I used the code in `wiiu/` to bootstrap my own WiiU homebrew app; this
PR reflects some changes I needed to make, that might be useful upstream.
1. Clean up filesystem initialization
Filesystem driver initialization was lumped in with filesystem mounting;
and that was a problem in my project, because I needed to be able to remount
the SD card on the fly. So, now it's split up.
I've added a callback object named "hooks" that can be used by consuming
applications to handle filesystem mounting and unmounting. If these hooks are
not provided, then the existing default behavior occurs.
2. Expand socket handling
- add `SO_NONBLOCK` flag for non-blocking socket I/O
- add normal errno defines like `EWOULDBLOCK` `EAGAIN`.
3. Remove RetroArch dependencies
- the exception handler protects usage of version_git with
`#ifdef HAVE_GIT_VERSION` but not the include, so I added that.
It also technically depends on version.h, but I'm not touching that.
It's easy enough to implement and I needed the same functionality. I'm
not sure what the best solution for that dependency is.
- missing_libc_functions.c included features/features_cpu.h which is
a libretro include. This appears to be a stale include though, because
everything compiles and works without it.
- an ifdef referencing the RA "WIIU" define, rather than the devkitpro
"__wiiu__" define
== DETAILS
The broadcast address is a standard part of TCP/IP that is used to
send messages to everyone on the subnet. This patch updates the
logging code to do the following:
1. Derive the broadcast address from the Wii U's own IP address
and subnet mask. These can all be obtained at runtime, which
means we can...
2. Remove the PC_DEVELOPMENT_IP_ADDRESS define from Wii U's
Makefile, because compiling in an IP is no longer needed.
3. Rewrite the net_listen script to listen for broadcast packets
and print them out with timestamps.
Since it's using the broadcast address, the only requirement is
that the PC be on the same network subnet as the Wii U.
Because of the low overhead of UDP, I've made logging on by
default. This will make it a ton easier to get useful bug
reports from users.
== DETAILS
The Wii U main entrypoints were embedded in the frontend driver,
which isn't a great place for them. Also, the `main()` method was
pretty long and monolithic. Now it's (much) less so.
Changes:
- Refactor out the main entrypoints into their own source files
(`wiiu/main.c` and `wiiu/main.h`)
- Optimize includes in both files, so only the minimum needed to
compile are included.
- The `main()` method is a lot easier to understand now. It's no longer
a confusing mess of ifdefs.
- There's a small amount of changes in the headers for future work, which
is switching kpad_driver to be callback-driven. The only change here is
to import the function that will be used, and define some data types.
Testing:
- Did local builds and confirmed build is successful
- Successfully loaded a core and switched among a few games
Add some timing functions so that we can know the time (woo-hoo?)
I use errno here, which worries me a bit. I wouldn't put it past
devkitPPC to have it as a null pointer or something.
Can confirm this makes XMB's clock work, much to @cucholix's delight;
I'm sure ;D
== DETAILS
We're trying to track down the source of crashes when switching cores.
To rule out the HID code, this commit does the following:
- Wraps the library imports in an ifdef
- Wraps the object files in conditionals in Makefile.wiiu
- In wiiu_joypad, calls into the hidpad driver are wrapped in ifdef
== TESTING
This didn't solve the "System memory error" crash I've been experiencing.
But, maybe it will impact the other flavors of crashes others are seeing.
== DETAILS
Starting to implement the I/O handling on the HID driver.
The old implementation basically had a never-ending HIDRead() callback
set up, so that callback N would start the read process for invocation
N+1.
We will take the same approach here; but now that the I/O thread is
happenning on its own core, we should be able to let it run full-
throttle without impacting emulator performance.
Of course, this hinges on the callback actually running on the same
core as the syscall was initiated on.
== TESTING
Can confirm that the read_loop_callback gets invoked on the same core
that the HIDRead() was invoked on.
Allow using the Gamepad's touch screen as a RETRO_DEVICE_POINTER.
Methodology could use some work, had to add an extra axis to
joypad in order to get the data transferred into the input driver.
Feel free to change this.
Needs to emulate RETRO_DEVICE_LIGHTGUN to really be useful.
Potential for Wiimote IR in future.
Partially addresses libretro/RetroArch#5294
