mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-08-26 23:29:22 +00:00
The new naming is (to me) much easier to understand. Here is a summary of the new state of the world: - '*Threshold' is the threshold for full unrolling. It is measured against the estimated unrolled cost as computed by getUserCost in TTI (or CodeMetrics, etc). We will exceed this threshold when unrolling loops where unrolling exposes a significant degree of simplification of the logic within the loop. - '*PercentDynamicCostSavedThreshold' is the percentage of the loop's estimated dynamic execution cost which needs to be saved by unrolling to apply a discount to the estimated unrolled cost. - '*DynamicCostSavingsDiscount' is the discount applied to the estimated unrolling cost when the dynamic savings are expected to be high. When actually analyzing the loop, we now produce both an estimated unrolled cost, and an estimated rolled cost. The rolled cost is notably a dynamic estimate based on our analysis of the expected execution of each iteration. While we're still working to build up the infrastructure for making these estimates, to me it is much more clear *how* to make them better when they have reasonably descriptive names. For example, we may want to apply estimated (from heuristics or profiles) dynamic execution weights to the *dynamic* cost estimates. If we start doing that, we would also need to track the static unrolled cost and the dynamic unrolled cost, as only the latter could reasonably be weighted by profile information. This patch is sadly not without functionality change for the new unroll analysis logic. Buried in the heuristic management were several things that surprised me. For example, we never subtracted the optimized instruction count off when comparing against the unroll heursistics! I don't know if this just got lost somewhere along the way or what, but with the new accounting of things, this is much easier to keep track of and we use the post-simplification cost estimate to compare to the thresholds, and use the dynamic cost reduction ratio to select whether we can exceed the baseline threshold. The old values of these flags also don't necessarily make sense. My impression is that none of these thresholds or discounts have been tuned yet, and so they're just arbitrary placehold numbers. As such, I've not bothered to adjust for the fact that this is now a discount and not a tow-tier threshold model. We need to tune all these values once the logic is ready to be enabled. Differential Revision: http://reviews.llvm.org/D9966 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@239164 91177308-0d34-0410-b5e6-96231b3b80d8
63 lines
3.6 KiB
LLVM
63 lines
3.6 KiB
LLVM
; In this test we check how heuristics for complete unrolling work. We have
|
|
; three knobs:
|
|
; 1) -unroll-threshold
|
|
; 3) -unroll-percent-dynamic-cost-saved-threshold and
|
|
; 2) -unroll-dynamic-cost-savings-discount
|
|
;
|
|
; They control loop-unrolling according to the following rules:
|
|
; * If size of unrolled loop exceeds the absoulte threshold, we don't unroll
|
|
; this loop under any circumstances.
|
|
; * If size of unrolled loop is below the '-unroll-threshold', then we'll
|
|
; consider this loop as a very small one, and completely unroll it.
|
|
; * If a loop size is between these two tresholds, we only do complete unroll
|
|
; it if estimated number of potentially optimized instructions is high (we
|
|
; specify the minimal percent of such instructions).
|
|
|
|
; In this particular test-case, complete unrolling will allow later
|
|
; optimizations to remove ~55% of the instructions, the loop body size is 9,
|
|
; and unrolled size is 65.
|
|
|
|
; RUN: opt < %s -S -loop-unroll -unroll-max-iteration-count-to-analyze=1000 -unroll-threshold=10 -unroll-percent-dynamic-cost-saved-threshold=20 -unroll-dynamic-cost-savings-discount=0 | FileCheck %s -check-prefix=TEST1
|
|
; RUN: opt < %s -S -loop-unroll -unroll-max-iteration-count-to-analyze=1000 -unroll-threshold=10 -unroll-percent-dynamic-cost-saved-threshold=20 -unroll-dynamic-cost-savings-discount=90 | FileCheck %s -check-prefix=TEST2
|
|
; RUN: opt < %s -S -loop-unroll -unroll-max-iteration-count-to-analyze=1000 -unroll-threshold=10 -unroll-percent-dynamic-cost-saved-threshold=80 -unroll-dynamic-cost-savings-discount=90 | FileCheck %s -check-prefix=TEST3
|
|
; RUN: opt < %s -S -loop-unroll -unroll-max-iteration-count-to-analyze=1000 -unroll-threshold=100 -unroll-percent-dynamic-cost-saved-threshold=80 -unroll-dynamic-cost-savings-discount=0 | FileCheck %s -check-prefix=TEST4
|
|
|
|
; If the absolute threshold is too low, or if we can't optimize away requested
|
|
; percent of instructions, we shouldn't unroll:
|
|
; TEST1: %array_const_idx = getelementptr inbounds [9 x i32], [9 x i32]* @known_constant, i64 0, i64 %iv
|
|
; TEST3: %array_const_idx = getelementptr inbounds [9 x i32], [9 x i32]* @known_constant, i64 0, i64 %iv
|
|
|
|
; Otherwise, we should:
|
|
; TEST2-NOT: %array_const_idx = getelementptr inbounds [9 x i32], [9 x i32]* @known_constant, i64 0, i64 %iv
|
|
|
|
; Also, we should unroll if the 'unroll-threshold' is big enough:
|
|
; TEST4-NOT: %array_const_idx = getelementptr inbounds [9 x i32], [9 x i32]* @known_constant, i64 0, i64 %iv
|
|
|
|
; And check that we don't crash when we're not allowed to do any analysis.
|
|
; RUN: opt < %s -loop-unroll -unroll-max-iteration-count-to-analyze=0 -disable-output
|
|
target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
|
|
|
|
@known_constant = internal unnamed_addr constant [9 x i32] [i32 0, i32 -1, i32 0, i32 -1, i32 5, i32 -1, i32 0, i32 -1, i32 0], align 16
|
|
|
|
define i32 @foo(i32* noalias nocapture readonly %src) {
|
|
entry:
|
|
br label %loop
|
|
|
|
loop: ; preds = %loop, %entry
|
|
%iv = phi i64 [ 0, %entry ], [ %inc, %loop ]
|
|
%r = phi i32 [ 0, %entry ], [ %add, %loop ]
|
|
%arrayidx = getelementptr inbounds i32, i32* %src, i64 %iv
|
|
%src_element = load i32, i32* %arrayidx, align 4
|
|
%array_const_idx = getelementptr inbounds [9 x i32], [9 x i32]* @known_constant, i64 0, i64 %iv
|
|
%const_array_element = load i32, i32* %array_const_idx, align 4
|
|
%mul = mul nsw i32 %src_element, %const_array_element
|
|
%add = add nsw i32 %mul, %r
|
|
%inc = add nuw nsw i64 %iv, 1
|
|
%exitcond86.i = icmp eq i64 %inc, 9
|
|
br i1 %exitcond86.i, label %loop.end, label %loop
|
|
|
|
loop.end: ; preds = %loop
|
|
%r.lcssa = phi i32 [ %r, %loop ]
|
|
ret i32 %r.lcssa
|
|
}
|