- xrdp_listen.c is refactored so we can create the
listening socket(s) before dropping privileges.
- The code which reads startup params from xrdp.ini
is moved from xrdp_listen.c to xrdp.c, so it
is only called once if we test the listen before
starting the daemon.
If ./configure is used with devel logging, but without --enable-pixman,
the stub pixman development files are used.
However, in this configuration, the pixman_region_selfcheck() function
is declared, but not defined.
This is a regression introduced in 7e58209b19
- xrdp is not now built with XRDP_IBUS to allow other input
methods to be more easily supported.
- chansrv is only aked to start an input method if the client
supports it.
- chansrv sends a status report back to xrdp when asked to start
and input method.
- ./configure without --enable-ibus now works.
When xrdp is built with IPv6 support it will only fall back to IPv4 if
IPv6 is not supported (EAFNOSUPPORT). However, if the system is IPv6
capable but doesn't have an IPv6 address set (at least inside a FreeBSD
jail) EPROTONOSUPPORT is returned from socket().
The '-lrt' added to the Makefile for the common library appears
to be unnecessary.
- On modern Linuxes, this library has been merged with libc, and the
supplied library is empty.
- On older ones (e.g. Devuan 4), the library contains routines we
do not use in xrdp (although we use 'shm_open()' in xorgxrdp).
- On FreeBSD 14 the library contains only mq_* and timer_* routines
which, again, are not required.
From [MS-RCPBCGR] 3.3.5.12.1:-
> ...The contents of this PDU SHOULD NOT be compressed.
>
> This PDU MUST NOT be sent to a client that has not indicated support for
> it by setting the RNS_UD_CS_SUPPORT_MONITOR_LAYOUT_PDU flag (0x0040)
> in the earlyCapabilityFlags field of the Client Core Data (section
> 2.2.1.3.2).
Also, 2.2.12.1 specifies the source channel must be zero.
In testing, a compressed monitor layout PDU causes mstsc.exe
to exit with a protocol error.
* Store EGFX state before entering resize state machine
At present the EGFX state is destroyed by states WMRZ_EGFX_DELETE_SURFACE
through WRMZ_EGFX_DELETE. This means that at WMRZ_EGFX_INITIALIZE we
cannot distinguish between EGFX not being ever used, and EGFX
having been torn down. Consequently, when running non-GFX, we don't
correctly recover the session.
* Allow multiple reasons for suppress_output
Replaces the single boolean for suppress_output with
a bitmask, to allow output to be suppressed for
more than one reason
* Disable output during resize
* Add states to dynamic resize
Adds states to the dynamic resize state machine so we wait for a
Deactivation-Reactivation sequence to finish before sending pointer
updates, etc.
* suppress module output during the dynamic resize
* Add support for dynamic resize to VNC backend
xrdp_mm needs to be informed when a resize has been performed so that
the resize stte machine can be updsate.
Since v0.9.9, xrdp has assumed that the "drdynvc" static virtual
channel is available for its exclusive use. With GFX support, it
is necessary to codify this to prevent this sequence of operations:-
- NeutrinoRDP target sends DVC Capabilities Request PDU
- target responds wih DVC Capabilities Response PDU
- xrdp processes this, starting the GFX virtual channel again
In the future, if NeutrinoRDP requires access to virtual channels,
data may somehow need to be passed through to the target while being
parsed and handled appropriately within xrdp.
* GFX: sort versions, flags to return the highest version we support in caps advertise
* GFX: simpify swtich in caps_advertise
* GFX: log skipped capability versions in caps_advertise
- Mostly base functions and utilities necessary to enable RFX
Progressive
- Add more EGFX work & mode flags.
- Update encoder.
- Does not yet include caps determination to enable RFX progressive
(yet).
- Update protocol constants
On Linux, the TCP send buffer size is increased to 32768 if it is less
that this (which it normally is). This however has the effect of disabling
dynamic buffer sizing, leading to a maximum available bandwidth of
max_bandwidth = 262144 (bits) / round_trip_time (secs)
This is not noticeable on a LAN with an RTT of around 0.5ms, but
very noticeable on a WAN with an RTT of 0.25s.
Comments in the config file and manpage in this area are improved, as
is the logging if the parameters are actually set.
- Use clearenv() if it exists
- Don't rely on <limits.h> being pulled in by <sys/param.h>
- Rename the DEFAULT_TYPE macro in sesrun.c. This name appears to be
used on Solaris. It's not a good choice.
Some clients appears to be sending cbClientAddress and/or cbClientDir
as 0 in the TS_EXTENDED_INFO_PACKET. This appears to be at odds with
[MS-RDPBCGR] which requires mandatory terminators for these fields.
The top level socket directory is now called XRDP_SOCKET_ROOT_PATH.
Below that are user-specific directories referred to with the
XRDP_SOCKET_PATH macro - this name is hard-coded into xorgxrdp and
the audio modules as an environment variable.
XRDP_SOCKET_PATH now looks like $XRDP_SOCKET_ROOT_PATH/<uid>
XRDP_SOCKET_PATH is only writeable by the user, and readable by the user
and the xrdp process.
The sockdir is only used when sesman is active. The
call g_mk_socket_path() is removed from os_calls and moved to
sesman.
We also change the permissions on this directory to
0755 rather than 01777 (01000 is the 'sticky bit', S_ISVTX).
The behaviour of g_create_dir() has been modified to not
set S_ISVTX on Linux directories. This is implementation-defined
behaviour according to 1003.1, and is no longer required for the
sockdir.
These calls are now replaced with explicit UTF conversion routines in
the common/string_calls.[hc] and common/parse.[hc] modules.
Also removed:-
- The support code in common/os_calls.c to set the locale to use
these routines.
- The twchar type in arch.h
On enterprise systems, using getgrouplist() (if available)
is more efficient than iterating over the members of the group,
and is also more likely to work
