Frame profiler
Go to file
2019-01-23 13:39:44 +01:00
capture Store failure reason strings in Worker. 2019-01-15 18:42:15 +01:00
client Enable tracing incline functions in callstacks. 2019-01-20 19:33:37 +01:00
common Make it possible to store multiple frames at single frame address. 2019-01-20 19:11:48 +01:00
doc Add histogram, compare screenshots to README. 2019-01-15 19:55:41 +01:00
extra X11 colors conversion program. 2018-07-04 18:26:57 +02:00
imgui Update imgui to 1.67. Also update imguicolortextedit. 2019-01-19 14:05:54 +01:00
imguicolortextedit Update imgui to 1.67. Also update imguicolortextedit. 2019-01-19 14:05:54 +01:00
libbacktrace Add modified libbacktrace. 2019-01-20 16:53:45 +01:00
manual Update manual. 2019-01-20 19:44:07 +01:00
nfd Workaround in nfd_win.cpp for MSVC problem in combaseapi.h. 2018-08-01 14:44:39 +02:00
profiler Update imgui to 1.67. Also update imguicolortextedit. 2019-01-19 14:05:54 +01:00
server Add zone child time getter. 2019-01-23 13:39:44 +01:00
test Function inlining test. 2019-01-20 16:55:09 +01:00
update Display dump file size change in the update utility. 2018-12-30 23:47:43 +01:00
.appveyor.yml Fix nuget. 2018-08-18 01:46:39 +02:00
.gitignore Use freetype to render fonts. 2018-08-17 21:40:15 +02:00
AUTHORS Update AUTHORS list. 2018-10-29 17:53:28 +01:00
FAQ.md Mention on-demand mode in FAQ. 2018-07-12 13:32:49 +02:00
LICENSE Update year in copyright notice. 2018-12-30 17:51:17 +01:00
NEWS Update NEWS. 2019-01-20 19:35:25 +01:00
README.md Add histogram, compare screenshots to README. 2019-01-15 19:55:41 +01:00
Tracy.hpp Allow forcing call stack capture. 2018-12-13 14:43:37 +01:00
TracyC.h Add ZoneText and ZoneName to the C API. 2019-01-16 02:10:21 +01:00
TracyClient.cpp Compile with libbacktrace on linux. 2019-01-20 16:55:33 +01:00
TracyClientDLL.cpp Fix builds with MingW. 2019-01-19 13:53:10 +02:00
TracyLua.hpp Explicitly cast size_t to uint32_t. 2018-08-22 16:30:37 +02:00
TracyOpenGL.hpp Allow forcing call stack capture. 2018-12-13 14:43:37 +01:00
TracyVulkan.hpp Hide internals behind TracyVkCtx typedef. 2019-01-14 12:40:54 +01:00

Tracy Profiler

Build status

Tracy is a real time, nanosecond resolution frame profiler that can be used for remote or embedded telemetry of your application. It can profile CPU (C, C++, Lua), GPU (OpenGL, Vulkan) and memory. It also can display locks held by threads and their interactions with each other.

Tracy requires compiler support for C++11, Thread Local Storage and a way to workaround static initialization order fiasco. There are no other requirements. The following platforms are confirmed to be working (this is not a complete list):

  • Windows (x86, x64)
  • Linux (x86, x64, ARM, ARM64)
  • Android (ARM, x86)
  • FreeBSD (x64)
  • Cygwin (x64)
  • WSL (x64)
  • OSX (x64)

The following compilers are supported:

  • MSVC
  • gcc
  • clang

Introduction to Tracy Profiler v0.2
New features in Tracy Profiler v0.3
New features in Tracy Profiler v0.4

A quick FAQ.
List of changes.

High-level overview

Tracy is split into client and server side. The client side collects events using a high-efficiency queue and awaits for an incoming connection. The server part connects to client and receives collected data from the client, which is then reconstructed into a viewable timeline. The transfer is performed using a TCP connection.

Performance impact

To check how much slowdown is introduced by using Tracy, I have profiled etcpak, which is the fastest ETC texture compression utility there is. I used an 8192×8192 test image as input data and instrumented everything down to the 4×4 pixel block compression function (that's 4 million blocks to compress). It should be noted that Tracy needs to calibrate its internal timers at each run. This introduces a delay of 115 ms (on my machine), which is negligible when doing lengthy profiling runs, but it skews the results of etcpak timing. The following times have this delay subtracted, to give focus on zone collection impact, which is the thing that really matters here.

Scenario Zones Clean run Profiling run Difference
Compression of an image to ETC1 format 4194568 0.94 s 1.003 s +0.063 s
Compression of an image to ETC2 format, with mip-maps 5592822 1.034 s 1.119 s +0.085 s

In both scenarios the per-zone time cost is at ~15 ns. This is in line with the measured 8 ns single event collection time (each zone has to report start and end event).

Usage instructions

The user manual for Tracy is available at the following address. It provides information about the integration process, required code markup and so on.

Features

Histogram of function execution times

Comparison of two profiling runs

Marking locks

Plotting data

Message log