[arm64] Use CMN for cmp(a,sub(0,b)) only when checking equality/inequality.

[arm64] Use CMN for cmp(a,sub(0,b)) only when checking equality/inequality.

We were previously incorrectly changing:
  sub r0, 0, r1
  cmp r2, r0
  b.cond <addr>
to:
  cmn r2, r1
  b.cond <addr>

for all conditions. This is incorrect for conditions involving the C (carry)
and V (overflow) flags, and in particular in the case where r1 = INT_MIN.
The optimization is still safe to perform for Equal and NotEqual since they
do not depend on the C and V flags.

BUG=

Committed: https://crrev.com/fdb0f07887ef70fa8e5fbb077d75687eb4b0aebe
Cr-Commit-Position: refs/heads/master@{#39246}

[georgia.kouveli, 2016-09-07 10:29:45]

georgia.kouveli@arm.com changed reviewers:
+ bmeurer@chromium.org, epertoso@chromium.org

[epertoso, 2016-09-07 11:45:28]

lgtm

 thanks!

[georgia.kouveli, 2016-09-07 12:19:03]

The CQ bit was checked by georgia.kouveli@arm.com

[commit-bot: I haz the power, 2016-09-07 12:19:19]

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, 2016-09-07 12:42:13]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-09-07 12:43:09]

Description was changed from

==========
[arm64] Use CMN for cmp(a,sub(0,b)) only when checking equality/inequality.

We were previously incorrectly changing:
  sub r0, 0, r1
  cmp r2, r0
  b.cond <addr>
to:
  cmn r2, r1
  b.cond <addr>

for all conditions. This is incorrect for conditions involving the C (carry)
and V (overflow) flags, and in particular in the case where r1 = INT_MIN.
The optimization is still safe to perform for Equal and NotEqual since they
do not depend on the C and V flags.

BUG=
==========

to

==========
[arm64] Use CMN for cmp(a,sub(0,b)) only when checking equality/inequality.

We were previously incorrectly changing:
  sub r0, 0, r1
  cmp r2, r0
  b.cond <addr>
to:
  cmn r2, r1
  b.cond <addr>

for all conditions. This is incorrect for conditions involving the C (carry)
and V (overflow) flags, and in particular in the case where r1 = INT_MIN.
The optimization is still safe to perform for Equal and NotEqual since they
do not depend on the C and V flags.

BUG=

Committed: https://crrev.com/fdb0f07887ef70fa8e5fbb077d75687eb4b0aebe
Cr-Commit-Position: refs/heads/master@{#39246}
==========

[commit-bot: I haz the power, 2016-09-07 12:43:10]

Patchset 1 (id:??) landed as
https://crrev.com/fdb0f07887ef70fa8e5fbb077d75687eb4b0aebe
Cr-Commit-Position: refs/heads/master@{#39246}