ARM: Assign "actuals" at call site to the appropriate GPR/stack slot.

ARM: Assign "actuals" at call site to the appropriate GPR/stack slot.

Actually assign arguments to r0-r3 at the call site. Previously
this was left unhandled. There was only logic for pulling
formal parameters out of r0-r3.

Refactor the GPR counter and move it into a class so that the
rounding up for i64 arguments is in one place for callsites
and for pulling out of parameters. We might be able to use a
similar pattern to count the FP/SIMD registers later.

BUG= https://code.google.com/p/nativeclient/issues/detail?id=4076
R=stichnot@chromium.org

Committed: https://gerrit.chromium.org/gerrit/gitweb?p=native_client/pnacl-subzero.git;a=commit;h=b0a8c24ecd98f4b6748f8fa8f8c4a254a48d4889

[jvoung (off chromium), 2015-06-18 17:08:51]

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

[Jim Stichnoth, 2015-06-18 19:47:34]

otherwise lgtm

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.cpp
File src/IceInstARM32.cpp (right):

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.cpp#ne...
src/IceInstARM32.cpp:648: << "\tsp, ";
There is both "sp" and getSrc(0).  Should they both be the same, maybe choose
getSrc(0)?

Also, maybe use "\t" "sp, " for greppablerability.

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.cpp#ne...
src/IceInstARM32.cpp:665: Str << "sp = sub.i32 sp, " << Amount;
Maybe add SrcAmount to the dump output for completeness?

Also, be consistent with emit() in the use of "sp" vs getSrc(0).

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.h
File src/IceInstARM32.h (right):

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.h#newc...
src/IceInstARM32.h:611: // updates the stack offset during code emission.
It would be helpful to document what appears at first glance to be a redundancy
between Amount and SrcAmount.  (It took me some digging to realize it was
because a large Amount has to be copied into a register.

I look forward to seeing the test of a non-Constant SrcAmount. :)

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.h#newc...
src/IceInstARM32.h:631: SizeT Amount;
const?

[jvoung (off chromium), 2015-06-18 21:18:30]

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.cpp
File src/IceInstARM32.cpp (right):

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.cpp#ne...
src/IceInstARM32.cpp:648: << "\tsp, ";
On 2015/06/18 19:47:33, stichnot wrote:
> There is both "sp" and getSrc(0).  Should they both be the same, maybe choose
> getSrc(0)?
> 
> Also, maybe use "\t" "sp, " for greppablerability.

Hmm yeah, I'll just use the actual operand.

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.cpp#ne...
src/IceInstARM32.cpp:665: Str << "sp = sub.i32 sp, " << Amount;
On 2015/06/18 19:47:33, stichnot wrote:
> Maybe add SrcAmount to the dump output for completeness?
> 
> Also, be consistent with emit() in the use of "sp" vs getSrc(0).

Done.

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.h
File src/IceInstARM32.h (right):

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.h#newc...
src/IceInstARM32.h:611: // updates the stack offset during code emission.
On 2015/06/18 19:47:33, stichnot wrote:
> It would be helpful to document what appears at first glance to be a
redundancy
> between Amount and SrcAmount.  (It took me some digging to realize it was
> because a large Amount has to be copied into a register.
> 

Yep -- okay done.

> I look forward to seeing the test of a non-Constant SrcAmount. :)

Ah yes =)

https://codereview.chromium.org/1187513006/diff/80001/src/IceInstARM32.h#newc...
src/IceInstARM32.h:631: SizeT Amount;
On 2015/06/18 19:47:33, stichnot wrote:
> const?

Done.

[jvoung (off chromium), 2015-06-18 22:00:21]

Committed patchset #6 (id:90001) manually as
b0a8c24ecd98f4b6748f8fa8f8c4a254a48d4889 (tree was closed).