display_channel_surface_id_unref resets the surface so
display_channel_surface_has_canvas will return false.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Store pointers to surface object in "surfaces" and allows to
have different surfaces with same ID in memory.
The surface was keep "busy" if there was pending drawing around.
Consider the following case:
1- receive drawing command
2- queue command on DCCs
3- destroy surface
4- send draw
Previously at point 4) you would have to use a surface from
"surfaces" which was destroyed, that is we would have to maintain
the pointer (and canvas) to the surface until reference counter
was 0.
However consider this case:
1- receive drawing command
2- queue command on DCCs
3- destroy surface
4- create surface
5- send draw
What would happen in point 4) ?
We could not change the surface as it will be used by point 5).
To avoid this the code attempts to wait the commands to release the
surface. However this can be an issue, you can't force the clients
to receive pending data if network is slow.
So this patch change this allowing to create surfaces while the old
version will still be used.
This is also more clean from the reference pointer prospective,
as the reference is increased for a specific surface.
Note that now instead of checking for canvas to not be NULL a
simple check for surface pointer is enough.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Mostly left on dcc-send.cpp.
Other minor too.
The change in BitmapData seems odd but the id for cached image
was not used so the only information left was the surface.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Instead of computing the value inside the function to then
compute also in the caller return it.
Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Acked-by: Victor Toso <victortoso@redhat.com>
As we use reference counting is more direct to use direct pointers.
Also this will allow to have a surface in a "released state"
reducing the complexity of code destroying a surface.
Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Similar to shared_ptr_counted but avoid atomic counter and
polymorphism.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Makes sure a message is sent with the proper message number.
Reduce mistakes associating types with wrong registered number.
Make easier to add new messages in the future.
Avoids having to type the message number every time, it's inferred
from the type.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Use a template to deduct type and avoid casts in every handler.
The reinterpret_cast seems strong but is safe as converting
a function with 2 typed pointers and a void return into a function
with 2 void pointers and a void return.
The unsafety left (not a regressions) is the association between
handler number and message type.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
When spice_server_destroy is called with pending clients (currently
not a big issue, usually the programs using SPICE server are
exiting then), some leaks can happen.
This is due to the fact that some dispatcher messages are queued
to handle some serialization but then they are neved executed as
the entire system is closed.
Close all connections and handle all main dispatcher messages
to remove these leaks.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
display_channel -> display.
In all the rest of the file display is used.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
This will allow to use not trivial objects inside the structure.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Since the conversion to a for range loop, there's no point to this
macro.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Acked-by: Frediano Ziglio <freddy77@gmail.com>
Since the conversion to a for range loop, there's no point to this
macro.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Acked-by: Frediano Ziglio <freddy77@gmail.com>
Using Unix sockets and no-Glibc C libraries (like Musl) getnameinfo
will fail causing SASL code to fail initialization.
Replicate Glibc behavior and report "localhost" as host and an
empty port string.
This fixes https://gitlab.freedesktop.org/spice/spice/-/issues/58.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
This removes:
In function ‘stream_channel_get_supported_codecs’,
inlined from ‘on_connect’ at ../server/stream-channel.cpp:364:60:
../server/stream-channel.cpp:326:31: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=]
326 | out_codecs[num++] = codec;
| ^
../server/stream-channel.cpp: In member function ‘on_connect’:
/usr/include/spice-1/spice/stream-device.h:209:13: note: destination object ‘codecs’ of size 0
209 | uint8_t codecs[0];
| ^
Reported by by Tomasz Kłoczko in
https://gitlab.freedesktop.org/spice/spice/-/issues/44
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
The build fails with slibtool while GNU libtool hide the issue
by silently ignoring -no-undefined.
ax_pthread.m4 is sourced from:
e68e8f6f62/m4/ax_pthread.m4
Downstream issue: https://bugs.gentoo.org/780027
Signed-off-by: orbea <orbea@riseup.net>
Acked-by: Frediano Ziglio <freddy77@gmail.com>
Found with readability-uppercase-literal-suffix
Avoids readability problems between lower case l and uppercase I.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Acked-by: Frediano Ziglio <freddy77@gmail.com>
Usually buffers to raw data are passed using void* pointers
to avoid casts and mark the buffer as raw.
Use them for read_safe and write_safe to avoid useless casts
in caller code.
As a minor convert a parameter to bool as changing the same
lines.
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Victor Toso <victortoso@redhat.com>
Found with google-readability-casting
https://google.github.io/styleguide/cppguide.html#Casting
Makes the operation clearer.
This commit uses const_cast where needed.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Acked-by: Frediano Ziglio <freddy77@gmail.com>
The leak is detected by Valgrind on Fedora 34as:
==19603== 400 bytes in 1 blocks are possibly lost in loss record 2,296 of 2,441
==19603== at 0x4845464: calloc (vg_replace_malloc.c:1117)
==19603== by 0x40135FB: _dl_allocate_tls (in /usr/lib64/ld-2.33.so)
==19603== by 0x57EB008: pthread_create@@GLIBC_2.2.5 (in /usr/lib64/libpthread-2.33.so)
==19603== by 0x53A1130: UnknownInlinedFun (gthread-posix.c:1323)
==19603== by 0x53A1130: g_thread_new_internal (gthread.c:931)
==19603== by 0x53C4953: g_thread_pool_start_thread.constprop.0 (gthreadpool.c:477)
==19603== by 0x53A2902: g_thread_pool_push (gthreadpool.c:691)
==19603== by 0x519AE11: g_task_run_in_thread_sync (gtask.c:1593)
==19603== by 0x80D8A74: ??? (in /usr/lib64/gio/modules/libgiolibproxy.so)
==19603== by 0x5181966: g_proxy_address_enumerator_next (gproxyaddressenumerator.c:176)
==19603== by 0x519281A: g_socket_client_connect (gsocketclient.c:1098)
Signed-off-by: Frediano Ziglio <freddy77@gmail.com>
Acked-by: Marc-André Lureau <marcandre.lureau@redhat.com>
