Subzero: Fix/improve -asm-verbose output.

Subzero: Fix/improve -asm-verbose output.

Fixes a bug where a num-uses counter wasn't being updated because of C
operator && semantics.  The code was something like "if (A && --B) ..."
but we want --B to happen even when A is false.

Sorts the LiveIn and LiveOut lists by regnum so that the lists always
display the set of registers in a consistent/familiar order.

BUG= none
R=jvoung@chromium.org

Committed: https://gerrit.chromium.org/gerrit/gitweb?p=native_client/pnacl-subzero.git;a=commit;h=9a05aea88a8d9484d5f0ea1f2eabf52ee5a3647f

[Jim Stichnoth, 2015-05-26 20:12:16]

stichnot@chromium.org changed reviewers:
+ jvoung@chromium.org, kschimpf@google.com, mtrofin@chromium.org

[Karl, 2015-05-26 20:47:47]

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp
File src/IceCfgNode.cpp (left):

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp#oldc...
src/IceCfgNode.cpp:890: ++LiveRegCount[Dest->getRegNum()];
I take it that there is a different reason for this code (i.e. the
++LiveRegCount..) being removed?

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp
File src/IceCfgNode.cpp (right):

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp#newc...
src/IceCfgNode.cpp:840: if (Var->hasReg()) {
Nit: make it more obvious that the -- is the special part that must be done, or
add a comment stating this, so that it doesn't get simplified back to the old
expression in some future date.

[jvoung (off chromium), 2015-05-26 22:24:41]

LGTM -- though I don't remember what one of the ++ was for

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp
File src/IceCfgNode.cpp (left):

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp#oldc...
src/IceCfgNode.cpp:890: ++LiveRegCount[Dest->getRegNum()];
On 2015/05/26 20:47:47, Karl wrote:
> I take it that there is a different reason for this code (i.e. the
> ++LiveRegCount..) being removed?

Hmm yeah I don't remember what this adjustment was for previously?

[Jim Stichnoth, 2015-05-26 22:57:17]

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp
File src/IceCfgNode.cpp (left):

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp#oldc...
src/IceCfgNode.cpp:890: ++LiveRegCount[Dest->getRegNum()];
On 2015/05/26 20:47:47, Karl wrote:
> I take it that there is a different reason for this code (i.e. the
> ++LiveRegCount..) being removed?

Yeah - I don't know why this code was there in the first place.  If it's a
redundant assignment like A=A, then the net effect on A's liveness is unchanged,
the count shouldn't be incremented.  Probably this was an incorrect "fix" to
what this CL actually fixes.

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp
File src/IceCfgNode.cpp (right):

https://codereview.chromium.org/1152813003/diff/40001/src/IceCfgNode.cpp#newc...
src/IceCfgNode.cpp:840: if (Var->hasReg()) {
On 2015/05/26 20:47:47, Karl wrote:
> Nit: make it more obvious that the -- is the special part that must be done,
or
> add a comment stating this, so that it doesn't get simplified back to the old
> expression in some future date.

Done.

[Jim Stichnoth, 2015-05-26 23:01:28]

Committed patchset #4 (id:60001) manually as
9a05aea88a8d9484d5f0ea1f2eabf52ee5a3647f (presubmit successful).