can be explicitly
> converted to any integral type large enough to hold it.').  If that requires
> changing a ton of signatures, keep it as int32_t.

Done.

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.h
File src/arm/macro-assembler-arm.h (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/macro-assembler-arm.h#n...
src/arm/macro-assembler-arm.h:100: RelocInfo::Mode rmode=
RelocInfo::CODE_TARGET,
On 2011/06/30 08:55:02, Kevin Millikin wrote:
> Missing a space before '='.

Done.

[Rodolph Perfetta, 2011-06-30 09:46:07]

minor drive comment.

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h
File src/arm/assembler-arm.h (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h#newcode...
src/arm/assembler-arm.h:1169: int SizeOfCodeGeneratedSince(Label* l) { return
pc_offset() - l->pos(); }
There is already a method 'InstructionsGeneratedSince' in this file, do we need
both?

[Sven Panne, 2011-06-30 11:26:02]

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h
File src/arm/assembler-arm.h (right):

http://codereview.chromium.org/7284029/diff/1/src/arm/assembler-arm.h#newcode...
src/arm/assembler-arm.h:1169: int SizeOfCodeGeneratedSince(Label* l) { return
pc_offset() - l->pos(); }
On 2011/06/30 09:46:07, Rodolph Perfetta wrote:
> There is already a method 'InstructionsGeneratedSince' in this file, do we
need both?

InstructionsGeneratedSince can at least use SizeOfCodeGeneratedSince. It is a
bit unfortunate that the different assemblers don't agree in their
signatures/names for similar methods...