I decided against sorting, since I have no idea what implications that
would
have wrt const pool entries that are marked as not sharing_ok.
Yes I would avoid this for now - I'm not sure how it would interact with ldr
range for entries which are reordered.
https://codereview.chromium.org/1217673003/diff/20001/src/arm/assembler-arm.cc
File src/arm/assembler-arm.cc (right):
https://codereview.chromium.org/1217673003/diff/20001/src/arm/assembler-arm.cc#newcode3801
src/arm/assembler-arm.cc:3801: entry.set_merged_index(j);
I think this the wrong way round, you should be setting
...constants[j].set_merged_index[i]. Otherwise the delta for entry will
get larger (since the j is larger than i) which means it might be out of
ldr range. I'm not sure if this is actually a problem though (I can't
rememeber how the assembler checks when it should emit the pool). Also
if there was another duplicate later then wouldnt entry j will get
merged with that other dup while i is still duped with j?
https://codereview.chromium.org/1217673003/
--
--
v8-dev mailing list
[email protected]
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 [email protected].
For more options, visit https://groups.google.com/d/optout.
