blender
/
blender
134
398
Fork 580
Code Issues 6.3k Pull Requests 758 Projects 19 Wiki Activity

Geometry Nodes: Face Group Boundaries node speedup #115138

Merged
Hans Goudey merged 12 commits from mod_moder/blender:speedup_face_group_boundaries into main 2023-12-05 14:45:15 +01:00
Conversation 4 Commits 12 Files Changed 1 +56 -14
Iliya Katushenock commented 2023-11-19 14:13:12 +01:00
Member
Copy Link

Parallel version of function to compute boundary edges.
New multithreaded algorithm is used concept from #112415, without creating jagged array of face indices though.
Array of indices is store last face for each edge. In case new face have different Group ID, face index
will be come to invalid state and will be bypassed in all next checks.

Speed improvement for attached example file is listed in table:

Cube resolution Main PR
20x20x20 71920 ns 97400 ns
100x100x100 1.27 ms 1.17 ms
500x500x500 79.37 ms 23.16 ms
1000x1000x1000 520.31 ms 142.21 ms
Parallel version of function to compute boundary edges. New multithreaded algorithm is used concept from #112415, without creating jagged array of face indices though. Array of indices is store last face for each edge. In case new face have different Group ID, face index will be come to invalid state and will be bypassed in all next checks. Speed improvement for attached example file is listed in table: | Cube resolution | Main | PR | | -- | -- | -- | | 20x20x20 | `71920 ns` | `97400 ns` | | 100x100x100 | `1.27 ms` | `1.17 ms` | | 500x500x500 | `79.37 ms` | `23.16 ms` | | 1000x1000x1000 | `520.31 ms` | `142.21 ms` |
Face Group Boundaries speedup test.blend
926 KiB
👍 2 ❤️ 1 🚀 1
Iliya Katushenock added 1 commit 2023-11-19 14:13:25 +01:00
Hans Goudey commented 2023-11-19 15:02:22 +01:00
Member
Copy Link

Nice numbers. It also might be worth considering using the edge to corner topology map?

Nice numbers. It also might be worth considering using the edge to corner topology map?
Iliya Katushenock commented 2023-11-19 15:07:25 +01:00
Author
Member
Copy Link

I also thinking about that. I haven't test this. But this have some similar ideas: Here is used the same atomic list of last indices as in #112415. But here coping of indices in to segmented span is not necessary due to we can check is all elements have the same group id during list building.

I also thinking about that. I haven't test this. But this have some similar ideas: Here is used the same atomic list of last indices as in https://projects.blender.org/blender/blender/pulls/112415. But here coping of indices in to segmented span is not necessary due to we can check is all elements have the same group id during list building.
Iliya Katushenock added 1 commit 2023-11-19 18:54:44 +01:00
Iliya Katushenock added this to the Nodes & Physics project 2023-11-19 18:58:50 +01:00
Iliya Katushenock added the
Interest
Geometry Nodes
 label 2023-11-19 18:58:55 +01:00
Iliya Katushenock requested review from Hans Goudey 2023-11-19 18:59:02 +01:00
Iliya Katushenock added 2 commits 2023-11-20 13:48:26 +01:00
Iliya Katushenock added 1 commit 2023-11-20 16:01:18 +01:00
Hans Goudey reviewed 2023-11-21 13:58:14 +01:00
Hans Goudey left a comment
Member
Copy Link

I'll still probably test a version using the topology map, but this is looking reasonable now.

I'll still probably test a version using the topology map, but this is looking reasonable now.
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc Outdated
@ -57,3 +58,1 @@
if (edge_face_set[edge] != face_set[i]) {
/* This edge is connected to two faces on different face sets. */
boundary[edge] = true;
Array<int> previos_face(mesh.totedge, no_face_yet);
Hans Goudey commented 2023-11-21 13:56:53 +01:00
Member
Copy Link

Using Array<std::atomic<int>> probably makes this code a bit nicer, worth looking into?

Using `Array<std::atomic<int>>` probably makes this code a bit nicer, worth looking into?
Iliya Katushenock commented 2023-11-21 14:50:01 +01:00
Author
Member
Copy Link

Will check this now. Mainly i try to avoid std::atomic<T> due to that is not real T instance, so any T -> atomic<T> conversions have to be done as copy... but in in pr case, not sure if there is any problem. I found that size of std::atomic<T> can be the same as T, but that is not guaranteed, but this is not really important for that case too i hope....

Will check this now. Mainly i try to avoid `std::atomic<T>` due to that is not real `T` instance, so any `T` -> `atomic<T>` conversions have to be done as copy... but in in pr case, not sure if there is any problem. I found that size of `std::atomic<T>` can be the same as `T`, but that is not guaranteed, but this is not really important for that case too i hope....
mod_moder marked this conversation as resolved
Iliya Katushenock added 2 commits 2023-11-21 16:11:52 +01:00
7b5c19c6fd progress 
cleanup, renaming, use std::atomic_int
Hans Goudey reviewed 2023-11-21 17:50:27 +01:00
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc
@ -60,0 +62,4 @@
const static constexpr int no_face_yet = -1;
const static constexpr int no_more_face = -2;
threading::parallel_for(edge_states.index_range(), 4096, [&](const IndexRange range) {
Hans Goudey commented 2023-11-21 17:50:27 +01:00
Member
Copy Link

Does this work?

Array<std::atomic_int> edge_states(mesh.totedge, no_face_yet);

I guess maybe not. Ideally this would compile down to a memset. If it doesn't, maybe worth using the old version. A bit tricky...

Does this work? ``` Array<std::atomic_int> edge_states(mesh.totedge, no_face_yet); ``` I guess maybe not. Ideally this would compile down to a `memset`. If it doesn't, maybe worth using the old version. A bit tricky...
Iliya Katushenock commented 2023-11-21 17:56:59 +01:00
Author
Member
Copy Link

There is no copy-constructor...
Also about memset (just found) https://devblogs.microsoft.com/oldnewthing/20180328-00/?p=98365
My mesure show the same speed, so not sure if this have any other problems unlike this manual initialization...

There is no copy-constructor... Also about `memset` (just found) https://devblogs.microsoft.com/oldnewthing/20180328-00/?p=98365 My mesure show the same speed, so not sure if this have any other problems unlike this manual initialization...
Hans Goudey requested review from Jacques Lucke 2023-11-21 18:07:40 +01:00
Iliya Katushenock added 2 commits 2023-11-21 18:39:30 +01:00
Hans Goudey requested changes 2023-11-27 04:34:59 +01:00
Hans Goudey left a comment
Member
Copy Link

Tested this with topology maps like I mentioned:

    Array<int> edge_to_face_offsets;
    Array<int> edge_to_face_indices;
    const GroupedSpan<int> edge_to_face = bke::mesh::build_edge_to_face_map(mesh.faces(),
                                                                            mesh.corner_edges(),
                                                                            mesh.totedge,
                                                                            edge_to_face_offsets,
                                                                            edge_to_face_indices);
    Array<bool> boundary(mask.min_array_size());
    mask.foreach_index(GrainSize(1024), [&](const int edge) {
      const Span<int> faces = edge_to_face[edge];
      if (UNLIKELY(faces.is_empty())) {
        boundary[edge] = false;
        return;
      }

      const int first = face_set_span[faces.first()];
      boundary[edge] = std::any_of(faces.begin() + 1, faces.end(), [&](const int face) {
        return face_set_span[face] != first;
      });
    });

20x20x20

  • Before: Average: 30472 ns, Min: 20530 ns, Last: 28770 ns
  • Topology map: Average: 0.11 ms, Min: 71328 ns, Last: 73218 ns
  • Topology map + VArraySpan: Average: 0.10 ms, Min: 58359 ns, Last: 99817 ns
  • Iliya's code (this PR): Average: 44942 ns, Min: 17620 ns, Last: 53579 ns

100x100x100

  • Before: Average: 0.61 ms, Min: 0.57 ms, Last: 0.61 ms
  • Topology map: Average: 0.91 ms, Min: 0.84 ms, Last: 0.85 ms
  • Topology map + VArraySpan: Average: 0.90 ms, Min: 0.83 ms, Last: 0.91 ms
  • Iliya's code (this PR): Average: 0.30 ms, Min: 0.25 ms, Last: 0.27 ms

500x500x500

  • Before: Average: 19.66 ms, Min: 19.26 ms, Last: 19.36 ms
  • Topology map: Average: 24.24 ms, Min: 23.05 ms, Last: 23.93 ms
  • Topology map + VArraySpan: Average: 24.01 ms, Min: 23.18 ms, Last: 24.18 ms
  • Iliya's code (this PR): Average: 6.28 ms, Min: 5.99 ms, Last: 6.18 ms

1000x1000x1000

  • Before: Average: 196.50 ms, Min: 193.72 ms, Last: 195.56 ms
  • Topology map: Average: 116.46 ms, Min: 114.49 ms, Last: 115.51 ms
  • Topology map + VArraySpan: Average: 114.31 ms, Min: 112.78 ms, Last: 114.75 ms
  • Iliya's code (this PR): Average: 30.97 ms, Min: 29.48 ms, Last: 30.83 ms

Unless the topology maps are already cached (edge to face maps currently aren't), the approach in this PR is much faster. So I think it's best to go with this for now. One reason it's better is probably because we end up reading from the face group ID varray much less. Nice job :)

I left some style comments though.

Tested this with topology maps like I mentioned: ```cpp Array<int> edge_to_face_offsets; Array<int> edge_to_face_indices; const GroupedSpan<int> edge_to_face = bke::mesh::build_edge_to_face_map(mesh.faces(), mesh.corner_edges(), mesh.totedge, edge_to_face_offsets, edge_to_face_indices); Array<bool> boundary(mask.min_array_size()); mask.foreach_index(GrainSize(1024), [&](const int edge) { const Span<int> faces = edge_to_face[edge]; if (UNLIKELY(faces.is_empty())) { boundary[edge] = false; return; } const int first = face_set_span[faces.first()]; boundary[edge] = std::any_of(faces.begin() + 1, faces.end(), [&](const int face) { return face_set_span[face] != first; }); }); ``` **20x20x20** - Before: Average: 30472 ns, Min: 20530 ns, Last: 28770 ns - Topology map: Average: 0.11 ms, Min: 71328 ns, Last: 73218 ns - Topology map + VArraySpan: Average: 0.10 ms, Min: 58359 ns, Last: 99817 ns - Iliya's code (this PR): Average: 44942 ns, Min: 17620 ns, Last: 53579 ns **100x100x100** - Before: Average: 0.61 ms, Min: 0.57 ms, Last: 0.61 ms - Topology map: Average: 0.91 ms, Min: 0.84 ms, Last: 0.85 ms - Topology map + VArraySpan: Average: 0.90 ms, Min: 0.83 ms, Last: 0.91 ms - Iliya's code (this PR): Average: 0.30 ms, Min: 0.25 ms, Last: 0.27 ms **500x500x500** - Before: Average: 19.66 ms, Min: 19.26 ms, Last: 19.36 ms - Topology map: Average: 24.24 ms, Min: 23.05 ms, Last: 23.93 ms - Topology map + VArraySpan: Average: 24.01 ms, Min: 23.18 ms, Last: 24.18 ms - Iliya's code (this PR): Average: 6.28 ms, Min: 5.99 ms, Last: 6.18 ms **1000x1000x1000** - Before: Average: 196.50 ms, Min: 193.72 ms, Last: 195.56 ms - Topology map: Average: 116.46 ms, Min: 114.49 ms, Last: 115.51 ms - Topology map + VArraySpan: Average: 114.31 ms, Min: 112.78 ms, Last: 114.75 ms - Iliya's code (this PR): Average: 30.97 ms, Min: 29.48 ms, Last: 30.83 ms Unless the topology maps are already cached (edge to face maps currently aren't), the approach in this PR is much faster. So I think it's best to go with this for now. One reason it's better is probably because we end up reading from the face group ID varray much less. Nice job :) I left some style comments though.
👍 1
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc Outdated
@ -58,2 +55,2 @@
/* This edge is connected to two faces on different face sets. */
boundary[edge] = true;
Array<std::atomic_int> edge_states(mesh.totedge);
Hans Goudey commented 2023-11-27 04:29:39 +01:00
Member
Copy Link

Array<std::atomic<int>> unless you had a reason to do it this way.

`Array<std::atomic<int>>` unless you had a reason to do it this way.
mod_moder marked this conversation as resolved
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc Outdated
@ -60,0 +55,4 @@
Array<std::atomic_int> edge_states(mesh.totedge);
/* State is index of face or one of invalid values: */
const static constexpr int no_face_yet = -1;
Hans Goudey commented 2023-11-27 04:29:18 +01:00
Member
Copy Link

A simple constexpr int seems fine to me

A simple `constexpr int` seems fine to me
Iliya Katushenock commented 2023-11-27 05:03:34 +01:00
Author
Member
Copy Link

This requered to use this as case in switch statment.

This requered to use this as case in `switch` statment.
Hans Goudey commented 2023-11-27 05:13:34 +01:00
Member
Copy Link

Ah, both const and static? I thought const was redundant with constexpr anyway

Ah, both `const` and `static`? I thought `const` was redundant with `constexpr` anyway
Iliya Katushenock commented 2023-11-27 05:16:11 +01:00
Author
Member
Copy Link

Well, static at least, maybe const can be omited..

Well, `static` at least, maybe `const` can be omited..
👍 1
mod_moder marked this conversation as resolved
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc Outdated
@ -60,0 +56,4 @@
Array<std::atomic_int> edge_states(mesh.totedge);
/* State is index of face or one of invalid values: */
const static constexpr int no_face_yet = -1;
const static constexpr int no_more_face = -2;
Hans Goudey commented 2023-11-27 04:33:04 +01:00
Member
Copy Link

Maybe no_more_face -> is_boundary

Maybe `no_more_face` -> `is_boundary`
mod_moder marked this conversation as resolved
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc
@ -60,0 +64,4 @@
}
});
const GroupedSpan<int> face_edges(mesh.face_offsets(), mesh.corner_edges());
Hans Goudey commented 2023-11-27 04:34:36 +01:00
Member
Copy Link

I usually reserve face_edges for the variable of the edge indices of a specific face. Keeping corner_edges.slice(faces[face_i]) inline for consistency seems better to me. Though I don't have a strong opinion.

I usually reserve `face_edges` for the variable of the edge indices of a specific face. Keeping `corner_edges.slice(faces[face_i])` inline for consistency seems better to me. Though I don't have a strong opinion.
mod_moder marked this conversation as resolved
source/blender/nodes/geometry/nodes/node_geo_mesh_face_group_boundaries.cc Outdated
@ -60,0 +71,4 @@
for (const int edge_i : face_edges[face_i]) {
std::atomic_int &edge_state = edge_states[edge_i];
while (true) {
int edge_state_v = edge_state.load(std::memory_order_relaxed);
Hans Goudey commented 2023-11-27 04:33:24 +01:00
Member
Copy Link

What does _v mean?

What does `_v` mean?
Iliya Katushenock commented 2023-11-27 05:04:52 +01:00
Author
Member
Copy Link

edge_state_value

`edge_state_value`
mod_moder marked this conversation as resolved
Iliya Katushenock requested review from Hans Goudey 2023-11-27 13:44:05 +01:00
Iliya Katushenock added 2 commits 2023-11-27 13:44:21 +01:00
buildbot/vexp-code-patch-coordinator Build done. Details
a04334b02c progress
Hans Goudey approved these changes 2023-11-28 16:43:28 +01:00
Hans Goudey commented 2023-12-04 22:28:32 +01:00
Member
Copy Link

@blender-bot build

@blender-bot build
Hans Goudey added 1 commit 2023-12-04 23:20:54 +01:00
buildbot/vexp-code-patch-coordinator Build done. Details
fedda5bdeb Merge branch 'main' into speedup_face_group_boundaries
Hans Goudey commented 2023-12-04 23:31:54 +01:00
Member
Copy Link

@blender-bot build

@blender-bot build
Hans Goudey merged commit 5429e006c8 into main 2023-12-05 14:45:15 +01:00
Iliya Katushenock deleted branch speedup_face_group_boundaries 2023-12-05 14:48:51 +01:00
Sign in to join this conversation.
Reviewers
No reviewers
Jacques Lucke
Hans Goudey
Labels
Clear labels   
Interest
Alembic

  
Interest
Animation & Rigging

  
Interest
Asset Browser

  
Interest
Asset Browser Project Overview

  
Interest
Audio

  
Interest
Automated Testing

  
Interest
Blender Asset Bundle

  
Interest
BlendFile

  
Interest
Collada

  
Interest
Compatibility
This issue affects/is about backward or forward compatibility

  
Interest
Compositing

  
Interest
Core

  
Interest
Cycles

  
Interest
Dependency Graph

  
Interest
Development Management

  
Interest
EEVEE

  
Interest
EEVEE & Viewport

  
Interest
Freestyle

  
Interest
Geometry Nodes

  
Interest
Grease Pencil

  
Interest
ID Management

  
Interest
Images & Movies

  
Interest
Import Export

  
Interest
Line Art

  
Interest
Masking

  
Interest
Metal

  
Interest
Modeling

  
Interest
Modifiers

  
Interest
Motion Tracking

  
Interest
Nodes & Physics

  
Interest
OpenGL

  
Interest
Overlay

  
Interest
Overrides

  
Interest
Performance

  
Interest
Physics

  
Interest
Pipeline, Assets & IO

  
Interest
Platforms, Builds & Tests

  
Interest
Python API

  
Interest
Render & Cycles

  
Interest
Render Pipeline

  
Interest
Sculpt, Paint & Texture

  
Interest
Text Editor

  
Interest
Translations

  
Interest
Triaging

  
Interest
Undo

  
Interest
USD

  
Interest
User Interface

  
Interest
UV Editing

  
Interest
VFX & Video

  
Interest
Video Sequencer

  
Interest
Virtual Reality

  
Interest
Vulkan

  
Interest
Wayland
Issues relating to Wayland windowing support.

  
Interest
Workbench

  
Interest: X11

Issues relating to Xorg/X11 support.

  
Legacy
Blender 2.8 Project

  
Legacy
Milestone 1: Basic, Local Asset Browser

  
Legacy
OpenGL Error

  
Meta
Good First Issue

  
Meta
Papercut

  
Meta
Retrospective

  
Meta
Security
Issues relating to security: https://wiki.blender.org/wiki/Process/Vulnerability_Reports

  
Module
Animation & Rigging

  
Module
Core

  
Module
Development Management

  
Module
EEVEE & Viewport

  
Module
Grease Pencil

  
Module
Modeling

  
Module
Nodes & Physics

  
Module
Pipeline, Assets & IO

  
Module
Platforms, Builds & Tests

  
Module
Python API

  
Module
Render & Cycles

  
Module
Sculpt, Paint & Texture

  
Module
Triaging

  
Module
User Interface

  
Module
VFX & Video

  
Platform
FreeBSD

  
Platform
Linux

  
Platform
macOS

  
Platform
Windows

  
Priority
High

  
Priority
Low

  
Priority
Normal

  
Priority
Unbreak Now!

  
Status
Archived

  
Status
Confirmed

  
Status
Duplicate

  
Status
Needs Info from Developers

  
Status
Needs Information from User

  
Status
Needs Triage

  
Status
Resolved

  
Type
Bug

  
Type
Design

  
Type
Known Issue

  
Type
Patch

  
Type
Report

  
Type
To Do

No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Interest: X11
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Nodes & Physics
Assignees
Clear assignees
No Assignees
2 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#115138
No description provided.
Delete Branch "mod_moder/blender:speedup_face_group_boundaries"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?