mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2025-01-31 20:51:52 +01:00
28625d59d9
Summary: Previously, in the NewPM pipeline, TailCallElim recalculates the DomTree when it modifies any instruction in the Function. For example, ``` CallInst *CI = dyn_cast<CallInst>(&I); ... CI->setTailCall(); Modified = true; ... if (!Modified || ...) return PreservedAnalyses::all(); ``` After applying this patch, the DomTree only recalculates if needed (plus an extra insertEdge() + an extra deleteEdge() call). When optimizing SQLite with `-passes="default<O3>"` pipeline of the newPM, the number of DomTree recalculation decreases by 6.2%, the number of nodes visited by DFS decreases by 2.9%. The time used by DomTree will decrease approximately 1%~2.5% after applying the patch. Statistics: ``` Before the patch: 23010 dom-tree-stats - Number of DomTree recalculations 489264 dom-tree-stats - Number of nodes visited by DFS -- DomTree After the patch: 21581 dom-tree-stats - Number of DomTree recalculations 475088 dom-tree-stats - Number of nodes visited by DFS -- DomTree ``` Reviewers: kuhar, dmgreen, brzycki, grosser, davide Reviewed By: kuhar, brzycki Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D49982 llvm-svn: 338954
83 lines
2.9 KiB
LLVM
83 lines
2.9 KiB
LLVM
; RUN: opt < %s -tailcallelim -verify-dom-info -S | grep call | count 4
|
|
; PR4323
|
|
|
|
; Several cases where tail call elimination should not move the load above the
|
|
; call, and thus can't eliminate the tail recursion.
|
|
|
|
|
|
@extern_weak_global = extern_weak global i32 ; <i32*> [#uses=1]
|
|
|
|
|
|
; This load can't be safely moved above the call because the load is from an
|
|
; extern_weak global and may trap, but the call may unwind before that happens.
|
|
define fastcc i32 @no_tailrecelim_1(i32* %a_arg, i32 %a_len_arg, i32 %start_arg) readonly {
|
|
entry:
|
|
%tmp2 = icmp sge i32 %start_arg, %a_len_arg ; <i1> [#uses=1]
|
|
br i1 %tmp2, label %if, label %else
|
|
|
|
if: ; preds = %entry
|
|
ret i32 37
|
|
|
|
else: ; preds = %entry
|
|
%tmp7 = add i32 %start_arg, 1 ; <i32> [#uses=1]
|
|
%tmp8 = call fastcc i32 @no_tailrecelim_1(i32* %a_arg, i32 %a_len_arg, i32 %tmp7) ; <i32> [#uses=1]
|
|
%tmp9 = load i32, i32* @extern_weak_global ; <i32> [#uses=1]
|
|
%tmp10 = add i32 %tmp9, %tmp8 ; <i32> [#uses=1]
|
|
ret i32 %tmp10
|
|
}
|
|
|
|
|
|
; This load can't be safely moved above the call because function may write to the pointer.
|
|
define fastcc i32 @no_tailrecelim_2(i32* %a_arg, i32 %a_len_arg, i32 %start_arg) nounwind {
|
|
entry:
|
|
%tmp2 = icmp sge i32 %start_arg, %a_len_arg ; <i1> [#uses=1]
|
|
br i1 %tmp2, label %if, label %else
|
|
|
|
if: ; preds = %entry
|
|
store i32 1, i32* %a_arg
|
|
ret i32 0
|
|
|
|
else: ; preds = %entry
|
|
%tmp7 = add i32 %start_arg, 1 ; <i32> [#uses=1]
|
|
%tmp8 = call fastcc i32 @no_tailrecelim_2(i32* %a_arg, i32 %a_len_arg, i32 %tmp7) ; <i32> [#uses=1]
|
|
%tmp9 = load i32, i32* %a_arg ; <i32> [#uses=1]
|
|
%tmp10 = add i32 %tmp9, %tmp8 ; <i32> [#uses=1]
|
|
ret i32 %tmp10
|
|
}
|
|
|
|
; This load can't be safely moved above the call because that would change the
|
|
; order in which the load volatiles are performed.
|
|
define fastcc i32 @no_tailrecelim_3(i32* %a_arg, i32 %a_len_arg, i32 %start_arg) nounwind {
|
|
entry:
|
|
%tmp2 = icmp sge i32 %start_arg, %a_len_arg ; <i1> [#uses=1]
|
|
br i1 %tmp2, label %if, label %else
|
|
|
|
if: ; preds = %entry
|
|
ret i32 0
|
|
|
|
else: ; preds = %entry
|
|
%tmp7 = add i32 %start_arg, 1 ; <i32> [#uses=1]
|
|
%tmp8 = call fastcc i32 @no_tailrecelim_3(i32* %a_arg, i32 %a_len_arg, i32 %tmp7) ; <i32> [#uses=1]
|
|
%tmp9 = load volatile i32, i32* %a_arg ; <i32> [#uses=1]
|
|
%tmp10 = add i32 %tmp9, %tmp8 ; <i32> [#uses=1]
|
|
ret i32 %tmp10
|
|
}
|
|
|
|
; This load can NOT be moved above the call because the a_arg is not
|
|
; sufficiently dereferenceable.
|
|
define fastcc i32 @no_tailrecelim_4(i32* dereferenceable(2) %a_arg, i32 %a_len_arg, i32 %start_arg) readonly {
|
|
entry:
|
|
%tmp2 = icmp sge i32 %start_arg, %a_len_arg ; <i1> [#uses=1]
|
|
br i1 %tmp2, label %if, label %else
|
|
|
|
if: ; preds = %entry
|
|
ret i32 0
|
|
|
|
else: ; preds = %entry
|
|
%tmp7 = add i32 %start_arg, 1 ; <i32> [#uses=1]
|
|
%tmp8 = call fastcc i32 @no_tailrecelim_4(i32* %a_arg, i32 %a_len_arg, i32 %tmp7) ; <i32> [#uses=1]
|
|
%tmp9 = load i32, i32* %a_arg ; <i32> [#uses=1]
|
|
%tmp10 = add i32 %tmp9, %tmp8 ; <i32> [#uses=1]
|
|
ret i32 %tmp10
|
|
}
|