This replaces all calls to SDL_free with a new macro, VVV_free, that
nulls the pointer afterwards. This mitigates any use-after-frees and
also completely eliminates double-frees. The same is done for any
function to free specific objects such as SDL_FreeSurface, with the
VVV_freefunc macro.
No exceptions for any of these calls, even if the pointer is discarded
or zeroed afterwards anyway. Better safe than sorry.
This is a macro rather than a function that takes in a
pointer-to-pointer because such a function would have type issues that
require casting and that's just not safe.
Even though SDL_free and other SDL functions already check for NULL, the
macro has a NULL check for other functions that don't. For example,
FAudioVoice_DestroyVoice does not check for NULL.
FILESYSTEM_freeMemory has been axed in favor of VVV_free because it
functionally does the same thing except for `unsigned char*` only.
The hilariously-named WIN32_LEAN_AND_MEAN slims down the number of
header files included in the already-massive `windows.h`. I know people
say Moore's law and precompiled headers and all that (well, we don't use
precompiled headers), but they kinda forgot about virtual machines, and
there's no reason not to define this and slim down the number of headers
anyway.
While fixing #885, I noticed that I had a bunch of
`special/stdin.vvvvvv` entries saved in my `levelstats.vvv`. At once I
knew that the dumb `special/stdin` hack that actually checks if the
filename passed is `special/stdin` was to blame.
STDIN playtesting was first merged, I knew in the back of my mind that
it was a bit of a dumb hack, but I didn't know it would cause
consequences like showing up in `levelstats.vvv`. For now, I'll just
have to patch it, but hopefully in the future I'll remove the dumb hack
entirely. Commenting both instances of the dumb hack with instructions
to grep for it should help maintainers out.
There are three different places where we call PHYSFS_openRead. This
commit makes sure all of them print a statement upon failure along with
the PhysFS reason for failure, and assigns the log level of each print
as so:
- FILESYSTEM_loadFileToMemory: Debug print (previously no
print existed in the first place), because some files (such as
font.txt) may or may not be needed, but if it isn't then no need to
print and worry the user. The game will error anyway if a critical
file like a graphics file is missing.
- FILESYSTEM_loadBinaryBlob: Debug print (previously info print),
because sometimes it's not needed, such as mmmmmm.vvv. I remember one
user being worried that the game printed "Unable to open file
mmmmmm.vvv" when it's not critical unlike vvvvvvmusic.vvv (and that
file is assumed to exist if data.zip exists anyways). Though maybe we
should move to loose-leaf files to save on memory usage (and so we
don't have to use special tools to modify a binary blob)...
- FILESYSTEM_loadZip: Error print. If we're calling this function, we
really do expect the zip to be loaded, and if it fails because we
can't open the file in the first place, then it would be good to know
why.
This fixes a limitation where the level filename had to be the exact
same name as the name of the zip, because the game used the name of the
level to identify the zip of which to load assets, and this also made it
impossible to use assets for more than one level in a zip.
Instead, we just look up where the level came from, so we can always
load its assets regardless of its filename.
Additionally, the zip structure checks can go away too, simplifying the
code further.
This WOULD be a huge breaking change, if it weren't for the fact that no
one uses them. Which is why I'm removing them, to simplify the code.
I asked on the VVVVVV Discord whether anyone used them or was even aware
of them and basically the answer was no. I go on Distractionware and no
one uses them. And why would they, when they'd have to distribute the
level .vvvvvv file separately? Better to just distribute everything in
one zip. And it's quite a bit obscure that you have to suffix the file
with .data.zip anyway.
This isn't necessary, but it does silence these annoying logs if you
pass an invalid argument or don't have data.zip:
[ERROR] Could not get window size: Invalid renderer
[WARN] Stats not loaded! Not writing unlock.vvv.
[ERROR] Could not get window size: Invalid renderer
[WARN] Settings not loaded! Not writing settings.vvv.
To do this, I've added FILESYSTEM_isInit().
Another cause of #870 is d0ffafe117, as a
bisect tells me. What that commit did is remove screenbuffer as a
pointer, since it's a statically-allocated object that _should_ always
exist, and it removed the `screenbuffer == NULL` guards in savestats()
and savesettings(). Unfortunately, those guards did something very
important - namely, they prevented writing to the save files when the
filesystem wasn't initialized. But that wasn't made clear, because it
seemed like the point of those guards was to prevent dereferencing NULL.
So instead, explicitly make it clear that
FILESYSTEM_saveTiXml2Document() needs to fail if the filesystem isn't
initialized. I've done this by adding an isInit bool to
FileSystemUtils.cpp.
While reviewing #272, I noticed that the PR was passing these two
arguments through a helper function, even though they really shouldn't
ever change. To obviate the need to pass these through, I'm making them
global variables.
pathSep is just a string literal from PhysFS, while basePath is a whole
complicated calculation from SDL and needs to be freed. It will be freed
upon filesystem deinit (as is done with PhysFS and the STDIN buffer).
Additionally the logic in FILESYSTEM_init is simplified by no longer
needing to keep a retval variable or use gotos to free basePath in
there.
This is mainly to make sure the game is definitely set to fullscreen in
Big Picture and on the Steam Deck, and to also remove windowed options
that wouldn't make sense if you're not on a desktop (toggling
fullscreen, resize to nearest). Those options would also be removed on
console and mobile too.
There's a bit of an annoying bug where if you launch the game in forced
fullscreen mode, but then exit and relaunch in normal mode, your game
will have fullscreen window sizes but it won't be fullscreen. This is
because forced fullscreen mode tries to preserve your non-forced
fullscreen setting, but due to the way window sizes are stored and
queried, it can't preserve the non-forced window size. This is a bit
difficult to work around, so I'm just putting in a FIXME here because we
can fix it later and I'd rather have a slightly buggy forced fullscreen
mode than not have one at all.
Closes#849.
These warnings are kinda spammy, and they make sense in principle.
vlog_error takes a format string, so passing it an arbitrary string
(even error messages from libraries) isn't a good idea.
This comment was referring to a now-deleted variable named mkdirResult
that was binary-"or"ed with all mkdir() results... except for the saves
directory. That variable was only used for save file migration, which is
now axed, so this comment is referring to nothing now.
I don't really know the answer to Ethan's question, but it doesn't
matter now.
Previously, loading STDIN used std::istreambuf_iterator and std::vector
and whatnot because... I guess it was less typing? But this isn't 1989;
we have the disk space to spare and we don't need to use fancy stuff
just to save on typing. It's not that hard to implement an array that
regrows to the nearest power of two every time.
Also, add a sync parameter to avoid calling syncfs too often.
Calling syncfs twice in a row is both inefficient and leads to errors
displaying twice. This allows us to bypass it when saving unlock.vvv as
part of savestatsandsettings.
In order to let callers provide their OWN callback functions through the
callback function WE provide to PhysFS, we casted the function pointer
to a void pointer.
Unfortunately, this is apparently undefined behavior... if your compiler
doesn't have an extension for it. And most compilers on most
architectures do. (In fact compilers on POSIX systems most certainly
have it due to dlsym() returning a void* which could actually be a
pointer to a function sometimes.)
But imo, it's better to be safe than sorry in this regard. Especially
when given GCC's approach to optimizing int + 100 > int (spoilers: they
remove it entirely! It's faster, but also broken!).
I've decided to wrap it in a struct. And as a nice side effect, if we
ever need more data to be passed through... well we already have this
struct.
Technically, it's also standards-compliant to cast a _pointer to_ a
function pointer to a void pointer. But that extra layer of pointer
indirection would get real confusing to conceptualize real fast (or at
least is more confusing than just putting it in a struct).
To match the option to nuke all main game save data, there is also now
an option to nuke all custom level save data separately (which is just
all custom level quicksaves, along with stars for level completion). It
has its own confirmation menu too. It does not delete any levels from
the levels folder.
This error case can happen, but if it does, non-console users get an
ERROR page with no further information. So use setLevelDirError if this
failure mode happens. And Menu::errorloadinglevel needs to be changed to
accomodate that.
Not sure why the original implementation decided to do things this way
instead of snprintf'ing a path to the .zip itself. Otherwise, if the
level is from data.zip, PHYSFS_getRealDir() will return the path of
data.zip, which then fails to mount for separate reasons.
If a zip file is improperly structured, a message will be displayed when
the player loads the level list.
This will only display the last-displayed improper zip, because there
only needs to be one displayed at a time. Also because doing anything
more would most likely require heap allocation, and I don't want to do
that.
Let's say you have a zip named LEVELNAME.zip, but the only .vvvvvv file
it contains is NOTLEVELNAME.vvvvvv. This zip would end up printing both
the 'LEVELNAME.vvvvvv is missing' and 'It has .vvvvvv file(s) other than
LEVELNAME.vvvvvv' messages, even though we already know there's
something wrong with the zip, and the 'other level files' message is
redundant, since in this case the problem here is simply just the
.vvvvvv file being named the wrong way.
The 'other level files' message is only intended to be printed when
LEVELNAME.vvvvvv *does* exist, but there's additional .vvvvvv files in
the zip on top of that, so don't print this message if LEVELNAME.vvvvvv
exists.
Two problems: the fRandom() range was from 0..36, but that's 37
characters, not 36. And the check to sort the lower 26 values into the
Latin alphabet used a 'lesser-than-or-equal-to 26' check, even though
that checks for the range of values of 0..26, which is 27 letters, even
though the alphabet only has 26 letters. So just drop the equals sign
from that check.
It was checking for .vvv-mnt-temp-XXXXXX/LEVELNAME.vvvvvv instead of
LEVELNAME.vvvvvv. When PhysFS enumerates the folder, it only gives us
LEVELNAME.vvvvvv, and not .vvv-mnt-temp-XXXXXX/LEVELNAME.vvvvvv.
- use fseeko and ftello like FreeBSD in tinyxml2
- use current directory as basePath if NULL (OpenBSD doesn't actually support this feature it is disabled via a patch in their ports)
If a level zip is named LEVELNAME.zip, the level file inside it must
also be named LEVELNAME.vvvvvv, else custom assets won't work.
This is because when we mount the zip file, we simply add
LEVELNAME.vvvvvv to the levels directory. Then whenever we load
LEVELNAME.vvvvvv, we look at the filename, remove the extension, and
look for the assets inside the zip of the same name, LEVELNAME.zip.
As a result, if someone were to make a level zip with assets but
mismatch the filename, the assets wouldn't load. Furthermore, if someone
were to add extra levels in the same zip, they wouldn't have any assets
load for them as well, which could be confusing.
To make things crystal-clear to the user, we now filter out any zips
that have incorrect structures like that, and print a message to the
terminal. Unfortunately nothing gets shown for non-terminal users, but
at least doing this and filtering out the zips is less confusing than
letting them through but with the issues mentioned above.
FILESYSTEM_mountAssets() has a big comment describing the magic numbers
needed to grab FILENAME from a string that looks like
"levels/FILENAME.vvvvvv".
Instead of doing that (and having to write a comment every time the
similar happens), I've written a macro (and helper function) instead
that does the same thing, but clearly conveys the intent.
I mean, just look at the diff. Using VVV_between() is much better than
having to read that comment, and the corresponding SDL_strlcpy().
This is so it can be reused without having to copy-paste.
generateBase36() is guaranateed to completely initialize and
null-terminate the buffer that is passed in.
Previously, if the game couldn't set the write dir to the base
directory, or couldn't make the base directory, or couldn't calculate
the base directory, it would probably dereference NULL or read from
uninitialized memory or murder your family or something. But now, I've
eliminated the potential Undefined Behavior from the code dealing with
the base path.
Previously, this function had a bug due to failing to account for array
decay. My solution was to just repeat the MAX_PATH again. But in
hindsight I realize that's bad because it hardcodes it, and introduces
the opportunity for an error where we update the size of the original
path but not the size in the function.
So instead, just pass the size through to the function.
FILESYSTEM_loadFileToMemory() dereferenced pointers without checking if
they were valid... I don't know of any cases where they could have been
NULL, but better safe than sorry.
So, the codebase was kind of undecided about who is responsible for
initializing the parameters passed to FILESYSTEM_loadFileToMemory() - is
it the caller? Is it FILESYSTEM_loadFileToMemory()? Sometimes callers
would initialize one variable but not the other, and it was always a
toss-up whether or not FILESYSTEM_loadFileToMemory() would end up
initializing everything in the end.
All of this is to say that the game dereferences an uninitialized
pointer if it can't load a sound effect. Which is bad. Now, I could
either fix that single case, or fix every case. Judging by the title of
this commit, you can infer that I decided to fix every case - fixing
every case means not just all cases that currently exist (which, as far
as I know, is only the sound effect one), but all cases that could exist
in the future.
So, FILESYSTEM_loadFileToMemory() is now guaranteed to initialize its
parameters even if the file fails to be loaded. This is better than
passing the responsibility to the caller anyway, because if the caller
initialized it, then that would be wasted work if the file succeeds
anyway because FILESYSTEM_loadFileToMemory() will overwrite it, and if
the file fails to load, well that's when the variables get initialized
anyway.
My next commit will involve using goto to jump to the end of a function
to initialize the variables to NULL, but that results in a compiler
error if we have initializations in the middle of the function. We might
as well put all declarations at the top of each block anyway, to help
the move to C, so I'm doing this now.
Since the length variable in the STDIN block now overshadows the length
variable in the outer block, I've renamed the length variable in the
block to stdin_length.
This seems to be a comment left by Ethan that he never got around to. So
I did it for him.
What I've done is made it so FileSystemUtils.cpp knows what a binary
blob is, and moved the binary blob loading code directly to
FileSystemUtils.cpp. To do this, I removed the private access modifier
from binaryBlob - I don't think we'll need it, and anyways when we move
to C we can't use it.
Along the way, I also cleaned up the style of the function a bit - the
null termination offset is no longer hardcoded, and the function no
longer mixes code and declarations together in the same block.
I also noticed that when printing all the filenames at the end, a single
invalid header would stop the whole loop instead of just being skipped
over... this seems to be a bug to me, so I've made it so invalid headers
just get skipped over instead of stopping the whole loop.
In FileSystemUtils.h, I used a forward declaration. In hindsight,
incomplete forward declarations should basically always be done in
header files if possible, otherwise this introduces the possibility of
transitive includes - if a file includes this header and it does a full
include, the file is silently able to use the full header, whereas if
it's a forward declaration, then the moment the file tries to use the
full header it fails, and then it's forced to include the full header
for itself. But uh, that's a code cleanup for later.
Misa
Dav999-v
leo60228
Ethan Lee
striker.sh