Ferrum development journal · Part 16 of 19
Getting Wine's Real 32-Bit Startup to Run
The ordinary Wine loader path needed real process, thread, locale, and shutdown state. A command prompt that prints and relaunches became a useful application milestone.
The generic 32-bit goal is to run the application's real Windows components, rather than keep adding host-side replacements for whichever import the next title reaches. That meant moving into Wine's actual startup path and supplying the process and thread state it expects to find.
The investigation followed the loader through its initialization, heap and process-parameter work, and kernel32 lookup. As it progressed, missing activation-context state, thread setup, locale data, and other startup details became concrete failures to resolve. These are ordinary runtime requirements, even when the visible application is only a command prompt.
Shutdown needed its own work. A process can print the expected line and still fail to terminate correctly, leave a peer thread running, or break the next launch. The adopted checkpoint included owner-aware process termination and waiting for the relevant cleanup before final exit. That made clean shutdown part of the result rather than an afterthought.
By the October 11 records, Wine's real 32-bit command prompt had produced the expected output, exited cleanly, and launched again. The current adopted checkpoint also had the original regression groups and additional native64 and locale-related groups rechecked. Those are bounded results on a named driver, not a blanket claim about the rest of the Windows catalog.
I like this milestone because it's easy to understand and difficult to replace with a superficial test. An ordinary executable is using real runtime code, doing work, and leaving the system in a state where it can do it again. The remaining failures can then be investigated from a more realistic starting point.
This still left asynchronous I/O, broader application behavior, Steam, graphics journeys, and a complete packaged release open. It was a meaningful step toward the generic path, and it gave us a better baseline for the archive and file tests that followed.
Ferrum development journal
View the journalLive19 entries so far·Updated New (19 entries updated in the last 14 days)
- Part 1Bringing FEX to macOS: The First Windows Program
- Part 2From Return Values to Pixels and Sound
- Part 3Building More of the Windows Runtime
- Part 4Following a Game Through Its Opening Video
- Part 5Installers, Saves, Fonts, and the Parts Around the Game
- Part 6Game Progress and Measuring the Right Thing
- Part 7Building the 32-Bit Graphics Path
- Part 8Making Memory, Files, and Graphics State Agree
- Part 9DirectX 12, Shader Caches, and the Steam Startup Work
- Part 10Working on Presentation and Controller Input
- Part 11A Launcher and a Freeze That Wasn't the Presenter
- Part 12Portal Reaches a Command Stream, With the Frame Still Ahead
- Part 13Correcting the Ferrum Progress Report
- Part 14Recovering the Project and Resuming Real Tests
- Part 15Making Memory Protection and Lifetime Tests More Useful
- Part 16Getting Wine's Real 32-Bit Startup to Run
- Part 17Making the Launcher Own Its Storage and Processes
- Part 18A Real Archive Roundtrip and Better File Behavior
- Part 19Asynchronous I/O and the Bytes a Cancelled Write Keeps