https://codereview.chromium.org/1303403004/diff/140001/src/arm64/builtins-arm64.cc
File src/arm64/builtins-arm64.cc (right):
https://codereview.chromium.org/1303403004/diff/140001/src/arm64/builtins-arm64.cc#newcode1032
src/arm64/builtins-arm64.cc:1032: __ Drop(x1);
On 2015/08/26 13:44:36, Michael Starzinger wrote:
Not sure but isn't the ARM64 macro assembler expecting "count" instead
of size
here when a register is passed in? Shouldn't this be "Drop(x1, 1)"
instead?
Yes, good catch thanks!
https://codereview.chromium.org/1303403004/diff/140001/src/frames.h
File src/frames.h (right):
https://codereview.chromium.org/1303403004/diff/140001/src/frames.h#newcode183
src/frames.h:183:
On 2015/08/26 13:44:36, Michael Starzinger wrote:
nit: Two empty newlines.
Done.
https://codereview.chromium.org/1303403004/diff/140001/src/objects-inl.h
File src/objects-inl.h (right):
https://codereview.chromium.org/1303403004/diff/140001/src/objects-inl.h#newcode4063
src/objects-inl.h:4063: return READ_INT_FIELD(this,
kParameterSizeOffset) >> kPointerSizeLog2;
On 2015/08/26 13:44:36, Michael Starzinger wrote:
Can we leave a short one-liner comment here that the "count" is stored
as a
"size" so that generated code can use it directly?
Done.
Also is this "trick" really worth it? Don't most of the architectures
should
have instructions with the appropriate scalling factors? Not an expert
on this,
so I don't mind either way.
I don't think they are always available AFAIKT. E.g., on Arm you can
scale and immediate add or scale the offset for an ldr, but you can't
scale a reg+reg add or the result of an ldr which is what would be
needed here.
https://codereview.chromium.org/1303403004/
--
--
v8-dev mailing list
v8-dev@googlegroups.com
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to v8-dev+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
