LGTM
http://codereview.chromium.org/1735007/diff/15001/16001 File src/arm/codegen-arm.cc (right): http://codereview.chromium.org/1735007/diff/15001/16001#newcode5286 src/arm/codegen-arm.cc:5286: // kayed load has been inlined. kayed -> keyed. http://codereview.chromium.org/1735007/diff/15001/16001#newcode5362 src/arm/codegen-arm.cc:5362: // Parts of this code is patched, so the exact instructions generated needs Parts ... is -> Parts ... are instructions ... needs -> instructions ... need http://codereview.chromium.org/1735007/diff/15001/16001#newcode5389 src/arm/codegen-arm.cc:5389: __ tst(r0, Operand(kSmiTagMask)); On 2010/04/23 10:37:41, antonm wrote:
r0 is an object, no? Should it be r1 and check that key is smi?
And sorry for being pushy but there is BranchOnSmi
I don't think it works for deferred. Perhaps we should have BranchOnSmi on deferred. http://codereview.chromium.org/1735007/diff/15001/16001#newcode5393 src/arm/codegen-arm.cc:5393: // is not a dictionary. Can it become a dictionary without changing the map? http://codereview.chromium.org/1735007/diff/15001/16001#newcode5410 src/arm/codegen-arm.cc:5410: __ cmp(r1, r3); Surely you should be comparing r3 against r0 here? http://codereview.chromium.org/1735007/show -- v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev
