Issue 5741003: Allow optimizing compiler to compute Math.log using untagged doubles.

Issue 5741003: Allow optimizing compiler to compute Math.log using untagged doubles. (Closed)

Created:
10 years ago by William Hesse

Modified:
9 years, 7 months ago

Reviewers:
fschneider

CC:
v8-dev

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

Visibility:
Public.

Description

Allow optimizing compiler to compute Math.log using untagged doubles. Committed: http://code.google.com/p/v8/source/detail?r=5989

Patch Set 3 : '' #

Created: 10 years ago

Download [raw] [tar.bz2]

	Unified diffs	Side-by-side diffs	Delta from patch set	Stats (+301 lines, -52 lines)			Patch
M	src/arm/lithium-arm.cc	View	1 2	1 chunk	+3 lines, -0 lines	0 comments	Download
M	src/code-stubs.h	View	1	1 chunk	+1 line, -0 lines	0 comments	Download
M	src/heap.h	View	1	1 chunk	+2 lines, -1 line	0 comments	Download
M	src/hydrogen.cc	View	1	1 chunk	+1 line, -0 lines	0 comments	Download
M	src/hydrogen-instructions.h	View	1	2 chunks	+2 lines, -0 lines	0 comments	Download
M	src/ia32/assembler-ia32.h	View	1	1 chunk	+4 lines, -1 line	0 comments	Download
M	src/ia32/assembler-ia32.cc	View	1	4 chunks	+39 lines, -2 lines	0 comments	Download
M	src/ia32/code-stubs-ia32.h	View	1	2 chunks	+19 lines, -0 lines	0 comments	Download
M	src/ia32/code-stubs-ia32.cc	View	1	1 chunk	+131 lines, -0 lines	0 comments	Download
M	src/ia32/disasm-ia32.cc	View	1	5 chunks	+64 lines, -32 lines	0 comments	Download
M	src/ia32/lithium-codegen-ia32.h	View	1	1 chunk	+1 line, -0 lines	0 comments	Download
M	src/ia32/lithium-codegen-ia32.cc	View	1	2 chunks	+11 lines, -0 lines	0 comments	Download
M	src/ia32/lithium-ia32.cc	View	1	1 chunk	+22 lines, -16 lines	0 comments	Download
M	src/math.js	View	1	1 chunk	+1 line, -0 lines	0 comments	Download

Messages

Total messages: 3 (0 generated)

Expand Messages | Collapse Messages

fschneider

LGTM. http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc File src/ia32/code-stubs-ia32.cc (right): http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc#newcode2696 src/ia32/code-stubs-ia32.cc:2696: // Input is a HeapNumber. Input is a ...

10 years ago (2010-12-13 13:16:32 UTC) #2

William Hesse

10 years ago (2010-12-13 14:15:47 UTC) #3

http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc
File src/ia32/code-stubs-ia32.cc (right):

http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc#new...
src/ia32/code-stubs-ia32.cc:2696: // Input is a HeapNumber.
On 2010/12/13 13:16:33, fschneider wrote:
> Input is a double in xmm1.

Done.

http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc#new...
src/ia32/code-stubs-ia32.cc:2718: // ST[0] == double value.
On 2010/12/13 13:16:33, fschneider wrote:
> That comments seems out-of-place.

Done.

http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc#new...
src/ia32/code-stubs-ia32.cc:2763: __ sub(Operand(esp), Immediate(2 *
kPointerSize));
On 2010/12/13 13:16:33, fschneider wrote:
> Rather use sizeof(double) instead of 2*kPointersize since it's better portable
> to 64-bit.

Done.

http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc#new...
src/ia32/code-stubs-ia32.cc:2795: Runtime::FunctionId
TranscendentalCacheSSE2Stub::RuntimeFunction() {
On 2010/12/13 13:16:33, fschneider wrote:
> I'm not sure what this function is used for. Is it used at all?

Started using it.  It is needed when more than one transcendental function is
involved.

http://codereview.chromium.org/5741003/diff/1/src/ia32/code-stubs-ia32.cc#new...
src/ia32/code-stubs-ia32.cc:2809: // (a newly allocated HeapNumber) is in eax.
On 2010/12/13 13:16:33, fschneider wrote:
> This comment does not fit here.

Some of it is needed - only edi may be used as a scratch register, and the input
is in ST[0] and xmm1.

Expand Messages | Collapse Messages