http://llvm.org/bugs/show_bug.cgi?id=22460
Bug ID: 22460
Summary: Canonicalizing load -> store to use integer type
confuses LICM
Product: libraries
Version: trunk
Hardware: PC
OS: Windows NT
Status: NEW
Severity: normal
Priority: P
Component: Scalar Optimizations
Assignee: [email protected]
Reporter: [email protected]
CC: [email protected], [email protected]
Classification: Unclassified
The canonicalization committed in r226781 creates a pattern that confuses LICM.
The minimal reproducer I managed to come up with is below:
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-pc-windows-msvc"
@in = internal unnamed_addr global i32* null, align 8
@out = internal unnamed_addr global i32* null, align 8
define i64 @bar(i32 %N) {
entry:
br label %do.body
do.body:
%i.0 = phi i32 [ 0, %entry ], [ %inc, %l2 ]
%total = phi i64 [ 0, %entry ], [ %next, %l2]
%c = icmp eq i32 %N, 6
br i1 %c, label %l1, label %l2
l1:
%v = load i32** @in, align 8
store i32* %v, i32** @out, align 8
br label %l2
l2:
%inval = load i32** @in, align 8
%int = ptrtoint i32* %inval to i64
%next = add i64 %total, %int
%inc = add nsw i32 %i.0, 1
%cmp = icmp slt i32 %inc, %N
br i1 %cmp, label %do.body, label %do.end
do.end:
ret i64 %total
}
Running opt -basicaa -gvn -licm produces this:
define i64 @bar(i32 %N) {
entry:
%c = icmp eq i32 %N, 6
%v = load i32** @in, align 8
%inval.pre = load i32** @in, align 8
br label %do.body
do.body: ; preds = %l2, %entry
%i.0 = phi i32 [ 0, %entry ], [ %inc, %l2 ]
%total = phi i64 [ 0, %entry ], [ %next, %l2 ]
br i1 %c, label %l1, label %do.body.l2_crit_edge
do.body.l2_crit_edge: ; preds = %do.body
br label %l2
l1: ; preds = %do.body
store i32* %v, i32** @out, align 8
br label %l2
l2: ; preds =
%do.body.l2_crit_edge, %l1
%inval = phi i32* [ %inval.pre, %do.body.l2_crit_edge ], [ %v, %l1 ]
%int = ptrtoint i32* %inval to i64
%next = add i64 %total, %int
%inc = add nsw i32 %i.0, 1
%cmp = icmp slt i32 %inc, %N
br i1 %cmp, label %do.body, label %do.end
do.end: ; preds = %l2
%total.lcssa = phi i64 [ %total, %l2 ]
ret i64 %total.lcssa
}
The load has been hoisted above the loop.
Adding instcombine before gvn gives us:
define i64 @bar(i32 %N) {
entry:
%c = icmp eq i32 %N, 6
%inval.pre = load i32** @in, align 8
br label %do.body
do.body: ; preds = %l2, %entry
%i.0 = phi i32 [ 0, %entry ], [ %inc, %l2 ]
%total = phi i64 [ 0, %entry ], [ %next, %l2 ]
br i1 %c, label %l1, label %do.body.l2_crit_edge
do.body.l2_crit_edge: ; preds = %do.body
br label %l2
l1: ; preds = %do.body
%v1 = load i64* bitcast (i32** @in to i64*), align 8
store i64 %v1, i64* bitcast (i32** @out to i64*), align 8
%0 = inttoptr i64 %v1 to i32*
br label %l2
l2: ; preds =
%do.body.l2_crit_edge, %l1
%inval = phi i32* [ %inval.pre, %do.body.l2_crit_edge ], [ %0, %l1 ]
%int = ptrtoint i32* %inval to i64
%next = add i64 %total, %int
%inc = add nsw i32 %i.0, 1
%cmp = icmp slt i32 %inc, %N
br i1 %cmp, label %do.body, label %do.end
do.end: ; preds = %l2
%total.lcssa = phi i64 [ %total, %l2 ]
ret i64 %total.lcssa
}
--
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
LLVMbugs mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
