hi Serya, May you take another look. Thanks, Zaheer
http://codereview.chromium.org/6170001/diff/65002/src/arm/frames-arm.h File src/arm/frames-arm.h (right): http://codereview.chromium.org/6170001/diff/65002/src/arm/frames-arm.h#newcode116 src/arm/frames-arm.h:116: static const int kMarkerOffset = -2 * kPointerSize; On 2011/01/21 09:52:36, SeRya wrote:
Actually the same technique (used fixed place in the stack frame for
the return
address) could be used to avoid this marker (and simplify exit frame
layout). I
don't think it worth to do it in this CL but you can update this TODO.
Done. http://codereview.chromium.org/6170001/diff/65002/src/arm/macro-assembler-arm.cc File src/arm/macro-assembler-arm.cc (right): http://codereview.chromium.org/6170001/diff/65002/src/arm/macro-assembler-arm.cc#newcode1401 src/arm/macro-assembler-arm.cc:1401: EnterExitFrame(false, arg_stack_space + 5); On 2011/01/21 09:52:36, SeRya wrote:
+1?
Done. updated EnterExitFrame to encapsulate the pushes in that function. Note currently it does not take in to account pushes with save_doubles=true (as the existing code) http://codereview.chromium.org/6170001/diff/65002/src/arm/simulator-arm.cc File src/arm/simulator-arm.cc (right): http://codereview.chromium.org/6170001/diff/65002/src/arm/simulator-arm.cc#newcode1590 src/arm/simulator-arm.cc:1590: SimulatorRuntimeApiCall target = On 2011/01/21 09:52:36, SeRya wrote:
May be to add ASSERT that where is indeed one argument?
PrepareCallApiFunction
looks quite general, but in simulator only one argument is actually
passed. simulator also uses stack space for the arguments (ie.. behavior is exactly same as in device). r0 points to Arguments& http://codereview.chromium.org/6170001/ -- v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev
