Issue 7992003: Porting r9392 to x64 (smi-only arrays).

Issue 7992003: Porting r9392 to x64 (smi-only arrays). (Closed)

Created:
9 years, 3 months ago by Yang

Modified:
9 years, 3 months ago

Reviewers:
William Hesse

CC:
v8-dev

Base URL:
https://v8.googlecode.com/svn/branches/bleeding_edge

Visibility:
Public.

Description

Porting r9392 to x64 (smi-only arrays). Committed: http://code.google.com/p/v8/source/detail?r=9416

Patch Set 2 : save one memory access. #

Total comments: 6

Created: 9 years, 3 months ago

Download [raw] [tar.bz2]

	Unified diffs	Side-by-side diffs	Delta from patch set	Stats (+118 lines, -24 lines)			Patch
M	src/ia32/stub-cache-ia32.cc	View		2 chunks	+2 lines, -2 lines	0 comments	Download
M	src/x64/full-codegen-x64.cc	View	1	1 chunk	+15 lines, -3 lines	0 comments	Download
M	src/x64/ic-x64.cc	View		1 chunk	+14 lines, -0 lines	2 comments	Download
M	src/x64/macro-assembler-x64.h	View		1 chunk	+12 lines, -0 lines	0 comments	Download
M	src/x64/macro-assembler-x64.cc	View		1 chunk	+24 lines, -0 lines	0 comments	Download
M	src/x64/stub-cache-x64.cc	View		5 chunks	+45 lines, -14 lines	2 comments	Download
M	test/mjsunit/element-kind.js	View		3 chunks	+6 lines, -5 lines	2 comments	Download

Messages

Total messages: 3 (0 generated)

Expand Messages | Collapse Messages

William Hesse

LGTM, with comments. http://codereview.chromium.org/7992003/diff/3001/src/x64/ic-x64.cc File src/x64/ic-x64.cc (right): http://codereview.chromium.org/7992003/diff/3001/src/x64/ic-x64.cc#newcode696 src/x64/ic-x64.cc:696: if (FLAG_smi_only_arrays) { The JumpIfNotSmi can ...

9 years, 3 months ago (2011-09-23 13:25:22 UTC) #2

Yang

9 years, 3 months ago (2011-09-23 14:19:12 UTC) #3

Addressed issues. Landing.

http://codereview.chromium.org/7992003/diff/3001/src/x64/ic-x64.cc
File src/x64/ic-x64.cc (right):

http://codereview.chromium.org/7992003/diff/3001/src/x64/ic-x64.cc#newcode696
src/x64/ic-x64.cc:696: if (FLAG_smi_only_arrays) {
On 2011/09/23 13:25:22, William Hesse wrote:
> The JumpIfNotSmi can be moved above the guard, together with the movq and ret
> for the two smi cases (now made one case).  I think this speeds and simplifies
> the code.

Done.

http://codereview.chromium.org/7992003/diff/3001/src/x64/stub-cache-x64.cc
File src/x64/stub-cache-x64.cc (right):

http://codereview.chromium.org/7992003/diff/3001/src/x64/stub-cache-x64.cc#ne...
src/x64/stub-cache-x64.cc:1522: __ movq(rcx, Operand(rsp, argc * kPointerSize));
On 2011/09/23 13:25:22, William Hesse wrote:
> Is this move to rcx now a dead value?  If so, remove it.

Done.

http://codereview.chromium.org/7992003/diff/3001/test/mjsunit/element-kind.js
File test/mjsunit/element-kind.js (right):

http://codereview.chromium.org/7992003/diff/3001/test/mjsunit/element-kind.js...
test/mjsunit/element-kind.js:117: assertKind(element_kind.fast_double_elements,
fast_double_array);
On 2011/09/23 13:25:22, William Hesse wrote:
> Why is the test for fast double elements in the change porting fast smi
> elements?

Yes, this should have been a separate CL. Kind of slipped this in when I was
looking through this test.

Expand Messages | Collapse Messages