Subzero: Auto-awesome iterators.

Subzero: Auto-awesome iterators.

Use C++11 'auto' where practical to make iteration more concise.
Use C++11 range-based for loops where possible.

BUG= none
R=jfb@chromium.org, kschimpf@google.com

Committed: https://gerrit.chromium.org/gerrit/gitweb?p=native_client/pnacl-subzero.git;a=commit;h=f44f371b7f3fab5683e6781873d71987e44fea2f

[Jim Stichnoth, 2014-10-01 13:15:15]

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

[Jim Stichnoth, 2014-10-01 13:15:27]

https://codereview.chromium.org/619893002/diff/1/src/IceCfg.cpp
File src/IceCfg.cpp (left):

https://codereview.chromium.org/619893002/diff/1/src/IceCfg.cpp#oldcode221
src/IceCfg.cpp:221: if (Variable *Var = *I)
This test was left over from the early days when the Variables list was allowed
to contain NULL values.

[Karl, 2014-10-01 15:54:08]

lgtm

https://codereview.chromium.org/619893002/diff/20001/src/IceConverter.cpp
File src/IceConverter.cpp (right):

https://codereview.chromium.org/619893002/diff/20001/src/IceConverter.cpp#new...
src/IceConverter.cpp:84: for (auto BBI = F->begin(), BBE = F->end(); BBI != BBE;
++BBI)
Why not range here?

https://codereview.chromium.org/619893002/diff/20001/src/IceConverter.cpp#new...
src/IceConverter.cpp:562: for (auto II = BB->begin(), II_e = BB->end(); II !=
II_e; ++II) {
Why not use a range here?

[jvoung (off chromium), 2014-10-01 16:10:53]

https://codereview.chromium.org/619893002/diff/20001/src/IceRegAlloc.cpp
File src/IceRegAlloc.cpp (right):

https://codereview.chromium.org/619893002/diff/20001/src/IceRegAlloc.cpp#newc...
src/IceRegAlloc.cpp:390: for (auto I = Unhandled.begin(), E = Unhandled.end();
could this be handled similar to line 315, and range-baseified?

[JF, 2014-10-01 16:21:01]

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp
File src/IceCfg.cpp (right):

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp#newcode85
src/IceCfg.cpp:85: for (auto &Node : Nodes)
for (Node *N : Nodes)

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp#newcode186
src/IceCfg.cpp:186: for (auto I = Nodes.rbegin(), E = Nodes.rend(); I != E; ++I)
{
LLVM has include/llvm/ADT/iterator_range.h which should allow you to
make_range(rbegin(), rend()).

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp#newcode305
src/IceCfg.cpp:305: for (auto I = Nodes.begin(), E = Nodes.end(); I != E; ++I) {
Range?

[Jim Stichnoth, 2014-10-01 19:15:08]

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp
File src/IceCfg.cpp (right):

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp#newcode85
src/IceCfg.cpp:85: for (auto &Node : Nodes)
On 2014/10/01 16:21:00, JF wrote:
> for (Node *N : Nodes)

Done.

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp#newcode186
src/IceCfg.cpp:186: for (auto I = Nodes.rbegin(), E = Nodes.rend(); I != E; ++I)
{
On 2014/10/01 16:21:00, JF wrote:
> LLVM has include/llvm/ADT/iterator_range.h which should allow you to
> make_range(rbegin(), rend()).

Added TODOs to add this after the LLVM 3.5 merge.

https://codereview.chromium.org/619893002/diff/20001/src/IceCfg.cpp#newcode305
src/IceCfg.cpp:305: for (auto I = Nodes.begin(), E = Nodes.end(); I != E; ++I) {
On 2014/10/01 16:21:00, JF wrote:
> Range?

I don't think so, because CfgNode::doBranchOpt() below needs the next node in
the iteration.

https://codereview.chromium.org/619893002/diff/20001/src/IceConverter.cpp
File src/IceConverter.cpp (right):

https://codereview.chromium.org/619893002/diff/20001/src/IceConverter.cpp#new...
src/IceConverter.cpp:84: for (auto BBI = F->begin(), BBE = F->end(); BBI != BBE;
++BBI)
On 2014/10/01 15:54:06, Karl wrote:
> Why not range here?

Done.

https://codereview.chromium.org/619893002/diff/20001/src/IceConverter.cpp#new...
src/IceConverter.cpp:562: for (auto II = BB->begin(), II_e = BB->end(); II !=
II_e; ++II) {
On 2014/10/01 15:54:06, Karl wrote:
> Why not use a range here?

Done.

https://codereview.chromium.org/619893002/diff/20001/src/IceRegAlloc.cpp
File src/IceRegAlloc.cpp (right):

https://codereview.chromium.org/619893002/diff/20001/src/IceRegAlloc.cpp#newc...
src/IceRegAlloc.cpp:390: for (auto I = Unhandled.begin(), E = Unhandled.end();
On 2014/10/01 16:10:53, jvoung wrote:
> could this be handled similar to line 315, and range-baseified?

Done.

[Karl, 2014-10-01 19:44:51]

lgtm

[JF, 2014-10-01 20:42:21]

lgtm

https://codereview.chromium.org/619893002/diff/60001/src/IceCfg.cpp
File src/IceCfg.cpp (right):

https://codereview.chromium.org/619893002/diff/60001/src/IceCfg.cpp#newcode338
src/IceCfg.cpp:338: for (uint8_t I : getTarget()->getNonExecBundlePadding())
typedef for bundle padding?

[Karl, 2014-10-01 21:02:02]

lgtm

[Jim Stichnoth, 2014-10-01 21:05:33]

https://codereview.chromium.org/619893002/diff/60001/src/IceCfg.cpp
File src/IceCfg.cpp (right):

https://codereview.chromium.org/619893002/diff/60001/src/IceCfg.cpp#newcode338
src/IceCfg.cpp:338: for (uint8_t I : getTarget()->getNonExecBundlePadding())
On 2014/10/01 20:42:21, JF wrote:
> typedef for bundle padding?

Done.

[Jim Stichnoth, 2014-10-01 21:06:13]

Committed patchset #5 (id:80001) manually as
f44f371b7f3fab5683e6781873d71987e44fea2f (presubmit successful).