Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(181)

Issue 419803009: [x64] Get rid of the stupid SmiConstantRegister. (Closed)

Created:
6 years, 4 months ago by Benedikt Meurer
Modified:
5 years, 9 months ago
CC:
v8-dev
Project:
v8
Visibility:
Public.

Description

[x64] Get rid of the stupid SmiConstantRegister.

Patch Set 1 #

Total comments: 1
Unified diffs Side-by-side diffs Delta from patch set Stats (+11 lines, -157 lines) Patch
M src/objects.h View 1 chunk +0 lines, -1 line 0 comments Download
M src/x64/code-stubs-x64.cc View 1 chunk +0 lines, -1 line 0 comments Download
M src/x64/deoptimizer-x64.cc View 1 chunk +0 lines, -1 line 0 comments Download
M src/x64/macro-assembler-x64.h View 4 chunks +1 line, -14 lines 0 comments Download
M src/x64/macro-assembler-x64.cc View 5 chunks +10 lines, -106 lines 1 comment Download
M test/cctest/test-macro-assembler-x64.cc View 2 chunks +0 lines, -34 lines 0 comments Download

Messages

Total messages: 3 (0 generated)
Benedikt Meurer
FYI: This CL removes the stupid "global nano optimization" SmiConstantRegister crap on x64; once we ...
6 years, 4 months ago (2014-08-01 12:32:44 UTC) #1
rmcilroy
Awesome, looks good to me. I don't know the x64 code too well though so ...
6 years, 4 months ago (2014-08-01 14:46:57 UTC) #2
titzer
6 years, 4 months ago (2014-08-01 15:06:16 UTC) #3
On 2014/08/01 14:46:57, rmcilroy wrote:
> Awesome, looks good to me.  I don't know the x64 code too well though so I'll
> leave it to someone else to lg.
> 
>
https://codereview.chromium.org/419803009/diff/1/src/x64/macro-assembler-x64.cc
> File src/x64/macro-assembler-x64.cc (right):
> 
>
https://codereview.chromium.org/419803009/diff/1/src/x64/macro-assembler-x64....
> src/x64/macro-assembler-x64.cc:2958: // TODO(bmeurer): r12 is
> kSmiConstantRegister.
> nit - r12 is currently unused (will be kConstantPoolPointer)

I think we decided not to do this until we're exactly sure what we're going to
do with the constant pool in TurboFan. In principle we could free up R12 to be
allocatable, but then we might have to take back again. So probably best to
leave it this way.

Powered by Google App Engine
This is Rietveld 408576698