ARM: Don't generate memory instructions with writeback where the data and address registers are the…

ARM: Don't generate memory instructions with writeback where the data and address registers are the same.

BUG=http://dartbug.com/24855
R=fschneider@google.com

Committed: https://github.com/dart-lang/sdk/commit/c315da5a06a6cbbe2b3424b5313254721cd40785

[rmacnak, 2015-11-12 20:32:25]

rmacnak@google.com changed reviewers:
+ zra@google.com

[zra, 2015-11-12 20:37:48]

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc
File runtime/vm/stub_code_arm.cc (right):

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc...
runtime/vm/stub_code_arm.cc:442: __ mov(IP, Operand(SP));
Does this clobber IP before it has been saved? Does IP need to be saved? The
CODE_REG case seems to be worried about that.

[rmacnak, 2015-11-12 21:15:51]

rmacnak@google.com changed reviewers:
+ fschneider@google.com

[rmacnak, 2015-11-12 21:15:52]

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc
File runtime/vm/stub_code_arm.cc (right):

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc...
runtime/vm/stub_code_arm.cc:442: __ mov(IP, Operand(SP));
On 2015/11/12 20:37:48, zra wrote:
> Does this clobber IP before it has been saved? Does IP need to be saved? The
> CODE_REG case seems to be worried about that.

Hm, this does clobber IP before it has been saved.

Florian, does deopt ever use the saved value of IP? Isn't IP also potentially
clobbered by the call to this stub if the stub is at large pool offset?

[Florian Schneider, 2015-11-13 09:04:07]

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc
File runtime/vm/stub_code_arm.cc (right):

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc...
runtime/vm/stub_code_arm.cc:442: __ mov(IP, Operand(SP));
On 2015/11/12 21:15:52, rmacnak wrote:
> On 2015/11/12 20:37:48, zra wrote:
> > Does this clobber IP before it has been saved? Does IP need to be saved? The
> > CODE_REG case seems to be worried about that.
> 
> Hm, this does clobber IP before it has been saved.
> 
> Florian, does deopt ever use the saved value of IP? Isn't IP also potentially
> clobbered by the call to this stub if the stub is at large pool offset?

IP holds the live value of LR at entry where they are swapped - so the assert
above is overly strict and can be removed.
IP will hold the return address which is stored in the frame on EnterFrame, so
it does not matter.

[Florian Schneider, 2015-11-13 11:44:27]

Lgtm otherwise.

[rmacnak, 2015-11-13 20:09:12]

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc
File runtime/vm/stub_code_arm.cc (right):

https://codereview.chromium.org/1434323003/diff/1/runtime/vm/stub_code_arm.cc...
runtime/vm/stub_code_arm.cc:442: __ mov(IP, Operand(SP));
On 2015/11/13 09:04:06, Florian Schneider wrote:
> On 2015/11/12 21:15:52, rmacnak wrote:
> > On 2015/11/12 20:37:48, zra wrote:
> > > Does this clobber IP before it has been saved? Does IP need to be saved?
The
> > > CODE_REG case seems to be worried about that.
> > 
> > Hm, this does clobber IP before it has been saved.
> > 
> > Florian, does deopt ever use the saved value of IP? Isn't IP also
potentially
> > clobbered by the call to this stub if the stub is at large pool offset?
> 
> IP holds the live value of LR at entry where they are swapped - so the assert
> above is overly strict and can be removed.
> IP will hold the return address which is stored in the frame on EnterFrame, so
> it does not matter.

Removing the assert for the CODE case.

[rmacnak, 2015-11-13 20:13:59]

Committed patchset #1 (id:1) manually as
c315da5a06a6cbbe2b3424b5313254721cd40785 (presubmit successful).