It's just a tuning parameter (that nobody touched for aeons) for
DoGetImage(), inside dix/dispatch.c
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
It's assigned a const char* value and not writing into it,
so it should be const, too (compiler correctly warning about that)
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Instead of internally faking requests, factor out the actual logic
into separate function, which is getting everything it needs as
parameters, so no need to fiddle with request buffer anymore.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Instead of directly accessing the global screenInfo.screens[] array,
let everybody go through a little inline helper. This one also checks
for array bounds - if the screen doesn't exist, return NULL.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
a) move to doxygen-style docs, inside the header instead of code
b) mention that the function can be called many times with the
same parameters (already registered keys are silently tolerated)
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Call DeleteCallbackList(&pScreen->hookPostCreateResources) during
dixFreeScreen, because otherwise it will be heap-use-after-free during
DeleteCallbackManager call.
Signed-off-by: Tautvis <gtautvis@gmail.com>
In order to reduce complexity of wrapped core request handlers with PanoramiX,
split the ProcCreateWindow() function into two pieces: the upper half is the
usual (non-PanoramiX) handler, while the lower one is what's called by both
the usual handler, as well as the PanoramiX' one.
We're already passing in the request parameters as separate pointers, so
follow-up commits can easily change PanoramiX handler to not tweaking the
request buffer directly anymore. Another one is letting PanoramiXCreateWindow()
be called by ProcCreateWindow explicitly (when enabled), so we don't need to
wrap the core request proc vector anymore. Once that's done, the swapping can
also be moved into ProcCreateWindow().
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Document some non-trivial aspects and add warnings what not to
touch in order to not breaking ABI.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
These structs are only used inside dixutils, the actual callback handling
functions. Therefore no need to keep them in public header.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Don't rely on this file just being included indirectly by somebody else
just by accident.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Yet another step for getting rid of the unnecessarily complicated SProc*
machinery. Later, all those swap*() calls will be replaced by macros.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
In the doListFontsWithInfo function in dixfonts.c, when a font alias is
encountered (err == FontNameAlias), the code saves the current state
and allocates memory for c->savedName.
If the malloc(namelen + 1) call fails, c->savedName remains NULL,
but c->haveSaved is still set to TRUE. Later, when a font is
successfully resolved (err == Successful), the code uses c->savedName
without checking if it is NULL, so there is potential null ptr
dereference. XNFalloc will check result of malloc and stop
program execution if allocation was failed.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1842
Signed-off-by: default avatarMikhail Dmitrichenko <m.dmitrichenko222@gmail.com>
Part-of: <https://gitlab.freedesktop.org/xorg/xserver/-/merge_requests/2062>
Not used by any drivers, so no need to keep it in public SDK.
Since it's not used by any drivers, effectively no ABI change, so
can be safely done within ABI-25.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
iterating over screen list via lambda-esque macros calls like this
DIX_FOR_EACH_SCREEN({
do_something
});
withing the body, the iterator variables `walkScreenIdx` and `walkScreen`
are defined and can be directly used (read-only). the code inside the body
is running in a separate scope.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
CoreEnterLeaveEvent calls FixUpEventFromWindow with a pointer to a stack-allocated xEvent structure, which may get later casted into an xXIDeviceEvent*, with writes done by FixUpXI2DeviceEventFromWindow at an offset larger than sizeof(xEvent). This code-path is protected by xi2_get_type() but the following warning is generated by building with -O3 and LTO:
In function 'FixUpXI2DeviceEventFromWindow',
inlined from 'FixUpEventFromWindow' at ../dix/events.c:2716:13,
inlined from 'CoreEnterLeaveEvent' at ../dix/events.c:4679:5:
../dix/events.c:2628:48: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=]
2628 | ((xXIEnterEvent *) event)->same_screen =
| ^
../dix/events.c: In function 'CoreEnterLeaveEvent':
../dix/events.c:4652:12: note: at offset 48 into destination object 'event' of size 32
4652 | xEvent event = {
| ^
This PR suppresses this warning, by tracking the level of the event (ie., core event, XI or XI2) and ensuring that fix-up is performed only on XI2 events.
Signed-off-by: alex14fr <alex14fr@gmail.com>
Move the callbacks directly into DIX, since it's actually core infrastructure.
Also simplifying the whole machinery, by just using a simpel CallbackListPtr.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
This function is really doing nothing except of calling into the
actual request handler, so not needed at all.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Not used by any external drivers, so no need to keep them in public
SDK headers. Since they're never used by drivers, it's effectively
not an ABI change, so can safely be done within ABI-25.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
We already have a function for callback list deletion, so use this one
instead of doing it manually.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Only used by Xselinux extension, not by any drivers, so no need to
keep it exported.
Since it's never been used by drivers at all, it's effectively no ABI change,
so can safely be done within ABI-25.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
It's only used inside dixutil.c, nowhere else, especially not drivers,
so no need to keep it in public SDK. Safe for ABI-25.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Move the callbacks directly into DIX, since it's actually core infrastructure.
Also simplifying the whole machinery, by just using a simpel CallbackListPtr.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Move functions/macros dealing with request parsing or reply assembly/write
out of the big dix_priv.h into their own headers. This new header will also
get more of those function/macros soon (yet still in the pipeline).
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Move the callbacks directly into DIX, since it's actually core infrastructure.
Also simplifying the whole machinery, by just using a simpel CallbackListPtr.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Nobody outside tables.c is actually using it, so can be static.
The function used to be _X_EXPORT'ed, but there's no indication that any
external driver (not even proprietary NVidia) ever using it, so we can
get away w/o ABI version bump.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
These values are assigned to `unsigned short` fields, and we're only
getting in smaller numbers that fit well into it. So fixing compiler
warning on type size mismatch by using short args.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
Existing client-state hook isn't sufficient for this, and so easy to
be extended cleanly (*1). Adding a new callback is trivial and cheap,
so preferring this way, instead of trying to tweak the existing hook
for something it's never been designed for.
*1) see discussion here: https://github.com/X11Libre/xserver/pull/1077
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
These are used as counters / sizes, and we're getting lots of warnings
on signedness mismatches. So use size_t instead of (signed) int here.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
It's a counter and size value - we're getting lots of signedness warnings,
so use size_t instead of (signed) int.
Signed-off-by: Enrico Weigelt, metux IT consult <info@metux.net>
