Displaying 3 results from an estimated 3 matches for "i154".
Did you mean:
154
2007 Feb 06
2
state of blur plugin
I forgot to give everyone an update on this.
The blur plugin is in a reasonable state now. It requires an OpenGL
implementation with an accelerated CopyTexSubImage2D path for usable
performance and that's likely not going to change sometime soon. One
major performance improvement can still be made, which allow us to avoid
the CopyTexSubImage2D path whenever the area behind a blurred region is
2019 Aug 23
3
Vectorization fails when dealing with a lot of for loops.
Hello, could you please have a look at this code posted on godbolt.org:
https://godbolt.org/z/O-O-Q7
The problem is that inside the compute function, only the first loop vectorizes while the rest copies of it don't. But if I remove any of the for loops, then the rest vectorize successfully. Could you please confirm that this is a bug, otherwise give me more insight on why the vectorization
2013 Feb 14
1
[LLVMdev] LiveIntervals analysis problem
...ec.ptr.6.i151 = getelementptr inbounds i16* %b, i32 7
%27 = load i16* %incdec.ptr.6.i151, align 2, !tbaa !5
%cmp5.7.i152 = icmp eq i16 %27, 0
br i1 %cmp5.7.i152, label %eisnan.exit158, label %if.then3
eisnan.exit158: ; preds = %for.inc.6.i153
%incdec.ptr.7.i154 = getelementptr inbounds i16* %b, i32 8
%28 = load i16* %incdec.ptr.7.i154, align 2, !tbaa !5
%not.cmp5.8.i155 = icmp eq i16 %28, 0
br i1 %not.cmp5.8.i155, label %if.end4, label %if.then3
if.then3: ; preds = %for.cond.preheader.i130, %eisnan.exit158, %...