[ignition] Use absolute values for jump offsets

[ignition] Use absolute values for jump offsets

Since JumpLoop is always backwards, and other jumps are always forwards,
we can store the jump offset as an always positive integer and decide on
the jump direction based on the bytecode. This will save a small amount
of space for large-ish for loops (>128 bytecodes).

Review-Url: https://codereview.chromium.org/2641443002
Cr-Commit-Position: refs/heads/master@{#42638}
Committed: https://chromium.googlesource.com/v8/v8/+/e56437b630198d86a3114cba34cd181345f3a5b4

[Leszek Swirski, 2017-01-17 15:59:27]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-17 15:59:29]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-17 16:15:06]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-17 16:15:07]

Dry run: Try jobs failed on following builders:
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux64_gyp_rel_ng_triggered on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng_trigg...)

[Leszek Swirski, 2017-01-17 16:56:29]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-17 16:56:33]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-17 17:00:03]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-17 17:00:03]

Dry run: Try jobs failed on following builders:
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux_mips64el_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_mips64el_compile_r...)
  v8_linux_mipsel_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_mipsel_compile_rel...)
  v8_linux_nodcheck_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng/bu...)

[Leszek Swirski, 2017-01-17 17:06:13]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-17 17:06:18]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-17 17:31:50]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-17 17:31:51]

Dry run: This issue passed the CQ dry run.

[Leszek Swirski, 2017-01-17 17:33:22]

leszeks@chromium.org changed reviewers:
+ rmcilroy@chromium.org

[Leszek Swirski, 2017-01-17 17:33:23]

Hey Ross,

What do you think of this small change? I mentioned this over coffee, the
regression I'd seen was due to incorrect interrupt handling.

[rmcilroy, 2017-01-17 17:58:39]

https://codereview.chromium.org/2641443002/diff/40001/src/interpreter/interpr...
File src/interpreter/interpreter-assembler.cc (right):

https://codereview.chromium.org/2641443002/diff/40001/src/interpreter/interpr...
src/interpreter/interpreter-assembler.cc:911: if (reverse) {
Could this be done separately (or maybe not at all depending on the impact it
has). Doing this will make our interrupt budget calculation less accurate since
we will assume we've executed any bytecodes that we actually jump over. This is
what FCG did, but it is probably better that we keep this accurate unless
updating the budget on forward jumps has a significant perf impact.

[Leszek Swirski, 2017-01-18 16:53:30]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-18 16:53:35]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-18 16:56:25]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-18 16:56:26]

Dry run: Try jobs failed on following builders:
  v8_linux64_avx2_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel_ng/buil...)
  v8_linux64_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_rel_ng/builds/19447)
  v8_win_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng/builds/21157)

[Leszek Swirski, 2017-01-18 16:57:22]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-18 16:57:25]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-18 17:21:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-18 17:21:58]

Dry run: This issue passed the CQ dry run.

[Leszek Swirski, 2017-01-18 17:22:28]

https://codereview.chromium.org/2641443002/diff/40001/src/interpreter/interpr...
File src/interpreter/interpreter-assembler.cc (right):

https://codereview.chromium.org/2641443002/diff/40001/src/interpreter/interpr...
src/interpreter/interpreter-assembler.cc:911: if (reverse) {
On 2017/01/17 17:58:38, rmcilroy wrote:
> Could this be done separately (or maybe not at all depending on the impact it
> has). Doing this will make our interrupt budget calculation less accurate
since
> we will assume we've executed any bytecodes that we actually jump over. This
is
> what FCG did, but it is probably better that we keep this accurate unless
> updating the budget on forward jumps has a significant perf impact.

Sure, added a "reverse" flag same as for advance. I was just doing the TODO in
UpdateInterruptBudget, should we simply get rid of that TODO?

[rmcilroy, 2017-01-19 09:22:38]

Yeah, let's get rid of the todo unless you saw some significant perf improvement
from it (I don't think we would get one, and it would probably cause more
unecessary optimization in top25).

Could you also update the description to remove mention of that change.

https://codereview.chromium.org/2641443002/diff/80001/src/interpreter/interpr...
File src/interpreter/interpreter-assembler.h (right):

https://codereview.chromium.org/2641443002/diff/80001/src/interpreter/interpr...
src/interpreter/interpreter-assembler.h:157: compiler::Node*
Jump(compiler::Node* jump_offset, bool backwards = false);
Nit, could you just make this a seperate JumpBackwards bytecode.

https://codereview.chromium.org/2641443002/diff/80001/test/cctest/interpreter...
File test/cctest/interpreter/bytecode_expectations/BasicLoops.golden (left):

https://codereview.chromium.org/2641443002/diff/80001/test/cctest/interpreter...
test/cctest/interpreter/bytecode_expectations/BasicLoops.golden:91: B(JumpLoop),
U8(-40), U8(0),
Could you update the bytecode disassembler to show JumpLoop with backward
offsets. You could also update the bytecode expectation printer which prints
this, but Inn less worried about that if it's tricky.

[Leszek Swirski, 2017-01-19 10:47:34]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-19 10:47:43]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Leszek Swirski, 2017-01-19 10:52:59]

https://codereview.chromium.org/2641443002/diff/80001/src/interpreter/interpr...
File src/interpreter/interpreter-assembler.h (right):

https://codereview.chromium.org/2641443002/diff/80001/src/interpreter/interpr...
src/interpreter/interpreter-assembler.h:157: compiler::Node*
Jump(compiler::Node* jump_offset, bool backwards = false);
On 2017/01/19 09:22:38, rmcilroy wrote:
> Nit, could you just make this a seperate JumpBackwards bytecode.

Done, though this function still exists (private) as a helper to remove
duplicate code.

Also, named it JumpBackward instead of JumpBackwards since we optimize rather
than optimising.

https://codereview.chromium.org/2641443002/diff/80001/test/cctest/interpreter...
File test/cctest/interpreter/bytecode_expectations/BasicLoops.golden (left):

https://codereview.chromium.org/2641443002/diff/80001/test/cctest/interpreter...
test/cctest/interpreter/bytecode_expectations/BasicLoops.golden:91: B(JumpLoop),
U8(-40), U8(0),
On 2017/01/19 09:22:38, rmcilroy wrote:
> Could you update the bytecode disassembler to show JumpLoop with backward
> offsets. You could also update the bytecode expectation printer which prints
> this, but Inn less worried about that if it's tricky.

Not sure what you mean here, do you think we should twiddle with the operand
value printing, to special case it for JumpLoop? Seems a bit messy, and the
target address printed is already correct.

[Leszek Swirski, 2017-01-19 10:53:21]

Description was changed from

==========
[ignition] Use absolute values for jump offsets

Since JumpLoop is always backwards, and other jumps are always forwards,
we can store the jump offset as an always positive integer and decide on
the jump direction based on the bytecode. This will save a small amount
of space for large-ish for loops (>128 bytecodes), and resolves a TODO
for updating the interrupt budget for reverse jumps only.
==========

to

==========
[ignition] Use absolute values for jump offsets

Since JumpLoop is always backwards, and other jumps are always forwards,
we can store the jump offset as an always positive integer and decide on
the jump direction based on the bytecode. This will save a small amount
of space for large-ish for loops (>128 bytecodes).
==========

[commit-bot: I haz the power, 2017-01-19 11:02:42]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-19 11:02:43]

Dry run: Try jobs failed on following builders:
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux64_gyp_rel_ng_triggered on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng_trigg...)

[rmcilroy, 2017-01-19 12:13:46]

Looks like you've got some redness, but LGTM once it runs.

https://codereview.chromium.org/2641443002/diff/80001/test/cctest/interpreter...
File test/cctest/interpreter/bytecode_expectations/BasicLoops.golden (left):

https://codereview.chromium.org/2641443002/diff/80001/test/cctest/interpreter...
test/cctest/interpreter/bytecode_expectations/BasicLoops.golden:91: B(JumpLoop),
U8(-40), U8(0),
On 2017/01/19 10:52:59, Leszek Swirski wrote:
> On 2017/01/19 09:22:38, rmcilroy wrote:
> > Could you update the bytecode disassembler to show JumpLoop with backward
> > offsets. You could also update the bytecode expectation printer which prints
> > this, but Inn less worried about that if it's tricky.
> 
> Not sure what you mean here, do you think we should twiddle with the operand
> value printing, to special case it for JumpLoop? Seems a bit messy, and the
> target address printed is already correct.

I thought we already did that in some circumstances (e.g., CallRuntime) but it
seems we factored that all out into operand types, so fair enough let's not do
this, I guess that would be messy.

[Leszek Swirski, 2017-01-19 12:29:41]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-19 12:29:50]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-19 12:45:05]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-19 12:45:05]

Dry run: Try jobs failed on following builders:
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux64_gyp_rel_ng_triggered on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng_trigg...)

[Leszek Swirski, 2017-01-20 14:07:55]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-20 14:07:58]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Leszek Swirski, 2017-01-20 14:16:40]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-20 14:16:48]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-20 14:36:51]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-20 14:36:53]

Dry run: Try jobs failed on following builders:
  v8_linux_dbg_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_dbg_ng/builds/19693)
  v8_linux_dbg_ng_triggered on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_dbg_ng_triggered/b...)

[Leszek Swirski, 2017-01-20 14:50:42]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-20 14:50:53]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-20 15:27:10]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-20 15:27:11]

Dry run: This issue passed the CQ dry run.

[Leszek Swirski, 2017-01-24 20:09:47]

The CQ bit was checked by leszeks@chromium.org

[Leszek Swirski, 2017-01-24 20:09:48]

The patchset sent to the CQ was uploaded after l-g-t-m from
rmcilroy@chromium.org
Link to the patchset: https://codereview.chromium.org/2641443002/#ps160001
(title: "Fix tests")

[commit-bot: I haz the power, 2017-01-24 20:09:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-24 20:44:50]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-01-24 20:44:52]

Failed to apply patch for
test/cctest/interpreter/bytecode_expectations/Generators.golden:
While running git apply --index -p1;
  error: patch failed:
test/cctest/interpreter/bytecode_expectations/Generators.golden:493
  error: test/cctest/interpreter/bytecode_expectations/Generators.golden: patch
does not apply

Patch:       test/cctest/interpreter/bytecode_expectations/Generators.golden
Index: test/cctest/interpreter/bytecode_expectations/Generators.golden
diff --git a/test/cctest/interpreter/bytecode_expectations/Generators.golden
b/test/cctest/interpreter/bytecode_expectations/Generators.golden
index
a53f7f622f2354936908d09ec4f77b05e7223eb5..b47a841a10a6c892ad12cf977e45c84315f7218d
100644
--- a/test/cctest/interpreter/bytecode_expectations/Generators.golden
+++ b/test/cctest/interpreter/bytecode_expectations/Generators.golden
@@ -138,7 +138,7 @@ bytecodes: [
                 B(JumpIfTrue), U8(60),
                 B(LdaSmi), U8(1),
                 B(TestEqualStrict), R(1), U8(0),
-                B(JumpIfTrueConstant), U8(0),
+                B(JumpIfTrue), U8(130),
                 B(LdaSmi), U8(78),
                 B(Star), R(3),
                 B(CallRuntime), U16(Runtime::kAbort), R(3), U8(1),
@@ -268,7 +268,6 @@ bytecodes: [
   /*   25 S> */ B(Return),
 ]
 constant pool: [
-  Smi [130],
 ]
 handlers: [
   [53, 228, 234],
@@ -281,7 +280,7 @@ snippet: "
 "
 frame size: 18
 parameter count: 1
-bytecode array length: 778
+bytecode array length: 775
 bytecodes: [
                 B(Ldar), R(new_target),
                 B(JumpIfUndefined), U8(35),
@@ -294,7 +293,7 @@ bytecodes: [
                 B(JumpIfTrue), U8(60),
                 B(LdaSmi), U8(1),
                 B(TestEqualStrict), R(3), U8(0),
-                B(JumpIfTrueConstant), U8(3),
+                B(JumpIfTrue), U8(154),
                 B(LdaSmi), U8(78),
                 B(Star), R(5),
                 B(CallRuntime), U16(Runtime::kAbort), R(5), U8(1),
@@ -338,7 +337,7 @@ bytecodes: [
                 B(Star), R(6),
                 B(LdaZero),
                 B(Star), R(5),
-                B(JumpConstant), U8(18),
+                B(JumpConstant), U8(13),
                 B(Ldar), R(10),
   /*   11 E> */ B(Throw),
                 B(Ldar), R(closure),
@@ -363,13 +362,13 @@ bytecodes: [
                 B(JumpIfTrue), U8(18),
                 B(LdaSmi), U8(1),
                 B(TestEqualStrict), R(3), U8(0),
-                B(JumpIfTrueConstant), U8(8),
+                B(JumpIfTrue), U8(134),
                 B(LdaSmi), U8(78),
                 B(Star), R(12),
                 B(CallRuntime), U16(Runtime::kAbort), R(12), U8(1),
   /*   27 S> */ B(LdaContextSlot), R(1), U8(7), U8(0),
                 B(Star), R(13),
-                B(LdaNamedProperty), R(13), U8(4), U8(8),
+                B(LdaNamedProperty), R(13), U8(3), U8(8),
                 B(Star), R(12),
   /*   27 E> */ B(CallProperty), R(12), R(13), U8(1), U8(6),
   /*   27 E> */ B(StaContextSlot), R(1), U8(8), U8(0),
@@ -382,11 +381,11 @@ bytecodes: [
                 B(CallRuntime), U16(Runtime::kThrowIteratorResultNotAnObject),
R(12), U8(1),
                 B(LdaContextSlot), R(1), U8(8), U8(0),
                 B(Star), R(12),
-                B(LdaNamedProperty), R(12), U8(5), U8(10),
-                B(JumpIfToBooleanTrueConstant), U8(9),
+                B(LdaNamedProperty), R(12), U8(4), U8(10),
+                B(JumpIfToBooleanTrue), U8(147),
                 B(LdaContextSlot), R(1), U8(8), U8(0),
                 B(Star), R(12),
-                B(LdaNamedProperty), R(12), U8(6), U8(12),
+                B(LdaNamedProperty), R(12), U8(5), U8(12),
                 B(StaContextSlot), R(1), U8(10), U8(0),
                 B(LdaSmi), U8(2),
                 B(StaContextSlot), R(1), U8(9), U8(0),
@@ -394,7 +393,7 @@ bytecodes: [
                 B(StaContextSlot), R(1), U8(6), U8(0),
   /*   16 E> */ B(StackCheck),
                 B(Ldar), R(closure),
-                B(CreateBlockContext), U8(7),
+                B(CreateBlockContext), U8(6),
                 B(PushContext), R(2),
                 B(LdaTheHole),
                 B(StaCurrentContextSlot), U8(4),
@@ -438,17 +437,17 @@ bytecodes: [
                 B(Star), R(9),
                 B(LdaZero),
                 B(Star), R(8),
-                B(Jump), U8(74),
+                B(Jump), U8(71),
                 B(Ldar), R(14),
   /*   36 E> */ B(Throw),
                 B(PopContext), R(2),
                 B(LdaZero),
                 B(StaContextSlot), R(1), U8(9), U8(0),
-                B(Wide), B(JumpLoop), U16(-219), U16(0),
+                B(JumpLoop), U8(218), U8(0),
                 B(Jump), U8(44),
                 B(Star), R(12),
                 B(Ldar), R(closure),
-                B(CreateCatchContext), R(12), U8(10), U8(11),
+                B(CreateCatchContext), R(12), U8(7), U8(8),
                 B(Star), R(11),
                 B(PushContext), R(2),
                 B(LdaContextSlot), R(1), U8(9), U8(0),
@@ -475,16 +474,16 @@ bytecodes: [
                 B(Star), R(11),
                 B(LdaZero),
                 B(TestEqualStrict), R(11), U8(15),
-                B(JumpIfTrueConstant), U8(17),
+                B(JumpIfTrue), U8(159),
                 B(LdaContextSlot), R(1), U8(7), U8(0),
                 B(Star), R(11),
-                B(LdaNamedProperty), R(11), U8(12), U8(16),
+                B(LdaNamedProperty), R(11), U8(9), U8(16),
                 B(StaContextSlot), R(1), U8(11), U8(0),
                 B(LdaContextSlot), R(1), U8(11), U8(0),
                 B(Star), R(11),
                 B(TestUndetectable), R(11),
                 B(JumpIfFalse), U8(4),
-                B(JumpConstant), U8(16),
+                B(Jump), U8(133),
                 B(LdaContextSlot), R(1), U8(9), U8(0),
                 B(Star), R(11),
                 B(LdaSmi), U8(1),
@@ -493,13 +492,13 @@ bytecodes: [
                 B(LdaContextSlot), R(1), U8(11), U8(0),
                 B(TypeOf),
                 B(Star), R(11),
-                B(LdaConstant), U8(13),
+                B(LdaConstant), U8(10),
                 B(TestEqualStrict), R(11), U8(20),
                 B(JumpIfFalse), U8(4),
                 B(Jump), U8(18),
                 B(Wide), B(LdaSmi), U16(129),
                 B(Star), R(11),
-                B(LdaConstant), U8(14),
+                B(LdaConstant), U8(11),
                 B(Star), R(12),
                 B(CallRuntime), U16(Runtime::kNewTypeError), R(11), U8(2),
                 B(Throw),
@@ -512,7 +511,7 @@ bytecodes: [
                 B(Jump), U8(20),
                 B(Star), R(12),
                 B(Ldar), R(closure),
-                B(CreateCatchContext), R(12), U8(10), U8(15),
+                B(CreateCatchContext), R(12), U8(7), U8(12),
                 B(Star), R(11),
                 B(LdaTheHole),
                 B(SetPendingMessage),
@@ -612,27 +611,22 @@ constant pool: [
   FIXED_ARRAY_TYPE,
   CONSTANT_ELEMENTS_PAIR_TYPE,
   SYMBOL_TYPE,
-  Smi [154],
   ONE_BYTE_INTERNALIZED_STRING_TYPE ["next"],
   ONE_BYTE_INTERNALIZED_STRING_TYPE ["done"],
   ONE_BYTE_INTERNALIZED_STRING_TYPE ["value"],
   FIXED_ARRAY_TYPE,
-  Smi [134],
-  Smi [150],
   ONE_BYTE_INTERNALIZED_STRING_TYPE [".catch"],
   FIXED_ARRAY_TYPE,
   ONE_BYTE_INTERNALIZED_STRING_TYPE ["return"],
   ONE_BYTE_INTERNALIZED_STRING_TYPE ["function"],
   ONE_BYTE_INTERNALIZED_STRING_TYPE [""],
   FIXED_ARRAY_TYPE,
-  Smi [133],
-  Smi [159],
-  Smi [582],
+  Smi [579],
 ]
 handlers: [
-  [53, 697, 703],
-  [149, 448, 454],
-  [152, 404, 406],
-  [550, 566, 568],
+  [53, 694, 700],
+  [149, 445, 451],
+  [152, 401, 403],
+  [547, 563, 565],
 ]

[Leszek Swirski, 2017-01-24 21:44:58]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-24 21:45:01]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Leszek Swirski, 2017-01-24 21:45:07]

The CQ bit was checked by leszeks@chromium.org

[Leszek Swirski, 2017-01-24 21:45:08]

The patchset sent to the CQ was uploaded after l-g-t-m from
rmcilroy@chromium.org
Link to the patchset: https://codereview.chromium.org/2641443002/#ps180001
(title: "Rebase")

[commit-bot: I haz the power, 2017-01-24 21:45:12]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Leszek Swirski, 2017-01-24 21:45:15]

The CQ bit was checked by leszeks@chromium.org

[commit-bot: I haz the power, 2017-01-24 21:45:22]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-24 22:09:03]

CQ is committing da patch.

Bot data: {"patchset_id": 180001, "attempt_start_ts": 1485294315984590,
"parent_rev": "82631263c614286b7ebacfb89d09a3619d6e1ade", "commit_rev":
"e56437b630198d86a3114cba34cd181345f3a5b4"}

[commit-bot: I haz the power, 2017-01-24 22:09:08]

Description was changed from

==========
[ignition] Use absolute values for jump offsets

Since JumpLoop is always backwards, and other jumps are always forwards,
we can store the jump offset as an always positive integer and decide on
the jump direction based on the bytecode. This will save a small amount
of space for large-ish for loops (>128 bytecodes).
==========

to

==========
[ignition] Use absolute values for jump offsets

Since JumpLoop is always backwards, and other jumps are always forwards,
we can store the jump offset as an always positive integer and decide on
the jump direction based on the bytecode. This will save a small amount
of space for large-ish for loops (>128 bytecodes).

Review-Url: https://codereview.chromium.org/2641443002
Cr-Commit-Position: refs/heads/master@{#42638}
Committed:
https://chromium.googlesource.com/v8/v8/+/e56437b630198d86a3114cba34cd181345f...
==========

[commit-bot: I haz the power, 2017-01-24 22:09:10]

Committed patchset #10 (id:180001) as
https://chromium.googlesource.com/v8/v8/+/e56437b630198d86a3114cba34cd181345f...