In Jak 3, the default PC settings file would have a language of 255
because it runs before the first settings update. This would cause the
game to crash the second time it is started.
I also added this simple imgui window to see the names of streams in the
"SPU" memory, which has been useful for debugging
---------
Co-authored-by: water111 <awaterford1111445@gmail.com>
Accidentally incrementing `s` twice in the loop, so queued streams in
slot 1, and 3 were skipped by the overlord.
This would usually cause the audio to never start playing because the
game would wait for the audio to successfully get queued.
Co-authored-by: water111 <awaterford1111445@gmail.com>
Change jak 3 SPU DMA to run the interrupt handler "immediately" (or at
least before `DMA_SendToSPUAndSync` returns).
This fixes an issue where audio can hang during fast cutscene playback.
I'm hoping it fixes more issues with looping/stuck audio as well, but
this needs more testing.
I originally wanted to do it this way, but thought that it didn't work -
from Jak 2 it seemed like things broke if the DMA was too fast. But, at
least for Jak 3, everything seems to work like this. This will remove a
huge source of non-deterministic timing in audio stuff and hopefully
make things easier to debug. It also means that a large portion of
streaming audio code will never have to run - from the game's point of
view there's always the next SPU buffer available.
If this works well, I might revisit jak 2 as well.
Co-authored-by: water111 <awaterford1111445@gmail.com>
Fix compiler warnings, and a bug where the `snd_handle` of
`SoundBankInfo` was never set, leading to sound banks never unloading.
The game relies on unloading soundbanks to make sure certain sounds
don't play, like the blue gun 1 fire noise when using blue gun 2.
---------
Co-authored-by: water111 <awaterford1111445@gmail.com>
Similar to the workaround added in jak 3, if too many sounds are
playing, a sound will be removed. If there are multiple instances of the
same sound being played, those will be removed first. Within that, older
sounds are removed first.
It's not exactly the same as the instance limits of 989snd, but it seems
to work well. It's at least better than what we had before.
Co-authored-by: water111 <awaterford1111445@gmail.com>
A few minor fixes:
- Fix crash in overlord3 during final boss
https://github.com/open-goal/jak-project/issues/3605
- Update goal_src for `scene-actor.gc`, which was not updated after a
bug fix for decompiling skelgroups, making some cutscene actors
invisible due to using the wrong joint for culling checks.
- Stop using `-1` as an invalid value for texture id's in Merc.cpp. This
could sometimes cause Merc2.cpp to accidentally skip updating the OpenGL
texture. This fixes the bug where skull gems sometimes didn't have the
animated textures.
I found two issues with Jak 3 eyes. The first was simple - we were
missing a `-pc` texture upload in `texture.gc` for `pris2` textures,
which has eye textures for a few characters, like torn or damas.
The second was a little more annoying. Unlike jak 2 and jak 1, jak 3 can
dynamically assign eye slots when merc models are loaded. This involves
modifying eye data to tell the eye renderer where to render, and
modifying the merc model's adgif shaders to point to the correct eye
texture. The modification to the merc adgif shader is problematic since
our PC port of merc assumes this slot is constant.
My solution here was to bypass this whole slot system entirely for jak
3. I modified the GOAL eye renderer to tell the c++ eye renderer the
name of the merc-ctrl containing the eye. Then, the PC C++ Merc renderer
can just look up the merc-ctrl by name. To make this fit nicely in the
existing memory layout, I used a 64-bit fnv hash of the name. (which
honestly is how we should have handled a lot of other texture/model
names stuff...)
Unrelated fix to Overlord2 so it handles the case where file size
changes after the game starts, I had this in jak2/jak1 and forgot it for
jak 3.
This adds hfrag, but with a few remaining issues:
- The textures aren't animated. Instead, it just uses one texture.
- The texture filtering isn't as good as at it could be.
I also cleaned up a few issues with the background renderers:
- Cleaned up some stuff that is common to hfrag, tie, tfrag, shrub
- Moved time-of-day color packing stuff to FR3 creation, rather than at
level load. This appears to reduce the frame time spikes when a level is
first drawn by about 5 or 6 ms in big levels.
- Cleaned up the x86 specific stuff used in time of day. Now there's
only one place where we have an `ifdef`, rather than spreading it all
over the rendering code.
This includes all the collision stuff needed to spawn `target`,
decompiles the sparticle code and adds some of the PC hacks needed for
merc to run (it doesn't work quite right and looks bad, likely due to a
combination of code copied from Jak 2 and the time of day hacks).
There are a bunch of temporary hacks (see commits) in place to prevent
the game from crashing quite as much, but it is still extremely prone to
doing so due to lots of missing functions/potentially bad decomp.
---------
Co-authored-by: water <awaterford111445@gmail.com>
This is not a fix, but it temporarily works around the potential crash
discussed here https://github.com/open-goal/jak-project/issues/2988
As long as this doesn't impact other things, this feels like a good
interim fix, as having a broken cutscene is better than a fatal crash
and it might be some time before this is properly identified and fixed.
This updates `fmt` to the latest version and moves to just being a copy
of their repo to make updating easier (no editing their cmake / figuring
out which files to minimally include).
The motivation for this is now that we switched to C++ 20, there were a
ton of deprecated function usages that is going away in future compiler
versions. This gets rid of all those warnings.
It was narrowed down recently that a lot of people have issues with the
controller input because of Steam Input working as intended. Steam Input
can be configured to replicate controller inputs as keyboard inputs (for
example, pressing X on your controller presses Enter on the keyboard).
This results in the problem of "jumping pauses the game" and similar
issues. This is a consequence of the intended behaviour of the game
listening to all input sources at the same time.
Since the vast majority of players are using controllers over keyboards,
it makes sense to disable the keyboard input by default to solve this
problem. However that makes things awkward for users that want to use
the keyboard (how do they enable the setting). The solution is a new
imgui option in the settings menu:
**Known issue that I don't care about** -- in Jak 1's menu code, since
the flags are controlled by pointers to values instead of a lambda like
in jak 2, the menu won't update live with the imgui option. This has no
functional impact and I don't care enough to fix it.
I also made the pc-settings.gc file persist on first load if the file
wasn't found. Hopefully this helps diagnose the support issues related
to the black screen.
# Why not just ignore the keyboard inputs for a period of time?
This won't work, the keyboard is polled every frame. Therefore if you
hold down the X button on your controller, steam is continuously
signaling that `Enter` is held down on the keyboard.
Yes it would be possible to completely disable the keyboard while the
controller is being used, but this defeats the purpose of creating an
input system that allows multiple input sources at the same time.
With an explicit option, not only can the user decide the behaviour they
want (do they want the keyboard ignored or simultaneously listened to)
but we avoid breaking strange edge-cases in usage leading to never
ending complexity:
- ie. imagine steam input sends events to the mouse, well you can't
disable the mouse while using the keyboard because most times people are
using mouse and keyboard
- ie. a user that wants to hold a direction with the keyboard and press
buttons on the controller in tandem (something i frequently do while
TAS'ing, to move in a perfect straight line)
This sets up the C Kernel for Jak 3, and makes it possible to build and
load code built with `goalc --jak3`.
There's not too much interesting here, other than they switched to a
system where symbol IDs (unique numbers less than 2^14) are generated at
compile time, and those get included in the object file itself.
This is kind of annoying, since it means all tools that produce a GOAL
object file need to work together to assign unique symbol IDs. And since
the symbol IDs can't conflict, and are only a number between 0 and 2^14,
you can't just hash and hope for no collisions.
We work around this by ignoring the IDs and re-assigning our own. I
think this is very similar to what the C Kernel did on early builds of
Jak 3 which supported loading old format level files, which didn't have
the IDs included.
As far as I can tell, this shouldn't cause any problems. It defeats all
of their fancy tricks to save memory by not storing the symbol string,
but we don't care.
Currently only tracks enemy kills, and how they were killed. There is
currently no menu for this, but I've already added most of the text for
it. Also did a bunch of misc decompilation fixes and renamed some
methods.
Fixes#3277Fixes#3278
There was a single static path buffer being shared between multiple file
i/o threads. So sometimes you would end up using the wrong path for the
file, and getting size/data for the wrong file.
I think the original reason to have this buffer was just me being lazy
when we changed how project paths works a long time ago. It was a bad
idea in the first place.
Fixes skull gems using a low resolution texture.
Fixes issue where jak in cutscenes is dark after watching oracle-level-1
(and likely other bugs with texture animations getting stuck)
---------
Co-authored-by: ManDude <7569514+ManDude@users.noreply.github.com>
This fixes the crash reported in
https://github.com/open-goal/jak-project/issues/2833
There was a memory bug here for a long time where our array of `VagCmd`
in `iso_queue.cpp` was too small. This caused GetVagCommand to return
bogus pointers, and sound code would write over other parts of memory.
Avoids blocking other IOP threads on IO.
Idk if it'll really help anything, but at least theoretically it might
stop some pathologically slow IO case from blocking the VAG buffer
switching.
Normally, when they allocate a VagCmd, they do a bunch of stuff to clear
all the status bits and reset things
in particular the InitVAGCmd function does a lot
but for the stereo command, they do a lot less:
Which means that the new_stereo_command can just have random status bits
left over from whatever the last user had.
we seem to end up in a state where byte21 is set, and this causes
everything else to be wrong and off-by-one dma transfer. My guess is
that the original game avoided this bug due to lucky timing that I don't
understand.
I think the fix of just clearing byte21 is ok because there's no way
that the old value of the byte is useful after the command is
repurposed.
Also fixed an original game bug in `loader.gc` on a method that's called
quite often, though I have no clue what erroneous behavior it could have
even caused.
Saves 16 bits and lets us align the `color_index` field properly.
This shouldn't improve or decrease performance by any noticeable amount
except maybe in really low end systems.
water111