[llvm] [MCA] Do not allocate space for DependenceEdge by default in DependencyGraphNode (NFC) (PR #125080)
Anton Sidorenko via llvm-commits
llvm-commits at lists.llvm.org
Thu Jan 30 08:13:26 PST 2025
https://github.com/asi-sc created https://github.com/llvm/llvm-project/pull/125080
For each instruction from the input assembly sequence, DependencyGraph has a dedicated node (DGNode). Outgoing edges (data, resource and memory dependencies) are tracked as SmallVector<..., 8> for each DGNode in the graph. However, it's unlikely that a usual input instruction will have approximately eight dependent instructions. Below is my statistics for several RISC-V input sequences:
```
Number of | Number of nodes with
edges | this # of edges
---------------------------------
0 | 8239447
1 | 464252
2 | 6164
3 | 6783
4 | 939
5 | 500
6 | 545
7 | 116
8 | 2
9 | 1
10 | 1
```
Approximately the same distribution is produced by llvm-mca lit tests for X86, AArch and RISC-V (even modified ones with extra dependencies added).
On a rather big input asm sequences, the use of SmallVector<..., 8> dramatically increases memory consumption without any need for it. In my case, replacing it with SmallVector<...,0> reduces memory usage by ~28% or ~1700% of input file size (2.2GB in absolute values).
There is no change in execution time, I verified it on mca lit-tests and on my big test (execution time is ~30s in both cases).
This change was made with the same intention as #124904 and optimizes I believe quite an unusual scenario. However, if there is no negative impact on other known scenarios, I'd like to have the change in llvm-project.
>From 77145440e95ec6e797cf2cfb2c10eb8dc777df69 Mon Sep 17 00:00:00 2001
From: Anton Sidorenko <anton.sidorenko at syntacore.com>
Date: Thu, 16 Jan 2025 15:22:36 +0300
Subject: [PATCH] [MCA] Do not allocate space for DependenceEdge by default in
DependencyGraphNode (NFC)
For each instruction from the input assembly sequence, DependencyGraph has a
dedicated node (DGNode). Outgoing edges (data, resource and memory dependencies)
are tracked as SmallVector<..., 8> for each DGNode in the graph. However, it's
rather unlikely that a usual input instruction will have approximately eight
dependent instructions. Below is my statistics for several RISC-V input
sequences:
Number of | Number of nodes with
edges | this # of edges
--------------------------------------------
0 | 8239447
1 | 464252
2 | 6164
3 | 6783
4 | 939
5 | 500
6 | 545
7 | 116
8 | 2
9 | 1
10 | 1
Approximately the same distribution is produced by llvm-mca lit tests (even
modified ones with extra dependencies added).
On a rather big input asm sequences, the use of SmallVector<..., 8> dramatically
increases memory consumption without any need for it. In my case, replacing it
with SmallVector<...,0> reduces memory usage by ~28% or ~1700% of input file size
(2.2GB in absolute values).
There is no change in execution time, I verified it on mca lit-tests and on my big
test (execution time is ~30s in both cases).
---
llvm/tools/llvm-mca/Views/BottleneckAnalysis.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/llvm/tools/llvm-mca/Views/BottleneckAnalysis.h b/llvm/tools/llvm-mca/Views/BottleneckAnalysis.h
index 529090cf543fc4..2621efb0413ae2 100644
--- a/llvm/tools/llvm-mca/Views/BottleneckAnalysis.h
+++ b/llvm/tools/llvm-mca/Views/BottleneckAnalysis.h
@@ -228,7 +228,7 @@ class DependencyGraph {
unsigned Depth;
DependencyEdge CriticalPredecessor;
- SmallVector<DependencyEdge, 8> OutgoingEdges;
+ SmallVector<DependencyEdge, 0> OutgoingEdges;
};
SmallVector<DGNode, 16> Nodes;
More information about the llvm-commits
mailing list