CI: deploy wasm #389
Merged
CI: deploy wasm #389
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Since wasm is deployed on Github Pages, thus it is reasonable to use CI to deploy automatically. The CI supports two event: merged pull request and workflow_dispatch. The former only run after merging and latter is used to deploy by maintainer manually. It also verifies whether any associated files have changed. If so, carry out the CI; if not, stop. Related: sysprog21#75
|
Please merge PR #388 first before merging this PR. |
|
Before merging this PR, please do the following to make the CI having permission:
|
|
After merging this PR, please run the CI manually for the first time using workflow_dispatch event. Then, we should see the wasm deployed on rv32emu-demo |
Please indicate the minimal requirements for the scopes of the access for personal tokens such as |
Please select the permissions |
vestata
pushed a commit
to vestata/rv32emu
that referenced
this pull request
Jan 24, 2025
CI: Deploy wasm
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 4, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later, sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host's SDL backend. Now, it is time to bridge all these components together. Leverage xterm.js as the frontend terminal in the web browser and bridge it with the backend VM shell through custom buffer management. This mechanism handles both standard ASCII input and escape sequences (such as arrow keys), providing a shell experience in the browser that closely resembles a real terminal. The SDL backend is also supported. After booting the guest Linux system, can run doom-riscv, quake, or smolnes to experience graphical applications. Can press Ctrl+C to exit SDL-based programs, or use their built-in exit commands. To reduce the size of the WASM file, the build is now separated into user and system targets. As a result, there are two HTML files: - user.html - system.html and two corresponding preload JavaScript: - user-pre.js and - system-pre.js Tested on latest Chrome, Firefox and Safari.
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 4, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later, sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host's SDL backend. Now, it is time to bridge all these components together. Leverage xterm.js as the frontend terminal in the web browser and bridge it with the backend VM shell through custom buffer management. This mechanism handles both standard ASCII input and escape sequences (such as arrow keys), providing a shell experience in the browser that closely resembles a real terminal. The SDL backend is also supported. After booting the guest Linux system, can run doom-riscv, quake, or smolnes to experience graphical applications. Can press Ctrl+C to exit SDL-based programs, or use their built-in exit commands. To reduce the size of the WASM file, the build is now separated into user and system targets. As a result, there are two HTML files: - user.html - system.html and two corresponding preload JavaScript: - user-pre.js - system-pre.js Tested on latest Chrome, Firefox and Safari.
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 5, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later, sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host's SDL backend. Now, it is time to bridge all these components together. Leverage xterm.js as the frontend terminal in the web browser and bridge it with the backend VM shell through custom buffer management. This mechanism handles both standard ASCII input and escape sequences (such as arrow keys), providing a shell experience in the browser that closely resembles a real terminal. The SDL backend is also supported. After booting the guest Linux system, can run doom-riscv, quake, or smolnes to experience graphical applications. Can press Ctrl+C to exit SDL-based programs, or use their built-in exit commands. To reduce the size of the WASM file, the build is now separated into user and system targets. As a result, there are two HTML files: - user.html - system.html and two corresponding preload JavaScript: - user-pre.js - system-pre.js Tested on latest Chrome, Firefox and Safari.
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 5, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism handles both standard ASCII input and escape sequences (such as arrow keys), providing a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed t 10000 ogether, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h header to store Emscripten-related variables and function declarations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 8, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h header to store Emscripten-related variables and function declarations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 13, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h,c to store Emscripten-related variables and function declarations and implementations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 13, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h,c to store Emscripten-related variables and function declarations and implementations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 13, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h,c to store Emscripten-related variables and function declarations and implementations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 13, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h,c to store Emscripten-related variables and function declarations and implementations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 13, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/emsc.h,c to store Emscripten-related variables and function declarations and implementations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
ChinYikMing
added a commit
to ChinYikMing/rv32emu
that referenced
this pull request
Jul 17, 2025
User-space emulation has been supported and deployable in WebAssembly since sysprog21#389, but system emulation was not yet available. With sysprog21#508, system emulation was introduced, and later sysprog21#551 added support for trap-and-emulate of guest Linux SDL syscalls, enabling offloading to the host’s SDL backend. This commit bridges these components together to enable full system emulation in the browser. xterm.js is leveraged as the frontend terminal, bridged with the backend VM shell through a custom buffer management mechanism. This mechanism provides a shell experience in the browser that closely resembles a real terminal. To handle terminal input without blocking the browser’s cooperative multitasking model, the original approach of mapping the read() system call to window.prompt() is avoided. Blocking behavior would freeze the browser’s event loop and degrade responsiveness. Instead, input from xterm.js is stored in a shared input buffer. The rv32emu periodically checks this buffer when handling external interrupts, and if input is available, it is read and consumed by the guest OS shell. The SDL graphic and sound backend are also supported. After booting the guest Linux system, users can run graphical applications such as doom-riscv, quake, or smolnes. These programs can be exited using Ctrl+C or their built-in exit funtionality. To reduce the size of the WebAssembly build and for the sake of the modularity, the project is now separated into user and system targets. As a result, two dedicated HTML files and corresponding preload JavaScript files are introduced: - user.html with user-pre.js - system.html with system-pre.js Navigation buttons are added to the index pages of both user and system demos, allowing users to switch easily between the two modes. Note that these navigation buttons are only functional when both user and system demos are deployed together, otherwise, the target pages may not be reachable. To improve usability, logic is implemented to disable and enable the "Run" button at appropriate times, preventing accidental re-execution when the process is already running. Additional improvements: - Ensure xterm.js uses \r\n instead of \n when logging to correctly move the cursor to the beginning of the line. - Add a new src/em_runtime.h,c to store Emscripten-related variables and function declarations and implementations for better management. This implementation has been tested on the latest versions of Chrome, Firefox, and Safari. To serve user space emulation index page: $ make start-web CC=emcc ENABLE_SDL=1 -j8 To serve system emulation index page: $ make start-web CC=emcc ENABLE_SYSTEM=1 ENABLE_SDL=1 INITRD_SIZE=32 -j8
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Since wasm is deployed on Github Pages, thus it is reasonable to use CI to deploy automatically.
The CI supports two events: merged pull request and workflow_dispatch. The former only run after merging and latter is used to deploy by maintainer manually. It also verifies whether any associated files have changed. If so, carry out the CI; if not, stop.
Related: #75