- User Since
- Sep 12 2012, 3:50 AM (365 w, 4 d)
Adding Simon and Krzysztof as the original authors of InfoByHwMode.
Fri, Sep 13
Thu, Sep 12
Thanks for the great feedback Jinsong! I've addressed your comments.
Mon, Sep 9
Fri, Sep 6
Thu, Sep 5
Wed, Sep 4
No-asserts build fixed in rL370894. Apologies for the breakage, I was holding git wrong.
On the other hand, I don't really want to spend much time thinking about how to make SelectionDAG better at this point.
Tue, Sep 3
Your problem sounds more like we need a better system for checking if a store/trunc store is legal
(for more context, this is in the context of https://reviews.llvm.org/rL370576 . It's entirely possible I picked a suboptimal target hook, but allowsMemoryAccess seemed correct to me.).
Aren't the targets intended to override allowsMisalignedMemoryAccesses? Does that not work for your usage?
Thanks for the review Thomas!
Landed in rL370705. Thanks Thomas!
Mon, Sep 2
Abandoning as this patch has been superseded.
Abandoning this revision; a more incremental approach will be landing soon.
Sat, Aug 31
Fri, Aug 30
Remove commented out code, add CMakeLists.txt entry.
Thu, Aug 29
Thanks! One nit.
I think I'm happy to stamp. I can't think of why this patch might be wrong, and it's easy to back out if someone finds a counterexample. LGTM.
Wed, Aug 28
I can't think of a counterexample for the other architectures I know about.
Fri, Aug 16
Aug 15 2019
Aug 14 2019
Aug 13 2019
Brendon, I've got a bunch of changes that I'd like to push based on this change. it's the gating factor for allowing predicated emission and allowing targets to adapt how they emit a schedule. I also have a different version of SMS to upstream (will be upstream very shortly!) that I'd love to reuse this emission logic for.
Aug 12 2019
Really sorry about that Jinsong. If you want the embarrassing details I scp'd the patchfile from one machine to another and then uploaded the stale patch from a different directory :(
Aug 9 2019
LGTM. Note for other reviewers; the canonical way to avoid this in LLVM is to use a stable_sort. The reason we can't do this is because this functor is used as the comparison for a std::priority_queue (heap) rather than a plain sort.
Aug 6 2019
Aug 2 2019
Aug 1 2019
Hi Brendon and Jinsong,
Jul 30 2019
LGTM. I had a look at changing this to use APInt instead, because that seems a little nicer, but this turns out to be nontrivial.
Jul 29 2019
Jul 12 2019
(excise some leftover debugging code).
Jul 3 2019
Not adding new predicates SGTM.
Most of this looks mechanical; I don't think the MVT changes should be contentious. However, the "is65536BitVector()" predicates are starting to get a bit unwieldy to read.
LGTM, thanks for the change Oliver!
Jul 2 2019
Jul 1 2019
LGTM with one simplification. Thanks!
May 22 2019
Thanks for the context Hans.
May 20 2019
I think I've convinced myself that this is correct. LGTM.
May 19 2019
May 10 2019
May 9 2019
Please clang-format. There should be a space between the + operator and its operands.
May 3 2019
Looking much better. I think the TTI hook could be described better.
May 2 2019
Undo approval; was looking at an incorrect diff.
This seems reasonable to me.
May 1 2019
Apr 30 2019
LGTM, thanks for making the changes!
I don't agree; if you're going to add this canonicalization, please do it in InstCombine first.
Apr 29 2019
LGTM with a couple of nits. Thanks!
Thanks for the changes, just two comments!
Apr 26 2019
Thanks for implementing this! It's really cool.
Yes, I also have the same concerns as Roman. If we don't canonicalize to fshr in instcombine, then we shouldn't canonicalize to fshr in simplifycfg. They're both canonicalization passes.
Thanks! Much easier to review this time.
Thanks for splitting these up! They're really much easier to read.
Apr 25 2019
Please reupload with full context as described in the developer's guide.
I am not the right person to review this change. I have never made changes in this directory and don't know enough about potential impact to LGTM.
Apr 23 2019
Seems reasonable to me.