Downloads & Free Reading Options - Results

Explicitly or implicitly, programming languages mirror domains. The best languages weave the concerns of a domain through a compatible computational model to offer programmers the best of both worlds. This statement naturally raises the question: What is the appropriate programming language for Web applications in the Ajax style? Our answer, Flapjax, is layered atop JavaScript. Flapjax demonstrates that the Web has a natural abstraction: event-driven reactivity. After explaining this we will build on it to examine connections to Web services, data binding, and access-control security. Depending on time, I can also cover some highlights of our work in this area on implementation techniques, principles for interfacing to legacy components, program transformations to improve performance, applications, and more. http://www.flapjax-lang.org/ ©2009 Microsoft Corporation. All rights reserved.

Presented at Code & Supply #golang Matthew Liegey will be presenting. The talk will include: ~ Quick intro to Go ~ Pros/Cons ~ Overview and examples of relevant std libs ~ Overview and examples of frameworks/libraries ~ Compared to: - Node - Ruby - Python

xvi, 480 p. : 24 cm

xvi, 480 p. : 24 cm

Get started with Dart and learn to program with this language suitable for high-performing, modern applications. You'll gain the basics and be ready to move to the next level: web and mobile apps. While you won't learn the specifics of programming web and mobile apps, you will have the foundation to take your Dart skills in that direction.This book starts with an introduction to the Dart IDE, after which you will take a look at the various components of the Dart programming language. You will look at types and variables, and get to know the significance of collections and arrays in Dart. Once you've familiarized yourself with the initial components of Dart, you will see how flow of control and looping can be achieved by using if, else, and conditional expressions. Moving on to functions and objects, you will dig deeper into the concepts of object oriented programming to gauge the importance of constructors in Dart. You will then discover more about inheritance and mixins, seeing how they contain methods for use by other classes. After gauging the importance of abstract classes and methods, you will learn more about anonymous functions such as lambdas and closures. You will then take a look at key data structures including lists and maps to aid you in organizing your information for storage and retrieval. After all this you’ll move on to managing exceptions arising from executing your program. Finally, Dart programming relies heavily on libraries to perform a variety of functions. You will cover some of these packages and libraries, including dart:core and dart:math, and also build a backend server with the help of the Dart core libraries.What You Will LearnUse variables and conditionals in DartWork with arrays and collections Apply flow control and loopingExplore data structures and their useHandle exceptionsUse Dart packages and libraries to build a backend serverWho This Book Is ForThose new to Dart programming who aim to get a quick introduction to its concepts and programming principles. Readers with no coding experience can also take advantage of this book....

356 pages : 28 cm +

Tri 2, 2015/2016

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Core PHP Programming, Second Edition is the first complete, practical guide to PHP 4 for experienced Web developers. Fully updated to reflect PHP 4.0's breakthrough features and performance, it covers every technique you need to build robust, fast Web applications. Leading PHP developer Leon Atkinson starts with a high-level overview of PHP: how it evolved, and how it appears to the developer. Next, Atkinson reviews the key building blocks of PHP scripts, including variables, operators, and expressions; conditional branches and loops. Review every PHP function, including I/O, data, and math functions; time, date, configuration, database, graphics, and network functions. Core PHP Programming, Second Edition also contains detailed coverage of effective program design and debugging. For all Web developers and system administrators.

Includes index

Q.1 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  List and explain important applications of internet in brief. (B)  Explain different approaches to Style Sheets. (C)  Write a short note on Internet Address. (D)  How hyperlinks are created in HTML? Explain with the help of an example. (E)  What is Search Engine? Explain its working. (F)  Write HTML Code to design following Web Page: Q.2 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  What is Image Map? Write the difference between Client-Side and Server-Side Image Mapping. (B)  Write the purpose of using: rowspan, colspan, cellspacing in Table tag. Give example. (C)  Write HTML Code for Embedding an Audio and Video in a Web Page. (D)  Explain the semantic tags of HTML5. (E)  Write HTML Code to design a form to enter name of the user, his/her password, and gender (radio button for male/female), hobbies (checkbox for reading/singing/sports), favorite color (list box red/green/blue) and submit and reset button. (F)  Explain < DIV > Tag with the help of suitable example. Q.3 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  Differentiate between Client Side and Server Side JavaScript. (B)  Write a JavaScript program to display all the prime numbers between 1 and 100. (C)  List and explain the methods of string object of JavaScript. (D)  Explain the following operator of JavaScript: (i)   new (ii)   delete (iii)   this (iv)   void (v)   ,(comma) (E)  Write the codes to design following web page and validate all the controls placed on the form using JavaScript as given: Name should not be blank, Check email-id is valid, Check Pin code is 6 digits long. (F)  Define events and event handlers. List various types of Mouse Event. Q.4 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  Write the difference between GET and POST methods in PHP. (B)  Explain different types of arrays available in PHP. (C)  Write PHP Program to demonstrate the use of different string functions. (D)  Explain Error Handling in PHP. (E)  Write a short note on variables in PHP. (F)  Write a PHP Program to create One Dimensional Array. Q.5 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  What is a cookie? How to store and retrieve the values in cookie in PHP? (B)  Explain any five PHP/MYSQL functions with example. (C)  Write a PHP program to send email with attachment. (D)  How to start and destroy a session and how to store a session variable in PHP? Explain. (E)  Write a short note on regular expressions in PHP. (F)  Write a PHP code to create a database "company" and to create a table "Employee" (emp_Id, emp_name, emp_dept, emp_salary).

Q.1 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  List and explain important applications of internet in brief. (B)  Explain different approaches to Style Sheets. (C)  Write a short note on Internet Address. (D)  How hyperlinks are created in HTML? Explain with the help of an example. (E)  What is Search Engine? Explain its working. (F)  Write HTML Code to design following Web Page: Q.2 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  What is Image Map? Write the difference between Client-Side and Server-Side Image Mapping. (B)  Write the purpose of using: rowspan, colspan, cellspacing in Table tag. Give example. (C)  Write HTML Code for Embedding an Audio and Video in a Web Page. (D)  Explain the semantic tags of HTML5. (E)  Write HTML Code to design a form to enter name of the user, his/her password, and gender (radio button for male/female), hobbies (checkbox for reading/singing/sports), favorite color (list box red/green/blue) and submit and reset button. (F)  Explain < DIV > Tag with the help of suitable example. Q.3 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  Differentiate between Client Side and Server Side JavaScript. (B)  Write a JavaScript program to display all the prime numbers between 1 and 100. (C)  List and explain the methods of string object of JavaScript. (D)  Explain the following operator of JavaScript: (i)   new (ii)   delete (iii)   this (iv)   void (v)   ,(comma) (E)  Write the codes to design following web page and validate all the controls placed on the form using JavaScript as given: Name should not be blank, Check email-id is valid, Check Pin code is 6 digits long. (F)  Define events and event handlers. List various types of Mouse Event. Q.4 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  Write the difference between GET and POST methods in PHP. (B)  Explain different types of arrays available in PHP. (C)  Write PHP Program to demonstrate the use of different string functions. (D)  Explain Error Handling in PHP. (E)  Write a short note on variables in PHP. (F)  Write a PHP Program to create One Dimensional Array. Q.5 ATTEMPT ANY THREE QUESTIONS: (15 MARKS) (A)  What is a cookie? How to store and retrieve the values in cookie in PHP? (B)  Explain any five PHP/MYSQL functions with example. (C)  Write a PHP program to send email with attachment. (D)  How to start and destroy a session and how to store a session variable in PHP? Explain. (E)  Write a short note on regular expressions in PHP. (F)  Write a PHP code to create a database "company" and to create a table "Employee" (emp_Id, emp_name, emp_dept, emp_salary).

Waitomo is an experimental programming language derived from Java. It aims to strengthen the guarantees of type safety by eliminating the need for casting without sacrificing the flexibility of the object-oriented programming style. The vital ingredients of the language are generics combined with a new approach to interfaces that includes a notion of self-types and union types. Waitomo interfaces are not types, but rather constraints which can be imposed on generics. This choice enables elegant solutions to standard programming problems. Besides an introduction to the language, the paper contains the full formalization of a core language with proofs of type soundness and decidability of subtyping. A prototype implementation of the language exists. ©2006 Microsoft Corporation. All rights reserved.

Waitomo is an experimental programming language derived from Java. It aims to strengthen the guarantees of type safety by eliminating the need for casting without sacrificing the flexibility of the object-oriented programming style. The vital ingredients of the language are generics combined with a new approach to interfaces that includes a notion of self-types and union types. Waitomo interfaces are not types, but rather constraints which can be imposed on generics. This choice enables elegant solutions to standard programming problems. Besides an introduction to the language, the paper contains the full formalization of a core language with proofs of type soundness and decidability of subtyping. A prototype implementation of the language exists. ©2006 Microsoft Corporation. All rights reserved.

Waitomo is an experimental programming language derived from Java. It aims to strengthen the guarantees of type safety by eliminating the need for casting without sacrificing the flexibility of the object-oriented programming style. The vital ingredients of the language are generics combined with a new approach to interfaces that includes a notion of self-types and union types. Waitomo interfaces are not types, but rather constraints which can be imposed on generics. This choice enables elegant solutions to standard programming problems. Besides an introduction to the language, the paper contains the full formalization of a core language with proofs of type soundness and decidability of subtyping. A prototype implementation of the language exists. ©2006 Microsoft Corporation. All rights reserved.

Waitomo is an experimental programming language derived from Java. It aims to strengthen the guarantees of type safety by eliminating the need for casting without sacrificing the flexibility of the object-oriented programming style. The vital ingredients of the language are generics combined with a new approach to interfaces that includes a notion of self-types and union types. Waitomo interfaces are not types, but rather constraints which can be imposed on generics. This choice enables elegant solutions to standard programming problems. Besides an introduction to the language, the paper contains the full formalization of a core language with proofs of type soundness and decidability of subtyping. A prototype implementation of the language exists. ©2006 Microsoft Corporation. All rights reserved.

The effects of knowledge of results (KR) and monetary reward on six hours of uninterrupted monitoring of a complex visual display were examined. Comparisons were made among groups receiving: no KR about response adequacy, KR, KR plus monetary reward or penalty determined by response adequacy, and KR plus reward in practice but not during the criterion session. In addition, comparison was made between the no-KR group and a similar one run by Webber and Adams (1), where a rest had been given after three hours of a six-hour monitoring period. All groups showed performance decrements of small magnitude. The manipulation of KR and reward failed to deter decrement; however, reward in addition to KR did enhance overall performance. KR alone did not facilitate performance, contrary to results from other studies. Training under KR plus reward did not enhance criterion performance when no KR or reward was provided. In support of previous research, man's monitoring capabilities over extended time periods seem adequate for modern systems.

Tri 1, 2017/2018

Tri 1, 2017/2018

Tri 1, 2017/2018

In addition to learning technical aspects of website creation, students must connect different knowledge, think critically, seek more information, and approach their work from various perspectives. These requirements, or specified competencies, can be developed through providing and receiving peer feedback, which is confirmed as an efficient learning strategy in various empirical studies on different subjects. The aim of this research is to investigate the impact of peer feedback on students’ achievements in the Web Programming Course. We conducted a quantitative study with two groups of students. Students from the experimental group participated in online giving and receiving peer feedback regarding students’ assignments, while students from the control group did not have this opportunity. For evaluating students’ work, we used rubrics. The results indicate that students, through analyzing their peers’ work, providing qualitative assessments, reflecting on their work in line with received comments, and making necessary corrections, create content that surpasses the quality of websites created by their peers who did not have the opportunity to participate in peer feedback activities. Since students extensively analyzed their work after receiving comments and suggestions, made specific corrections, and consequently improved their projects, organizing peer feedback activities in programming courses, especially web programming, should be given more attention. Enable Ginger Cannot connect to Ginger Check your internet connection or reload the browserDisable GingerRephraseRephrase with Ginger (Ctrl+Alt+E) Edit in Ginger Ginger is checking your text for mistakes... Disable Ginger in this text fieldDisable Ginger on this website×

In addition to learning technical aspects of website creation, students must connect different knowledge, think critically, seek more information, and approach their work from various perspectives. These requirements, or specified competencies, can be developed through providing and receiving peer feedback, which is confirmed as an efficient learning strategy in various empirical studies on different subjects. The aim of this research is to investigate the impact of peer feedback on students’ achievements in the Web Programming Course. We conducted a quantitative study with two groups of students. Students from the experimental group participated in online giving and receiving peer feedback regarding students’ assignments, while students from the control group did not have this opportunity. For evaluating students’ work, we used rubrics. The results indicate that students, through analyzing their peers’ work, providing qualitative assessments, reflecting on their work in line with received comments, and making necessary corrections, create content that surpasses the quality of websites created by their peers who did not have the opportunity to participate in peer feedback activities. Since students extensively analyzed their work after receiving comments and suggestions, made specific corrections, and consequently improved their projects, organizing peer feedback activities in programming courses, especially web programming, should be given more attention. Enable Ginger Cannot connect to Ginger Check your internet connection or reload the browserDisable GingerRephraseRephrase with Ginger (Ctrl+Alt+E) Edit in Ginger Ginger is checking your text for mistakes... Disable Ginger in this text fieldDisable Ginger on this website×

We describe a framework to support the implementation of web-based systems intended to manipulate data stored in relational databases. Since the conceptual model of a relational database is often specified as an entity-relationship (ER) model, we propose to use the ER model to generate a complete implementation in the declarative programming language Curry. This implementation contains operations to create and manipulate entities of the data model, supports authentication, authorization, session handling, and the composition of individual operations to user processes. Furthermore, the implementation ensures the consistency of the database w.r.t. the data dependencies specified in the ER model, i.e., updates initiated by the user cannot lead to an inconsistent state of the database. In order to generate a high-level declarative implementation that can be easily adapted to individual customer requirements, the framework exploits previous works on declarative database programming and web user interface construction in Curry.

We describe a framework to support the implementation of web-based systems intended to manipulate data stored in relational databases. Since the conceptual model of a relational database is often specified as an entity-relationship (ER) model, we propose to use the ER model to generate a complete implementation in the declarative programming language Curry. This implementation contains operations to create and manipulate entities of the data model, supports authentication, authorization, session handling, and the composition of individual operations to user processes. Furthermore, the implementation ensures the consistency of the database w.r.t. the data dependencies specified in the ER model, i.e., updates initiated by the user cannot lead to an inconsistent state of the database. In order to generate a high-level declarative implementation that can be easily adapted to individual customer requirements, the framework exploits previous works on declarative database programming and web user interface construction in Curry.

This is the first module for a course on the essentials of web authoring. See http://webprog.halavais.net

This is the first module for a course on the essentials of web authoring. See http://webprog.halavais.net

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

The Chromium Projects Home Chromium ChromiumOS Quick links Report bugs Discuss Other sites Chromium Blog Google Chrome Extensions Except as otherwise  noted , the content of this page is licensed under a  Creative Commons Attribution 2.5 license , and examples are licensed under the  BSD License . Privacy Edit this page For Developers  &gt;  How-Tos  &gt; Debugging Chromium on Windows First see  get the code  for checkout and build instructions. Getting started You can use Visual Studio's built-in debugger or  WinDBG  to debug Chromium. You don't need to use the IDE to build in order to use the debugger: autoninja is used to build Chromium and most developers invoke it from a command prompt, and then open the IDE for debugging as necessary. To start debugging an already-built executable with Visual Studio just launch Visual Studio (2019 or higher) and select File-&gt; Open-&gt; Project/Solution (Ctrl+Shift+O) and select the executable of interest. This will create a solution with that executable as the 'project'. You can then launch the debugger with F5 or F11 or from the Debug menu. If you right-click on the executable in Solution Explorer and select properties then you can edit things such as the executable path, command-line arguments, and working directory. You can add additional executables to the solution by using File-&gt; Add-&gt; Existing Project and selecting another already-built executable. You can select which one to debug by right-clicking on one of them in Solution Explorer and selecting Set as Startup Project. When your solution file is customized to your taste you can save it to a directory such as out\solutions. Saving it there helps ensure that relative paths to source files, printed from build commands, will correctly identify the source files. The Tools menu can be used to add commands to do things like invoke autoninja to build Chrome, compile the selected source file, or other things. Visual Studio 2017 is not recommended for debugging of Chromium - use a newer version for best performance and stability. symbol_level=2  is the default on Windows and gives full debugging information with types, locals, globals, function names, and source/line information.  symbol_level=1  creates smaller PDBs with just function names, and source/line information - source-level debugging is still supported (new from June 2019), but local variables and type information are missing.  symbol_level=0  gives extremely limited debugging abilities, mostly just viewing call stacks when Chromium crashes. Browsing source code If you use a solution file generated by gn ( gn gen --ide=vs ) then Intellisense may help you navigate the code. If this doesn't work or if you use a solution created as above then you may want to install  VsChromium  to help navigate the code, as well as using  https://source.chromium.org . Profiles It's a good idea to use a different Chrome profile for your debugging. If you are debugging Google Chrome branded builds, or use a Chromium build as your primary browser, the profiles can collide so you can't run both at once, and your stable browser might see profile versions from the future (Google Chrome and Chromium use different profile directories by default so won't collide). Use the command-line option: --user-data-dir =C:\tmp\my_debug_profile (replace the path as necessary) Using the IDE, go to the  Debugging  tab of the properties of the chrome project, and set the  Command Arguments. Chrome debug log Enable Chrome debug logging to a file by passing  --enable-logging --v=1  command-line flags at startup. Debug builds place the  chrome_debug.log  file in the  out\Debug  directory. Release builds place the file in the top level of the user data Chromium app directory, which is OS-version-dependent. For more information, see  logging  and  user data directory  details. Symbol server If you are debugging official Google Chrome release builds, use the symbol server: https://chromium-browser-symsrv.commondatastorage.googleapis.com In Visual Studio, this goes in  Tools &gt; Options  under  Debugging &gt; Symbols . You should set up a local cache in a empty directory on your computer. In windbg you can add this to your symbol server search path with the command below, where C:\symbols is a local cache directory: .sympath+ SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Alternately, You can set the _NT_SYMBOL_PATH environment variable to include both the Microsoft and Google symbol servers - VS, windbg, and other tools should both respect this environment variable: _NT_SYMBOL_PATH =SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols ;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com Note that symbol servers will let the debuggers download both the PE files (DLLs and EXEs) and the PDB files. Chrome often loads third party libraries and partial symbols for some of these are also public. For example: AMD : https://download.amd.com/dir/bin Nvidia : https://driver-symbols.nvidia.com/ Intel : https://software.intel.com/sites/downloads/symbols/ For example, for completeness, the following symbol server environment variable will resolve all of the above sources - but this is more than is normally needed: _NT_SYMBOL_PATH=SRV\*C:\symbols\*https://msdl.microsoft.com/download/symbols;SRV\*C:\symbols\*https://chromium-browser-symsrv.commondatastorage.googleapis.com;SRV\*C:\symbols\*https://download.amd.com/dir/bin;SRV\*C:\symbols\*https://driver-symbols.nvidia.com/;SRV\*C:\symbols\*https://software.intel.com/sites/downloads/symbols/ Source indexing You should set up source indexing in your debugger ( .srcfix  in windbg, Tools-&gt; Options-&gt; Debugging-&gt; General-&gt;  Enable source server support  in Visual Studio) so that the correct source files will automatically be downloaded based on information in the downloaded symbols. Additionally, you must have  python  in your  path  in order for the  command that fetches source files  to succeed; launching the debugger from the same environment as where you build Chromium is an easy way to ensure it's present. This is highly recommended when debugging released Google Chrome builds or looking at crash dumps. Having the correct version of the source files automatically show up saves significant time so you should definitely set this. Multi-process issues Chromium can be challenging to debug because of its  multi-process architecture . When you select  Run  in the debugger, only the main browser process will be debugged. The code that actually renders web pages (the Renderer) and the plugins will be in separate processes that's not (yet!) being debugged. The  ProcessExplorer  tool has a process tree view where you can see how these processes are related. You can also get the process IDs associated with each tab from the Chrome Task Manager (right-click on an empty area of the window title bar to open). Automatically attach to child processes There are two Visual Studio extensions that enable the debugger to automatically attach to all Chrome processes, so you can debug all of Chrome at once. Microsoft's  Child Process Debugging Power Tool  is a standalone extension for this, and  VsChromium  is another option that bundles many other additional features. In addition to installing one of these extensions, you  must  run Visual Studio as Administrator, or it will silently fail to attach to some of Chrome's child processes. Single-process mode One way to debug issues is to run Chromium in single-process mode. This will allow you to see the entire state of the program without extra work (although it will still have many threads). To use single-process mode, add the command-line flag --single-process This approach isn't perfect because some problems won't manifest themselves in this mode and some features don't work and worker threads are still spawned into new processes. Manually attaching to a child process You can attach to the running child processes with the debugger. Select  Tools &gt; Attach to Process  and click the  chrome.exe  process you want to attach to. Before attaching, make sure you have selected only Native code when attaching to the process This is done by clicking Select... in the Attach to Process window and only checking Native. If you forget this, it may attempt to attach in "WebKit" mode to debug JavaScript, and you'll get an error message "An operation is not legal in the current state." You can now debug the two processes as if they were one. When you are debugging multiple processes, open the  Debug &gt; Windows &gt; Processes  window to switch between them. Sometimes you are debugging something that only happens on startup, and want to see the child process as soon as it starts. Use: --renderer-startup-dialog --no-sandbox You have to disable the sandbox or the dialog box will be prohibited from showing. When the dialog appears, visit Tools &gt; Attach to Process and attach to the process showing the Renderer startup dialog. Now you're debugging in the renderer and can continue execution by pressing OK in the dialog. Startup dialogs also exist for other child process types:  --gpu-startup-dialog ,  --ppapi-startup-dialog ,  --utility-startup-dialog ,  --plugin-startup-dialog  (for NPAPI). For utilities, you can add a service type  --utility-startup-dialog=data_decoder.mojom.DataDecoderService . You can also try  the vs-chromium plug-in  to attach to the right processes. Semi-automatically attaching the debugger to child processes The following flags cause child processes to wait for 60 seconds in a busy loop for a debugger to attach to the process. Once either condition is true, it continues on; no exception is thrown. --wait-for-debugger-children [=filter] The filter, if provided, will fire only if it matches the  --type  parameter to the process. Values include  renderer ,  plugin  (for NPAPI),  ppapi ,  gpu-process , and  utility . When using this option, it may be helpful to limit the number of renderer processes spawned, using: --renderer-process-limit = 1 Image File Execution Options Using Image File Execution Options (IFEO) will not work because CreateProcess() returns the handle to the debugger process instead of the intended child process. There are also issues with the sandbox. Time travel debugging You can do  time travel debugging using WinDbg Preview  (must be installed from the Microsoft Store). This lets you execute a program forward and backwards. After capturing a trace, you can set breakpoints and step through code as normal, but also provides 'backwards' commands (g-, t-, p-) so that you can go back and forth through the execution. It is especially useful to set data breakpoints ( ba command ) and reverse continuing, so you can see when a certain variable was last changed to its current value. Chromium specifics: The type of injection the time travel tracer needs to perform is incompatible with the Chromium sandbox. In order to record a trace, you'll need to run with  --no-sandbox . Chromium cannot run elevated with Administrator privileges, so the "Launch executable (advance)" option won't work, you'll need to attach after the process has already launched via the checkbox in the bottom right. If you need to record startup-like things, you'll have to use --{browser,gpu,renderer,utility}-startup-dialog, then attach (and hope the relevant code hasn't executed before that point). JsDbg -- data structure visualization You can install  JsDbg as a plugin for WinDbg or Visual Studio . It interactively lets you look at data structures (such as the DOM tree, Accessibility tree, layout object tree, and others) in a web browser as you debug. See the  JsDbg site  for some screen shots and usage examples. This also works when examining memory dumps (though not minidumps), and also works together with time travel debugging. Visual Studio hints Debug visualizers Chrome's custom debug visualizers should be added to the pdb files and automatically picked up by Visual Studio. The definitions are in  //tools/win/DebugVisualizers  if you need to modify them (the BUILD.gn file there has additional instructions). Don't step into trivial functions The debugger can be configured to automatically not step into functions based on regular expression. Edit  default.natstepfilter  in the following directory: For Visual Studio 2015:  C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2015\Visualizers  (for the current user only) For Visual Studio 2017 Pro:  C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\Common7\Packages\Debugger\Visualizers  (for all users) or  %USERPROFILE%\My Documents\Visual Studio 2017\Visualizers  (for the current user only) Add regular expressions of functions to not step into. Remember to regex-escape  and  XML-escape them, e.g. &lt; for &lt; and \. for a literal dot. Example: &lt; Function &gt; &lt; Name &gt; operator new &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt; Function &gt; &lt; Name &gt; operator delete &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- Skip everything in std --&gt; &lt; Function &gt; &lt; Name &gt; std::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; &lt;!-- all methods on WebKit OwnPtr and variants, ... WTF::*Ptr&lt;*&gt;::* --&gt; &lt; Function &gt; &lt; Name &gt; WTF::.*Ptr&lt;.*&gt;::.* &lt;/ Name &gt; &lt; Action &gt; NoStepInto &lt;/ Action &gt; &lt;/ Function &gt; This file is read at start of a debugging session (F5), so you don't need to restart Visual Studio after changing it. More info:  Microsoft email thread V8 and Chromium V8 supports many command-line flags that are useful for debugging. V8 command-line flags can be set via the Chromium command-line flag --js-flags; for instance: chrome.exe --js-flags= "--trace_exception --heap_stats" Note that some V8 command-line flags exist only in the debug build of V8. For a list of all V8 flags try: chrome.exe --js-flags= "--help" Graphics debugging GPU Acceleration of rendering can be more easily debugged with tools. See: Graphics Debugging in Visual Studio 2013 Graphical debugging with NVIDIA NSight Debugging on another machine Sometimes it's useful to debug installation and execution on a machine other than your primary build box. To run the installer on said other machine, first build the mini_installer target on your main build machine (e.g., autoninja -C out\Debug mini_installer). Next, on the debug machine: Make the build machine's build volume available on the debug machine either by mounting it locally (e.g., Z:\) or by crafting a UNC path to it (e.g., \\builder\src) Open up a command prompt and change to a local disk Run src\tools\win\ copy-installer.bat  in the remote checkout by way of the mount (e.g., Z:\PATHTOCHECKOUT\src\...) or UNC path (e.g., \\builder\src\...). This will copy the installer, DLLs, and PDBs into your debug machine's C:\out or C:\build (depending on if you're rocking the component=shared_library build or not) Run  C:\out\Debug\mini_installer.exe  with the flags of your choice to install Chrome. This can take some time, especially on a slow machine. Watch the Task Manager and wait until mini_installer.exe exits before trying to launch Chrome (by way of the shortcut(s) created by the installer) For extra pleasure, add C:\out\Debug to your _NT_SYMBOL_PATH environment variable Consider reading the documentation at the top of copy-installer.bat to see how you can run it. It tries to be smart and copy the right things, but you may need to be explicit (e.g., "copy-installer.bat out Debug"). It is safe to re-run the script to copy only modified files (after a rebuild, for example). You can also use the zip action of the isolate scripts (tools\mb\mb.py) to package all the files for a target into a single zip file, for example: python tools\mb\mb.py zip out/Release base_unittests base_unittests. zip Finding all memory allocations It is possible to use Heap Snapshots to get call stacks on all outstanding allocations that use the OS heap. This works particularly well if heap snapshots are started as soon as the Chrome browser process is created, but before it starts running. Details can be found in  this batch file . However, with  PartitionAlloc Everywhere  most Chromium allocations no longer use the Windows heap so this will only find a subset of allocations, mostly from OS DLLs. Find memory leaks Note: as with Heap Snapshots the utility of UMDH is greatly reduced now because PartitionAlloc Everywhere has mostly replaced the Windows heap. The Windows heap manager has a really useful debug flag, where it can be asked to capture and store a stack trace with every allocation. The tool to scrape these stack traces out of processes is UMDH, which comes with  WinDbg . UMDH is great. It will capture a snapshot of the heap state as many times as you like, and it'll do it fairly quickly. You then run it again against either a single snapshot, or a pair of snapshots, at which time it'll symbolize the stack traces and aggregate usage up to unique stack traces. Turning on the user stack trace database for chrome.exe with gflags.exe makes it run unbearably slowly; however, turning on the user stack trace database on for the browser alone is just fine. While it's possible to turn on the user stack database with the "!gflag" debugging extension, it's too late to do this by the time the initial debugger breakpoint hits. The only reasonable way to do this is to Launch GFlags.exe, Enable the user stack trace database (per image below), Launch Chrome under the debugger. Set a breakpont when chrome.dll loads with "sxe ld chrome.dll". Step up, to allow Chrome.dll to initialize. Disable the stack trace database in GFlags.exe. Continue chrome, optionally detaching the debugger. GFlags.exe settings for user mode stack trace database. If you then ever suffer a browser memory leak, you can snarf a dump of the process with umdh - p :&lt;my browser pid&gt; &gt; chrome-browser-leak-umdh-dump.txt which can then typically be "trivially" analyzed to find the culprit. Miscellaneous Note that by default Application Verifier only works with non-official builds of Chromium. To use Application Verifier on official builds you need to add --disable-features=RendererCodeIntegrity to avoid sandbox crashes in renderer processes. See  crbug.com/1004989  for details. See also  this page . Application Verifier  is a free tool from Microsoft (available as part of the Windows SDK) that can be used to flush out programming errors. Starting with M68 Application Verifier can be enabled for chrome.exe without needing to disable the sandbox. After adding chrome.exe to the list of applications to be stressed you need to expand the list of Basics checks and disable the  Leak  checks. You may also need to disable  Handles  and  Locks  checks depending on your graphics driver and specific Chrome version, but the eventual goal is to have Chrome run with  Handles  and  Locks  checks enabled. When bugs are found Chrome will trigger a breakpoint so running all Chrome processes under a debugger is recommended. Chrome will run much more slowly because Application Verifier puts every heap allocation on a separate page. Note that with PartitionAlloc Everywhere most Chromium allocations don't actually go through the Windows heap and are therefore unaffected by Application Verifier. You can check the undocumented 'Cuzz' checkbox in Application Verifier to get the Windows thread scheduler to add some extra randomness in order to help expose race conditions in your code. To put a breakpoint on CreateFile(), add this break point: {,,kernel32.dll}_CreateFileW@28 {,,kernel32.dll} specifies the DLL (context operator). _ prefix means extern "C". @28 postfix means _stdcall with the stack pop at the end of the function. i.e. the number of arguments in BYTES. You can use  DebugView  from SysInternals or  sawbuck  to view LOG() messages that normally go to stderr on POSIX.

xvi, 480 p. : 24 cm

Welcome to the Web, from Khan Academy

xxiii, 378 p. : 24 cm

Immerse yourself in learning Python and introductory data analytics with this book’s project-based approach. Through the structure of a ten-week coding bootcamp course, you’ll learn key concepts and gain hands-on experience through weekly projects. Each chapter in this book is presented as a full week of topics, with Monday through Thursday covering specific concepts, leading up to Friday, when you are challenged to create a project using the skills learned throughout the week. Topics include Python basics and essential intermediate concepts such as list comprehension, generators and iterators, understanding algorithmic complexity, and data analysis with pandas. From beginning to end, this book builds up your abilities through exercises and challenges, culminating in your solid understanding of Python. Challenge yourself with the intensity of a coding bootcamp experience or learn at your own pace. With this hands-on learning approach, you will gain the skills you need to jumpstart a new career in programming or further your current one as a software developer. What You Will Learn

Includes index

Includes index

Web database programming uisng PHP/MariaDb-MySQL/Javascript etc.

Web database programming uisng PHP/MariaDb-MySQL/Javascript etc.

Web database programming uisng PHP/MariaDb-MySQL/Javascript etc.

1Intoxicated Ghost And Other Stories

By Arlo Bates

A charming collection of short stories, dealing with ghosts, magic, and other-worldly events that even the faint of heart will enjoy.<br> 1. The Intoxicated Ghost - a woman tries to outsmart a ghost to save the family from financial ruin.<br> 2. A Problem In Portraiture - can a man's portrait influence the man he becomes?<br> 3. Knitters In The Sun - will a father's curse keep two lovers apart?<br> 4. A Comedy In Crape - the death of the town playboy causes a dispute over who is entitled to be chief mourner<br> 5. A Meeting Of The Psychical Club - who is the hooded stranger, and are his powers real?<br> 6. Tim Calligan's Grave-Money - a poor man's final sacrifice<br> 7. Miss Gaylord and Jenny - who will ultimately be George's bride, Alice or her alter-ego?<br> 8. Dr. Polnitzski - a man in a no-win situation and his resolution<br> 9. In The Virginia Room - a Southern woman visits a Confederate War Museum on the anniversary of her husband's death and meets a Northerner. Can she forgive?<br> (Summary by ppcunningham)

2Astounding Stories 14, February 1931

By Harry Bates, Charles Willard Diffin, Anthony Gilmore, Desmond Winter Hall, F. V. W. Mason and Captain S. P. Meek

This issue includes "Werewolves of War" by D. W. Hall, "The Tentacles from Below" by Anthony Gilmore, "The Black Lamp" by Captain S. P. Meek, "Phalanxes of Atlans" by F. V. W. Mason, and continues with "The Pirate Planet" by Charles W. Diffin, - Summary by Bill Boerst

3Astounding Stories 19, July 1931

By Harry Bates, Ray Cummings, A. R. Holmes, H. Thompson Rich, Nat Schachner, Jack Williamson and Arthur L. Zagat

THE DOOM FROM PLANET 4 JACK WILLIAMSON A Ray of Fire, Green, Mysterious, Stabs Through the Night to Dan on His Ship. It Leads Him to an Island of Unearthly Peril.<br> THE HANDS OF ATEN H. G. WINTER Out of the Solid Ice Craig Hews Three Long-Frozen Egyptians and Is at Once Caught Up into Amazing Adventure. (A Complete Novelette.)<br> THE DIAMOND THUNDERBOLT H. THOMPSON RICH Locked in a Rocket and Fired into Space! Such Was the Fate which Awaited Young Stoddard at the End of the Diamond Trail!<br> THE SLAVE SHIP FROM SPACE A. R. HOLMES Three Kidnapped Earthlings Show Xantra of the Tillas How "Docile" Earth Slaves Can Be.<br> THE REVOLT OF THE MACHINES NAT SCHACHNER AND ARTHUR L. ZAGAT Something in the Many-Faceted Mind of the Master Machine Spurs It to Diabolical Revolt Against the Authority of Its Human Masters.<br> THE EXILE OF TIME RAY CUMMINGS Only Near the End of the World Does Fate Catch Up with Tugh, the Cripple Who Ran Amuck Through Time. (Conclusion.)<br> THE READERS' CORNER ALL OF US A Meeting Place for Readers of Astounding Stories.

4Astounding Stories 20, August 1931

By Harry Bates, Ralph Milne Farley, Captain S. P. Meek, R. F. Starzl and Harl Vincent

This issue contains THE DANGER FROM THE DEEP by Ralph Milne Farley Marooned on the Sea-Floor, His Hoisting Cable Cut, Young Abbot Is Left at the Mercy of the Man-Sharks<br> [BROOD OF THE DARK MOON PART 1 .We are not inlcuding this as this is the last full issue available. All 4 episodes are available in our catalogue].<br> IF THE SUN DIED by R.F Starzl Tens of Millenniums After the Death of the Sun There Comes a Young Man Who Dares to Open the Frozen Gate of Subterranea.<br> THE MIDGET FROM THE ISLAND by H.G. Winter Garth Howard, Prey to Half the Animals of the Forest, Fights Valiantly to Regain His Lost Five Feet of Size. A Complete Novelette.<br> "Jazzing up the Universe" A Science Fact<br> THE MOON WEED by Harl Vincent Unwittingly the Traitor of the Earth, Van Pits Himself Against the Inexorably Tightening Web of Plant-Beasts He Has Released from the Moon.<br> THE PORT OF MISSING PLANES by Captain S.P. Meek In the Underground Caverns of the Selom, Dr. Bird Once Again Locks Wills with the Subversive Genius, Saranoff.<br> "A Classification of the Universe " A Science Fact<br> THE READERS' CORNER ALL OF US A Meeting Place for Readers of Astounding Stories.<br> " A living disembodied heart" - Summary by Annise

5On The Tree Top

By Clara Doty Bates

A collection of Nursery Rhymes retold by the author and others. - Summary by David Lawrence <br/> <br/>*Additional Proof-listening by Christine Lehman.

LibriVox volunteers bring you 19 recordings of Blood Road by Katharine Lee Bates.<br> This was the New Year's Weekly Poetry project for December 30. 2018. <br> ------<br> Katharine Lee Bates was an American writer, poet, professor, and social activist. Although she was a renowned author and professor during her lifetime, today she is primarily remembered as the author of the words to the anthem "America the Beautiful". For 25 years, she lived with her long-time friend and companion, Katharine Coman. This poem taken from 'America the beautiful and other poems' 1911. - Summary by Wikipedia

7Henry More Smith: The Mysterious Stranger

By Walter Bates

Sometime in the month of July, 1812, nearly a hundred years ago now, a well dressed, smooth spoken man, less than thirty years of age, made his appearance at Windsor, Nova Scotia. The story as told in subsequent pages by Sheriff Bates is unique in criminal annals and is worthy of careful perusal. - Summary Adapted from the Preface

LibriVox volunteers bring you 17 recordings of On Christmas Eve by Katharine Lee Bates.<br> This was the Weekly Poetry project for December 19, 2021.<br> -------<br> Best known as the author of "America the Beautiful", American professor and poet Katharine Lee Bates also wrote many books and articles on social reform. In this poem, she strips away some of the outer layers of the Christmas story to peer into the core of the event. - Summary by TriciaG

LibriVox volunteers bring you 13 recordings of The Empty Room by Katharine Lee Bates. This was the Fortnightly Poetry project for December 19, 2021.<br> -------<br> Best known as the author of "America the Beautiful", American professor and poet Katharine Lee Bates also wrote many books and articles on social reform. - Summary by TriciaG

10Fish Story

By Clara Doty Bates

LibriVox volunteers bring you 12 recordings of A Fish Story by Clara Doty Bates.<br> This was the Fortnightly Poetry project for May 8, 2022. <br> ------<br> Clara Doty Bates was a 19th-century American author who published a number of volumes of poetry and juvenile literature. Many of these works were illustrated, the designs being furnished by her sister. This Fortnightly offers a silly fish story. - Summary by Wikipedia

LibriVox volunteers bring you 12 recordings of America the Beautiful by Katharine Lee Bates.<br> This was the Weekly Poetry project for July 3, 2022. <br> ------<br> In honor of The United States' 246th birthday on July 4th. Bates wrote the words as a poem originally entitled "Pikes Peak". It was first published in the Fourth of July 1895 edition of the church periodical, The Congregationalist. It was at that time that the poem was first entitled "America". - Summary by TriciaG & Wikipedia

12Seen and Unseen

By E. Katharine Bates

Novelist. Psychic investigator. Globe trotter. The multi-plane traveler, E. Katharine Bates, shares her phenomenal adventures on the cusp of the Twentieth Century in the autobiographical “Seen and Unseen”. Journey with this worthy companion! Enjoy the otherworldly experience! - Summary by Brian Fullen