|
|
|
/*
|
|
|
|
* rfbserver.c - deal with server-side of the RFB protocol.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
* Copyright (C) 2011-2012 D. R. Commander
|
|
|
|
* Copyright (C) 2005 Rohit Kumar, Johannes E. Schindelin
|
|
|
|
* Copyright (C) 2002 RealVNC Ltd.
|
|
|
|
* OSXvnc Copyright (C) 2001 Dan McGuirk <mcguirk@incompleteness.net>.
|
|
|
|
* Original Xvnc code Copyright (C) 1999 AT&T Laboratories Cambridge.
|
|
|
|
* All Rights Reserved.
|
|
|
|
*
|
|
|
|
* This is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License as published by
|
|
|
|
* the Free Software Foundation; either version 2 of the License, or
|
|
|
|
* (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This software is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this software; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
|
|
|
|
* USA.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifdef __STRICT_ANSI__
|
|
|
|
#define _BSD_SOURCE
|
|
|
|
#endif
|
|
|
|
#include <string.h>
|
|
|
|
#include <rfb/rfb.h>
|
|
|
|
#include <rfb/rfbregion.h>
|
|
|
|
#include "private.h"
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_FCNTL_H
|
|
|
|
#include <fcntl.h>
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef WIN32
|
|
|
|
#define write(sock,buf,len) send(sock,buf,len,0)
|
|
|
|
#else
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_UNISTD_H
|
|
|
|
#include <unistd.h>
|
|
|
|
#endif
|
|
|
|
#include <pwd.h>
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_SYS_SOCKET_H
|
|
|
|
#include <sys/socket.h>
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_NETINET_IN_H
|
|
|
|
#include <netinet/in.h>
|
|
|
|
#include <netinet/tcp.h>
|
|
|
|
#include <netdb.h>
|
|
|
|
#include <arpa/inet.h>
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef DEBUGPROTO
|
|
|
|
#undef DEBUGPROTO
|
|
|
|
#define DEBUGPROTO(x) x
|
|
|
|
#else
|
|
|
|
#define DEBUGPROTO(x)
|
|
|
|
#endif
|
|
|
|
#include <stdarg.h>
|
|
|
|
#include <scale.h>
|
|
|
|
/* stst() */
|
|
|
|
#include <sys/types.h>
|
|
|
|
#include <sys/stat.h>
|
|
|
|
#include <unistd.h>
|
|
|
|
/* readdir() */
|
|
|
|
#include <dirent.h>
|
|
|
|
/* errno */
|
|
|
|
#include <errno.h>
|
|
|
|
/* strftime() */
|
|
|
|
#include <time.h>
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_WITH_WEBSOCKETS
|
|
|
|
#include "rfbssl.h"
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef __MINGW32__
|
|
|
|
static int compat_mkdir(const char *path, int mode)
|
|
|
|
{
|
|
|
|
return mkdir(path);
|
|
|
|
}
|
|
|
|
#define mkdir compat_mkdir
|
|
|
|
#endif
|
|
|
|
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
Add TurboVNC encoding support.
TurboVNC is a variant of TightVNC that uses the same client/server protocol (RFB version 3.8t),
and thus it is fully cross-compatible with TightVNC and TigerVNC (with one exception, which is noted below.)
Both the TightVNC and TurboVNC encoders analyze each rectangle, pick out regions of solid color to send
separately, and send the remaining subrectangles using mono, indexed color, JPEG, or raw encoding, depending
on the number of colors in the subrectangle. However, TurboVNC uses a fundamentally different selection
algorithm to determine the appropriate subencoding to use for each subrectangle. Thus, while it sends a
protocol stream that can be decoded by any TightVNC-compatible viewer, the mix of subencoding types in this
protocol stream will be different from those generated by a TightVNC server.
The research that led to TurboVNC is described in the following report:
http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf.
In summary: 20 RFB captures, representing "common" 2D and 3D application workloads (the 3D workloads were
run using VirtualGL), were studied using the TightVNC encoder in isolation. Some of the analysis features
in the TightVNC encoder, such as smoothness detection, were found to generate a lot of CPU usage with little
or no benefit in compression, so those features were disabled. JPEG encoding was accelerated using
libjpeg-turbo (which achieves a 2-4x speedup over plain libjpeg on modern x86 or ARM processors.) Finally,
the "palette threshold" (minimum number of colors that the subrectangle must have before it is compressed
using JPEG or raw) was adjusted to account for the fact that JPEG encoding is now quite a bit faster
(meaning that we can now use it more without a CPU penalty.) TurboVNC has additional optimizations,
such as the ability to count colors and encode JPEG images directly from the framebuffer without first
translating the pixels into RGB. The TurboVNC encoder compares quite favorably in terms of compression
ratio with TightVNC and generally encodes a great deal faster (often an order of magnitude or more.)
The version of the TurboVNC encoder included in this patch is roughly equivalent to the one found in version
0.6 of the Unix TurboVNC Server, with a few minor patches integrated from TurboVNC 1.1. TurboVNC 1.0
added multi-threading capabilities, which can be added in later if desired (at the expense of making
libvncserver depend on libpthread.)
Because TurboVNC uses a fundamentally different mix of subencodings than TightVNC, because it uses
the identical protocol (and thus a viewer really has no idea whether it's talking to a TightVNC or
TurboVNC server), and because it doesn't support rfbTightPng (and in fact conflicts with it-- see below),
the TurboVNC and TightVNC encoders cannot be enabled simultaneously.
Compatibility:
In *most* cases, a TurboVNC-enabled viewer is fully compatible with a TightVNC server, and vice versa.
TurboVNC supports pseudo-encodings for specifying a fine-grained (1-100) quality scale and specifying
chrominance subsampling. If a TurboVNC viewer sends those to a TightVNC server, then the TightVNC server
ignores them, so the TurboVNC viewer also sends the quality on a 0-9 scale that the TightVNC server can
understand. Similarly, the TurboVNC server checks first for fine-grained quality and subsampling
pseudo-encodings from the viewer, and failing to receive those, it then checks for the TightVNC 0-9
quality pseudo-encoding.
There is one case in which the two systems are not compatible, and that is when a TightVNC or TigerVNC
viewer requests compression level 0 without JPEG from a TurboVNC server. For performance reasons,
this causes the TurboVNC server to send images directly to the viewer, bypassing Zlib. When the
TurboVNC server does this, it also sets bits 7-4 in the compression control byte to rfbTightNoZlib (0x0A),
which is unfortunately the same value as rfbTightPng. Older TightVNC viewers that don't handle PNG
will assume that the stream is uncompressed but still encapsulated in a Zlib structure, whereas newer
PNG-supporting TightVNC viewers will assume that the stream is PNG. In either case, the viewer will
probably crash. Since most VNC viewers don't expose compression level 0 in the GUI, this is a
relatively rare situation.
Description of changes:
configure.ac
-- Added support for libjpeg-turbo. If passed an argument of --with-turbovnc, configure will now run
(or, if cross-compiling, just link) a test program that determines whether the libjpeg library being
used is libjpeg-turbo. libjpeg-turbo must be used when building the TurboVNC encoder, because the
TurboVNC encoder relies on the libjpeg-turbo colorspace extensions in order to compress images directly
out of the framebuffer (which may be, for instance, BGRA rather than RGB.) libjpeg-turbo can optionally
be used with the TightVNC encoder as well, but the speedup will only be marginal (the report linked
above explains why in more detail, but basically it's because of Amdahl's Law. The TightVNC encoder
was designed with the assumption that JPEG had a very high CPU cost, and thus JPEG is used only sparingly.)
-- Added a new configure variable, JPEG_LDFLAGS. This is necessitated by the fact that libjpeg-turbo
often distributes libjpeg.a and libjpeg.so in /opt/libjpeg-turbo/lib32 or /opt/libjpeg-turbo/lib64,
and many people prefer to statically link with it. Thus, more flexibility is needed than is provided
by --with-jpeg. If JPEG_LDFLAGS is specified, then it overrides the changes to LDFLAGS enacted by
--with-jpeg (but --with-jpeg is still used to set the include path.) The addition of JPEG_LDFLAGS
necessitated replacing AC_CHECK_LIB with AC_LINK_IFELSE (because AC_CHECK_LIB automatically sets
LIBS to -ljpeg, which is not what we want if we're, for instance, linking statically with libjpeg-turbo.)
-- configure does not check for PNG support if TurboVNC encoding is enabled. This prevents the
rfbSendRectEncodingTightPng() function from being compiled in, since the TurboVNC encoder doesn't
(and can't) support it.
common/turbojpeg.c, common/turbojpeg.h
-- TurboJPEG is a simple API used to compress and decompress JPEG images in memory. It was originally
implemented because it was desirable to use different types of underlying technologies to compress
JPEG on different platforms (mediaLib on SPARC, Quicktime on PPC Macs, Intel Performance Primitives, etc.)
These days, however, libjpeg-turbo is the only underlying technology used by TurboVNC, so TurboJPEG's
purpose is largely just code simplicity and flexibility. Thus, since there is no real need for
libvncserver to use any technology other than libjpeg-turbo for compressing JPEG, the TurboJPEG wrapper
for libjpeg-turbo has been included in-tree so that libvncserver can be directly linked with libjpeg-turbo.
This is convenient because many modern Linux distros (Fedora, Ubuntu, etc.) now ship libjpeg-turbo as
their default libjpeg library.
libvncserver/rfbserver.c
-- Added logic to check for the TurboVNC fine-grained quality level and subsampling encodings and to
map Tight (0-9) quality levels to appropriate fine-grained quality level and subsampling values if
communicating with a TightVNC/TigerVNC viewer.
libvncserver/turbo.c
-- TurboVNC encoder (compiled instead of libvncserver/tight.c)
rfb/rfb.h
-- Added support for the TurboVNC subsampling level
rfb/rfbproto.h
-- Added constants for the TurboVNC fine quality level and subsampling encodings as well as the rfbTightNoZlib
constant and notes on its usage.
13 years ago
|
|
|
/*
|
|
|
|
* Map of quality levels to provide compatibility with TightVNC/TigerVNC
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
* clients. This emulates the behavior of the TigerVNC Server.
|
Add TurboVNC encoding support.
TurboVNC is a variant of TightVNC that uses the same client/server protocol (RFB version 3.8t),
and thus it is fully cross-compatible with TightVNC and TigerVNC (with one exception, which is noted below.)
Both the TightVNC and TurboVNC encoders analyze each rectangle, pick out regions of solid color to send
separately, and send the remaining subrectangles using mono, indexed color, JPEG, or raw encoding, depending
on the number of colors in the subrectangle. However, TurboVNC uses a fundamentally different selection
algorithm to determine the appropriate subencoding to use for each subrectangle. Thus, while it sends a
protocol stream that can be decoded by any TightVNC-compatible viewer, the mix of subencoding types in this
protocol stream will be different from those generated by a TightVNC server.
The research that led to TurboVNC is described in the following report:
http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf.
In summary: 20 RFB captures, representing "common" 2D and 3D application workloads (the 3D workloads were
run using VirtualGL), were studied using the TightVNC encoder in isolation. Some of the analysis features
in the TightVNC encoder, such as smoothness detection, were found to generate a lot of CPU usage with little
or no benefit in compression, so those features were disabled. JPEG encoding was accelerated using
libjpeg-turbo (which achieves a 2-4x speedup over plain libjpeg on modern x86 or ARM processors.) Finally,
the "palette threshold" (minimum number of colors that the subrectangle must have before it is compressed
using JPEG or raw) was adjusted to account for the fact that JPEG encoding is now quite a bit faster
(meaning that we can now use it more without a CPU penalty.) TurboVNC has additional optimizations,
such as the ability to count colors and encode JPEG images directly from the framebuffer without first
translating the pixels into RGB. The TurboVNC encoder compares quite favorably in terms of compression
ratio with TightVNC and generally encodes a great deal faster (often an order of magnitude or more.)
The version of the TurboVNC encoder included in this patch is roughly equivalent to the one found in version
0.6 of the Unix TurboVNC Server, with a few minor patches integrated from TurboVNC 1.1. TurboVNC 1.0
added multi-threading capabilities, which can be added in later if desired (at the expense of making
libvncserver depend on libpthread.)
Because TurboVNC uses a fundamentally different mix of subencodings than TightVNC, because it uses
the identical protocol (and thus a viewer really has no idea whether it's talking to a TightVNC or
TurboVNC server), and because it doesn't support rfbTightPng (and in fact conflicts with it-- see below),
the TurboVNC and TightVNC encoders cannot be enabled simultaneously.
Compatibility:
In *most* cases, a TurboVNC-enabled viewer is fully compatible with a TightVNC server, and vice versa.
TurboVNC supports pseudo-encodings for specifying a fine-grained (1-100) quality scale and specifying
chrominance subsampling. If a TurboVNC viewer sends those to a TightVNC server, then the TightVNC server
ignores them, so the TurboVNC viewer also sends the quality on a 0-9 scale that the TightVNC server can
understand. Similarly, the TurboVNC server checks first for fine-grained quality and subsampling
pseudo-encodings from the viewer, and failing to receive those, it then checks for the TightVNC 0-9
quality pseudo-encoding.
There is one case in which the two systems are not compatible, and that is when a TightVNC or TigerVNC
viewer requests compression level 0 without JPEG from a TurboVNC server. For performance reasons,
this causes the TurboVNC server to send images directly to the viewer, bypassing Zlib. When the
TurboVNC server does this, it also sets bits 7-4 in the compression control byte to rfbTightNoZlib (0x0A),
which is unfortunately the same value as rfbTightPng. Older TightVNC viewers that don't handle PNG
will assume that the stream is uncompressed but still encapsulated in a Zlib structure, whereas newer
PNG-supporting TightVNC viewers will assume that the stream is PNG. In either case, the viewer will
probably crash. Since most VNC viewers don't expose compression level 0 in the GUI, this is a
relatively rare situation.
Description of changes:
configure.ac
-- Added support for libjpeg-turbo. If passed an argument of --with-turbovnc, configure will now run
(or, if cross-compiling, just link) a test program that determines whether the libjpeg library being
used is libjpeg-turbo. libjpeg-turbo must be used when building the TurboVNC encoder, because the
TurboVNC encoder relies on the libjpeg-turbo colorspace extensions in order to compress images directly
out of the framebuffer (which may be, for instance, BGRA rather than RGB.) libjpeg-turbo can optionally
be used with the TightVNC encoder as well, but the speedup will only be marginal (the report linked
above explains why in more detail, but basically it's because of Amdahl's Law. The TightVNC encoder
was designed with the assumption that JPEG had a very high CPU cost, and thus JPEG is used only sparingly.)
-- Added a new configure variable, JPEG_LDFLAGS. This is necessitated by the fact that libjpeg-turbo
often distributes libjpeg.a and libjpeg.so in /opt/libjpeg-turbo/lib32 or /opt/libjpeg-turbo/lib64,
and many people prefer to statically link with it. Thus, more flexibility is needed than is provided
by --with-jpeg. If JPEG_LDFLAGS is specified, then it overrides the changes to LDFLAGS enacted by
--with-jpeg (but --with-jpeg is still used to set the include path.) The addition of JPEG_LDFLAGS
necessitated replacing AC_CHECK_LIB with AC_LINK_IFELSE (because AC_CHECK_LIB automatically sets
LIBS to -ljpeg, which is not what we want if we're, for instance, linking statically with libjpeg-turbo.)
-- configure does not check for PNG support if TurboVNC encoding is enabled. This prevents the
rfbSendRectEncodingTightPng() function from being compiled in, since the TurboVNC encoder doesn't
(and can't) support it.
common/turbojpeg.c, common/turbojpeg.h
-- TurboJPEG is a simple API used to compress and decompress JPEG images in memory. It was originally
implemented because it was desirable to use different types of underlying technologies to compress
JPEG on different platforms (mediaLib on SPARC, Quicktime on PPC Macs, Intel Performance Primitives, etc.)
These days, however, libjpeg-turbo is the only underlying technology used by TurboVNC, so TurboJPEG's
purpose is largely just code simplicity and flexibility. Thus, since there is no real need for
libvncserver to use any technology other than libjpeg-turbo for compressing JPEG, the TurboJPEG wrapper
for libjpeg-turbo has been included in-tree so that libvncserver can be directly linked with libjpeg-turbo.
This is convenient because many modern Linux distros (Fedora, Ubuntu, etc.) now ship libjpeg-turbo as
their default libjpeg library.
libvncserver/rfbserver.c
-- Added logic to check for the TurboVNC fine-grained quality level and subsampling encodings and to
map Tight (0-9) quality levels to appropriate fine-grained quality level and subsampling values if
communicating with a TightVNC/TigerVNC viewer.
libvncserver/turbo.c
-- TurboVNC encoder (compiled instead of libvncserver/tight.c)
rfb/rfb.h
-- Added support for the TurboVNC subsampling level
rfb/rfbproto.h
-- Added constants for the TurboVNC fine quality level and subsampling encodings as well as the rfbTightNoZlib
constant and notes on its usage.
13 years ago
|
|
|
*/
|
|
|
|
|
|
|
|
static const int tight2turbo_qual[10] = {
|
|
|
|
15, 29, 41, 42, 62, 77, 79, 86, 92, 100
|
|
|
|
};
|
|
|
|
|
|
|
|
static const int tight2turbo_subsamp[10] = {
|
|
|
|
1, 1, 1, 2, 2, 2, 0, 0, 0, 0
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
|
|
|
|
static void rfbProcessClientProtocolVersion(rfbClientPtr cl);
|
|
|
|
static void rfbProcessClientNormalMessage(rfbClientPtr cl);
|
|
|
|
static void rfbProcessClientInitMessage(rfbClientPtr cl);
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPTHREAD
|
|
|
|
void rfbIncrClientRef(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
LOCK(cl->refCountMutex);
|
|
|
|
cl->refCount++;
|
|
|
|
UNLOCK(cl->refCountMutex);
|
|
|
|
}
|
|
|
|
|
|
|
|
void rfbDecrClientRef(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
LOCK(cl->refCountMutex);
|
|
|
|
cl->refCount--;
|
|
|
|
if(cl->refCount<=0) /* just to be sure also < 0 */
|
|
|
|
TSIGNAL(cl->deleteCond);
|
|
|
|
UNLOCK(cl->refCountMutex);
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
void rfbIncrClientRef(rfbClientPtr cl) {}
|
|
|
|
void rfbDecrClientRef(rfbClientPtr cl) {}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPTHREAD
|
|
|
|
static MUTEX(rfbClientListMutex);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
struct rfbClientIterator {
|
|
|
|
rfbClientPtr next;
|
|
|
|
rfbScreenInfoPtr screen;
|
|
|
|
rfbBool closedToo;
|
|
|
|
};
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbClientListInit(rfbScreenInfoPtr rfbScreen)
|
|
|
|
{
|
|
|
|
if(sizeof(rfbBool)!=1) {
|
|
|
|
/* a sanity check */
|
|
|
|
fprintf(stderr,"rfbBool's size is not 1 (%d)!\n",(int)sizeof(rfbBool));
|
|
|
|
/* we cannot continue, because rfbBool is supposed to be char everywhere */
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
rfbScreen->clientHead = NULL;
|
|
|
|
INIT_MUTEX(rfbClientListMutex);
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbClientIteratorPtr
|
|
|
|
rfbGetClientIterator(rfbScreenInfoPtr rfbScreen)
|
|
|
|
{
|
|
|
|
rfbClientIteratorPtr i =
|
|
|
|
(rfbClientIteratorPtr)malloc(sizeof(struct rfbClientIterator));
|
|
|
|
i->next = NULL;
|
|
|
|
i->screen = rfbScreen;
|
|
|
|
i->closedToo = FALSE;
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbClientIteratorPtr
|
|
|
|
rfbGetClientIteratorWithClosed(rfbScreenInfoPtr rfbScreen)
|
|
|
|
{
|
|
|
|
rfbClientIteratorPtr i =
|
|
|
|
(rfbClientIteratorPtr)malloc(sizeof(struct rfbClientIterator));
|
|
|
|
i->next = NULL;
|
|
|
|
i->screen = rfbScreen;
|
|
|
|
i->closedToo = TRUE;
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbClientPtr
|
|
|
|
rfbClientIteratorHead(rfbClientIteratorPtr i)
|
|
|
|
{
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPTHREAD
|
|
|
|
if(i->next != 0) {
|
|
|
|
rfbDecrClientRef(i->next);
|
|
|
|
rfbIncrClientRef(i->screen->clientHead);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
LOCK(rfbClientListMutex);
|
|
|
|
i->next = i->screen->clientHead;
|
|
|
|
UNLOCK(rfbClientListMutex);
|
|
|
|
return i->next;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbClientPtr
|
|
|
|
rfbClientIteratorNext(rfbClientIteratorPtr i)
|
|
|
|
{
|
|
|
|
if(i->next == 0) {
|
|
|
|
LOCK(rfbClientListMutex);
|
|
|
|
i->next = i->screen->clientHead;
|
|
|
|
UNLOCK(rfbClientListMutex);
|
|
|
|
} else {
|
|
|
|
IF_PTHREADS(rfbClientPtr cl = i->next);
|
|
|
|
i->next = i->next->next;
|
|
|
|
IF_PTHREADS(rfbDecrClientRef(cl));
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPTHREAD
|
|
|
|
if(!i->closedToo)
|
|
|
|
while(i->next && i->next->sock<0)
|
|
|
|
i->next = i->next->next;
|
|
|
|
if(i->next)
|
|
|
|
rfbIncrClientRef(i->next);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
return i->next;
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbReleaseClientIterator(rfbClientIteratorPtr iterator)
|
|
|
|
{
|
|
|
|
IF_PTHREADS(if(iterator->next) rfbDecrClientRef(iterator->next));
|
|
|
|
free(iterator);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbNewClientConnection is called from sockets.c when a new connection
|
|
|
|
* comes in.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbNewClientConnection(rfbScreenInfoPtr rfbScreen,
|
|
|
|
int sock)
|
|
|
|
{
|
|
|
|
rfbNewClient(rfbScreen,sock);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbReverseConnection is called to make an outward
|
|
|
|
* connection to a "listening" RFB client.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbClientPtr
|
|
|
|
rfbReverseConnection(rfbScreenInfoPtr rfbScreen,
|
|
|
|
char *host,
|
|
|
|
int port)
|
|
|
|
{
|
|
|
|
int sock;
|
|
|
|
rfbClientPtr cl;
|
|
|
|
|
|
|
|
if ((sock = rfbConnect(rfbScreen, host, port)) < 0)
|
|
|
|
return (rfbClientPtr)NULL;
|
|
|
|
|
|
|
|
cl = rfbNewClient(rfbScreen, sock);
|
|
|
|
|
|
|
|
if (cl) {
|
|
|
|
cl->reverseConnection = TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return cl;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbSetProtocolVersion(rfbScreenInfoPtr rfbScreen, int major_, int minor_)
|
|
|
|
{
|
|
|
|
/* Permit the server to set the version to report */
|
|
|
|
/* TODO: sanity checking */
|
|
|
|
if ((major_==3) && (minor_ > 2 && minor_ < 9))
|
|
|
|
{
|
|
|
|
rfbScreen->protocolMajorVersion = major_;
|
|
|
|
rfbScreen->protocolMinorVersion = minor_;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
rfbLog("rfbSetProtocolVersion(%d,%d) set to invalid values\n", major_, minor_);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbNewClient is called when a new connection has been made by whatever
|
|
|
|
* means.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static rfbClientPtr
|
|
|
|
rfbNewTCPOrUDPClient(rfbScreenInfoPtr rfbScreen,
|
|
|
|
int sock,
|
|
|
|
rfbBool isUDP)
|
|
|
|
{
|
|
|
|
rfbProtocolVersionMsg pv;
|
|
|
|
rfbClientIteratorPtr iterator;
|
|
|
|
rfbClientPtr cl,cl_;
|
|
|
|
#ifdef LIBVNCSERVER_IPv6
|
|
|
|
struct sockaddr_storage addr;
|
|
|
|
#else
|
|
|
|
struct sockaddr_in addr;
|
|
|
|
#endif
|
|
|
|
socklen_t addrlen = sizeof(addr);
|
|
|
|
rfbProtocolExtension* extension;
|
|
|
|
|
|
|
|
cl = (rfbClientPtr)calloc(sizeof(rfbClientRec),1);
|
|
|
|
|
|
|
|
cl->screen = rfbScreen;
|
|
|
|
cl->sock = sock;
|
|
|
|
cl->viewOnly = FALSE;
|
|
|
|
/* setup pseudo scaling */
|
|
|
|
cl->scaledScreen = rfbScreen;
|
|
|
|
cl->scaledScreen->scaledScreenRefCount++;
|
|
|
|
|
|
|
|
rfbResetStats(cl);
|
|
|
|
|
|
|
|
cl->clientData = NULL;
|
|
|
|
cl->clientGoneHook = rfbDoNothingWithClient;
|
|
|
|
|
|
|
|
if(isUDP) {
|
|
|
|
rfbLog(" accepted UDP client\n");
|
|
|
|
} else {
|
|
|
|
int one=1;
|
|
|
|
|
|
|
|
getpeername(sock, (struct sockaddr *)&addr, &addrlen);
|
|
|
|
#ifdef LIBVNCSERVER_IPv6
|
|
|
|
char host[1024];
|
|
|
|
if(getnameinfo((struct sockaddr*)&addr, addrlen, host, sizeof(host), NULL, 0, NI_NUMERICHOST) != 0) {
|
|
|
|
rfbLogPerror("rfbNewClient: error in getnameinfo");
|
|
|
|
cl->host = strdup("");
|
|
|
|
}
|
|
|
|
else
|
|
|
|
cl->host = strdup(host);
|
|
|
|
#else
|
|
|
|
cl->host = strdup(inet_ntoa(addr.sin_addr));
|
|
|
|
#endif
|
|
|
|
|
|
|
|
rfbLog(" other clients:\n");
|
|
|
|
iterator = rfbGetClientIterator(rfbScreen);
|
|
|
|
while ((cl_ = rfbClientIteratorNext(iterator)) != NULL) {
|
|
|
|
rfbLog(" %s\n",cl_->host);
|
|
|
|
}
|
|
|
|
rfbReleaseClientIterator(iterator);
|
|
|
|
|
|
|
|
if(!rfbSetNonBlocking(sock)) {
|
|
|
|
close(sock);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (setsockopt(sock, IPPROTO_TCP, TCP_NODELAY,
|
|
|
|
(char *)&one, sizeof(one)) < 0) {
|
|
|
|
rfbLogPerror("setsockopt failed");
|
|
|
|
close(sock);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
FD_SET(sock,&(rfbScreen->allFds));
|
|
|
|
rfbScreen->maxFd = max(sock,rfbScreen->maxFd);
|
|
|
|
|
|
|
|
INIT_MUTEX(cl->outputMutex);
|
|
|
|
INIT_MUTEX(cl->refCountMutex);
|
|
|
|
INIT_MUTEX(cl->sendMutex);
|
|
|
|
INIT_COND(cl->deleteCond);
|
|
|
|
|
|
|
|
cl->state = RFB_PROTOCOL_VERSION;
|
|
|
|
|
|
|
|
cl->reverseConnection = FALSE;
|
|
|
|
cl->readyForSetColourMapEntries = FALSE;
|
|
|
|
cl->useCopyRect = FALSE;
|
|
|
|
cl->preferredEncoding = -1;
|
|
|
|
cl->correMaxWidth = 48;
|
|
|
|
cl->correMaxHeight = 48;
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
cl->zrleData = NULL;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
cl->copyRegion = sraRgnCreate();
|
|
|
|
cl->copyDX = 0;
|
|
|
|
cl->copyDY = 0;
|
|
|
|
|
|
|
|
cl->modifiedRegion =
|
|
|
|
sraRgnCreateRect(0,0,rfbScreen->width,rfbScreen->height);
|
|
|
|
|
|
|
|
INIT_MUTEX(cl->updateMutex);
|
|
|
|
INIT_COND(cl->updateCond);
|
|
|
|
|
|
|
|
cl->requestedRegion = sraRgnCreate();
|
|
|
|
|
|
|
|
cl->format = cl->screen->serverFormat;
|
|
|
|
cl->translateFn = rfbTranslateNone;
|
|
|
|
cl->translateLookupTable = NULL;
|
|
|
|
|
|
|
|
LOCK(rfbClientListMutex);
|
|
|
|
|
|
|
|
IF_PTHREADS(cl->refCount = 0);
|
|
|
|
cl->next = rfbScreen->clientHead;
|
|
|
|
cl->prev = NULL;
|
|
|
|
if (rfbScreen->clientHead)
|
|
|
|
rfbScreen->clientHead->prev = cl;
|
|
|
|
|
|
|
|
rfbScreen->clientHead = cl;
|
|
|
|
UNLOCK(rfbClientListMutex);
|
|
|
|
|
|
|
|
#if defined(LIBVNCSERVER_HAVE_LIBZ) || defined(LIBVNCSERVER_HAVE_LIBPNG)
|
|
|
|
cl->tightQualityLevel = -1;
|
|
|
|
#if defined(LIBVNCSERVER_HAVE_LIBJPEG) || defined(LIBVNCSERVER_HAVE_LIBPNG)
|
|
|
|
cl->tightCompressLevel = TIGHT_DEFAULT_COMPRESSION;
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
cl->turboSubsampLevel = TURBO_DEFAULT_SUBSAMP;
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
for (i = 0; i < 4; i++)
|
|
|
|
cl->zsActive[i] = FALSE;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
|
|
|
cl->fileTransfer.fd = -1;
|
|
|
|
|
|
|
|
cl->enableCursorShapeUpdates = FALSE;
|
|
|
|
cl->enableCursorPosUpdates = FALSE;
|
|
|
|
cl->useRichCursorEncoding = FALSE;
|
|
|
|
cl->enableLastRectEncoding = FALSE;
|
|
|
|
cl->enableKeyboardLedState = FALSE;
|
|
|
|
cl->enableSupportedMessages = FALSE;
|
|
|
|
cl->enableSupportedEncodings = FALSE;
|
|
|
|
cl->enableServerIdentity = FALSE;
|
|
|
|
cl->lastKeyboardLedState = -1;
|
|
|
|
cl->cursorX = rfbScreen->cursorX;
|
|
|
|
cl->cursorY = rfbScreen->cursorY;
|
|
|
|
cl->useNewFBSize = FALSE;
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
cl->compStreamInited = FALSE;
|
|
|
|
cl->compStream.total_in = 0;
|
|
|
|
cl->compStream.total_out = 0;
|
|
|
|
cl->compStream.zalloc = Z_NULL;
|
|
|
|
cl->compStream.zfree = Z_NULL;
|
|
|
|
cl->compStream.opaque = Z_NULL;
|
|
|
|
|
|
|
|
cl->zlibCompressLevel = 5;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
cl->progressiveSliceY = 0;
|
|
|
|
|
|
|
|
cl->extensions = NULL;
|
|
|
|
|
|
|
|
cl->lastPtrX = -1;
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_WITH_WEBSOCKETS
|
|
|
|
/*
|
|
|
|
* Wait a few ms for the client to send one of:
|
|
|
|
* - Flash policy request
|
|
|
|
* - WebSockets connection (TLS/SSL or plain)
|
|
|
|
*/
|
|
|
|
if (!webSocketsCheck(cl)) {
|
|
|
|
/* Error reporting handled in webSocketsHandshake */
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
rfbClientConnectionGone(cl);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
sprintf(pv,rfbProtocolVersionFormat,rfbScreen->protocolMajorVersion,
|
|
|
|
rfbScreen->protocolMinorVersion);
|
|
|
|
|
|
|
|
if (rfbWriteExact(cl, pv, sz_rfbProtocolVersionMsg) < 0) {
|
|
|
|
rfbLogPerror("rfbNewClient: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
rfbClientConnectionGone(cl);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for(extension = rfbGetExtensionIterator(); extension;
|
|
|
|
extension=extension->next) {
|
|
|
|
void* data = NULL;
|
|
|
|
/* if the extension does not have a newClient method, it wants
|
|
|
|
* to be initialized later. */
|
|
|
|
if(extension->newClient && extension->newClient(cl, &data))
|
|
|
|
rfbEnableExtension(cl, extension, data);
|
|
|
|
}
|
|
|
|
rfbReleaseExtensionIterator();
|
|
|
|
|
|
|
|
switch (cl->screen->newClientHook(cl)) {
|
|
|
|
case RFB_CLIENT_ON_HOLD:
|
|
|
|
cl->onHold = TRUE;
|
|
|
|
break;
|
|
|
|
case RFB_CLIENT_ACCEPT:
|
|
|
|
cl->onHold = FALSE;
|
|
|
|
break;
|
|
|
|
case RFB_CLIENT_REFUSE:
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
rfbClientConnectionGone(cl);
|
|
|
|
cl = NULL;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return cl;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbClientPtr
|
|
|
|
rfbNewClient(rfbScreenInfoPtr rfbScreen,
|
|
|
|
int sock)
|
|
|
|
{
|
|
|
|
return(rfbNewTCPOrUDPClient(rfbScreen,sock,FALSE));
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbClientPtr
|
|
|
|
rfbNewUDPClient(rfbScreenInfoPtr rfbScreen)
|
|
|
|
{
|
|
|
|
return((rfbScreen->udpClient=
|
|
|
|
rfbNewTCPOrUDPClient(rfbScreen,rfbScreen->udpSock,TRUE)));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbClientConnectionGone is called from sockets.c just after a connection
|
|
|
|
* has gone away.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbClientConnectionGone(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
int i;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
LOCK(rfbClientListMutex);
|
|
|
|
|
|
|
|
if (cl->prev)
|
|
|
|
cl->prev->next = cl->next;
|
|
|
|
else
|
|
|
|
cl->screen->clientHead = cl->next;
|
|
|
|
if (cl->next)
|
|
|
|
cl->next->prev = cl->prev;
|
|
|
|
|
Fix deadlock in threaded mode when using nested rfbClientIteratorNext() calls.
Lengthy explanation follows...
First, the scenario before this patch:
We have three clients 1,2,3 connected. The main thread loops through
them using rfbClientIteratorNext() (loop L1) and is currently at
client 2 i.e. client 2's cl_2->refCount is 1. At this point we need to
loop again through the clients, with cl_2->refCount == 1, i.e. do a
loop L2 nested within loop L1.
BUT: Now client 2 disconnects, it's clientInput thread terminates its
clientOutput thread and calls rfbClientConnectionGone(). This LOCKs
clientListMutex and WAITs for cl_2->refCount to become 0. This means
this thread waits for the main thread to release cl_2. Waiting, with
clientListMutex LOCKed!
Meanwhile, the main thread is about to begin the inner
rfbClientIteratorNext() loop L2. The first call to rfbClientIteratorNext()
LOCKs clientListMutex. BAAM. This mutex is locked by cl2's clientInput
thread and is only released when cl_2->refCount becomes 0. The main thread
would decrement cl_2->refCount when it would continue with loop L1. But
it's waiting for cl2's clientInput thread to release clientListMutex. Which
never happens since this one's waiting for the main thread to decrement
cl_2->refCount. DEADLOCK.
Now, situation with this patch:
Same as above, but when client 2 disconnects it's clientInput thread
rfbClientConnectionGone(). This again LOCKs clientListMutex, removes cl_2
from the linked list and UNLOCKS clientListMutex. The WAIT for
cl_2->refCount to become 0 is _after_ that. Waiting, with
clientListMutex UNLOCKed!
Therefore, the main thread can continue, do the inner loop L2 (now only
looping through 1,3 - 2 was removed from the linked list) and continue with
loop L1, finally decrementing cl_2->refCount, allowing cl2's clientInput
thread to continue and terminate. The resources held by cl2 are not free()'d
by rfbClientConnectionGone until cl2->refCount becomes 0, i.e. loop L1 has
released cl2.
13 years ago
|
|
|
UNLOCK(rfbClientListMutex);
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPTHREAD
|
|
|
|
if(cl->screen->backgroundLoop != FALSE) {
|
|
|
|
int i;
|
|
|
|
do {
|
|
|
|
LOCK(cl->refCountMutex);
|
|
|
|
i=cl->refCount;
|
|
|
|
if(i>0)
|
|
|
|
WAIT(cl->deleteCond,cl->refCountMutex);
|
|
|
|
UNLOCK(cl->refCountMutex);
|
|
|
|
} while(i>0);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if(cl->sock>=0)
|
|
|
|
close(cl->sock);
|
|
|
|
|
|
|
|
if (cl->scaledScreen!=NULL)
|
|
|
|
cl->scaledScreen->scaledScreenRefCount--;
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
rfbFreeZrleData(cl);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
rfbFreeUltraData(cl);
|
|
|
|
|
|
|
|
/* free buffers holding pixel data before and after encoding */
|
|
|
|
free(cl->beforeEncBuf);
|
|
|
|
free(cl->afterEncBuf);
|
|
|
|
|
|
|
|
if(cl->sock>=0)
|
|
|
|
FD_CLR(cl->sock,&(cl->screen->allFds));
|
|
|
|
|
|
|
|
cl->clientGoneHook(cl);
|
|
|
|
|
|
|
|
rfbLog("Client %s gone\n",cl->host);
|
|
|
|
free(cl->host);
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
/* Release the compression state structures if any. */
|
|
|
|
if ( cl->compStreamInited ) {
|
|
|
|
deflateEnd( &(cl->compStream) );
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
if (cl->zsActive[i])
|
|
|
|
deflateEnd(&cl->zsStruct[i]);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if (cl->screen->pointerClient == cl)
|
|
|
|
cl->screen->pointerClient = NULL;
|
|
|
|
|
|
|
|
sraRgnDestroy(cl->modifiedRegion);
|
|
|
|
sraRgnDestroy(cl->requestedRegion);
|
|
|
|
sraRgnDestroy(cl->copyRegion);
|
|
|
|
|
|
|
|
if (cl->translateLookupTable) free(cl->translateLookupTable);
|
|
|
|
|
|
|
|
TINI_COND(cl->updateCond);
|
|
|
|
TINI_MUTEX(cl->updateMutex);
|
|
|
|
|
|
|
|
/* make sure outputMutex is unlocked before destroying */
|
|
|
|
LOCK(cl->outputMutex);
|
|
|
|
UNLOCK(cl->outputMutex);
|
|
|
|
TINI_MUTEX(cl->outputMutex);
|
|
|
|
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
TINI_MUTEX(cl->sendMutex);
|
|
|
|
|
|
|
|
rfbPrintStats(cl);
|
|
|
|
rfbResetStats(cl);
|
|
|
|
|
|
|
|
free(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbProcessClientMessage is called when there is data to read from a client.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbProcessClientMessage(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
switch (cl->state) {
|
|
|
|
case RFB_PROTOCOL_VERSION:
|
|
|
|
rfbProcessClientProtocolVersion(cl);
|
|
|
|
return;
|
|
|
|
case RFB_SECURITY_TYPE:
|
|
|
|
rfbProcessClientSecurityType(cl);
|
|
|
|
return;
|
|
|
|
case RFB_AUTHENTICATION:
|
|
|
|
rfbAuthProcessClientMessage(cl);
|
|
|
|
return;
|
|
|
|
case RFB_INITIALISATION:
|
|
|
|
case RFB_INITIALISATION_SHARED:
|
|
|
|
rfbProcessClientInitMessage(cl);
|
|
|
|
return;
|
|
|
|
default:
|
|
|
|
rfbProcessClientNormalMessage(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbProcessClientProtocolVersion is called when the client sends its
|
|
|
|
* protocol version.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void
|
|
|
|
rfbProcessClientProtocolVersion(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbProtocolVersionMsg pv;
|
|
|
|
int n, major_, minor_;
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, pv, sz_rfbProtocolVersionMsg)) <= 0) {
|
|
|
|
if (n == 0)
|
|
|
|
rfbLog("rfbProcessClientProtocolVersion: client gone\n");
|
|
|
|
else
|
|
|
|
rfbLogPerror("rfbProcessClientProtocolVersion: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
pv[sz_rfbProtocolVersionMsg] = 0;
|
|
|
|
if (sscanf(pv,rfbProtocolVersionFormat,&major_,&minor_) != 2) {
|
|
|
|
rfbErr("rfbProcessClientProtocolVersion: not a valid RFB client: %s\n", pv);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbLog("Client Protocol Version %d.%d\n", major_, minor_);
|
|
|
|
|
|
|
|
if (major_ != rfbProtocolMajorVersion) {
|
|
|
|
rfbErr("RFB protocol version mismatch - server %d.%d, client %d.%d",
|
|
|
|
cl->screen->protocolMajorVersion, cl->screen->protocolMinorVersion,
|
|
|
|
major_,minor_);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Check for the minor version use either of the two standard version of RFB */
|
|
|
|
/*
|
|
|
|
* UltraVNC Viewer detects FileTransfer compatible servers via rfb versions
|
|
|
|
* 3.4, 3.6, 3.14, 3.16
|
|
|
|
* It's a bad method, but it is what they use to enable features...
|
|
|
|
* maintaining RFB version compatibility across multiple servers is a pain
|
|
|
|
* Should use something like ServerIdentity encoding
|
|
|
|
*/
|
|
|
|
cl->protocolMajorVersion = major_;
|
|
|
|
cl->protocolMinorVersion = minor_;
|
|
|
|
|
|
|
|
rfbLog("Protocol version sent %d.%d, using %d.%d\n",
|
|
|
|
major_, minor_, rfbProtocolMajorVersion, cl->protocolMinorVersion);
|
|
|
|
|
|
|
|
rfbAuthNewClient(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbClientSendString(rfbClientPtr cl, const char *reason)
|
|
|
|
{
|
|
|
|
char *buf;
|
|
|
|
int len = strlen(reason);
|
|
|
|
|
|
|
|
rfbLog("rfbClientSendString(\"%s\")\n", reason);
|
|
|
|
|
|
|
|
buf = (char *)malloc(4 + len);
|
|
|
|
((uint32_t *)buf)[0] = Swap32IfLE(len);
|
|
|
|
memcpy(buf + 4, reason, len);
|
|
|
|
|
|
|
|
if (rfbWriteExact(cl, buf, 4 + len) < 0)
|
|
|
|
rfbLogPerror("rfbClientSendString: write");
|
|
|
|
free(buf);
|
|
|
|
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbClientConnFailed is called when a client connection has failed either
|
|
|
|
* because it talks the wrong protocol or it has failed authentication.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbClientConnFailed(rfbClientPtr cl,
|
|
|
|
const char *reason)
|
|
|
|
{
|
|
|
|
char *buf;
|
|
|
|
int len = strlen(reason);
|
|
|
|
|
|
|
|
rfbLog("rfbClientConnFailed(\"%s\")\n", reason);
|
|
|
|
|
|
|
|
buf = (char *)malloc(8 + len);
|
|
|
|
((uint32_t *)buf)[0] = Swap32IfLE(rfbConnFailed);
|
|
|
|
((uint32_t *)buf)[1] = Swap32IfLE(len);
|
|
|
|
memcpy(buf + 8, reason, len);
|
|
|
|
|
|
|
|
if (rfbWriteExact(cl, buf, 8 + len) < 0)
|
|
|
|
rfbLogPerror("rfbClientConnFailed: write");
|
|
|
|
free(buf);
|
|
|
|
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbProcessClientInitMessage is called when the client sends its
|
|
|
|
* initialisation message.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void
|
|
|
|
rfbProcessClientInitMessage(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbClientInitMsg ci;
|
|
|
|
union {
|
|
|
|
char buf[256];
|
|
|
|
rfbServerInitMsg si;
|
|
|
|
} u;
|
|
|
|
int len, n;
|
|
|
|
rfbClientIteratorPtr iterator;
|
|
|
|
rfbClientPtr otherCl;
|
|
|
|
rfbExtensionData* extension;
|
|
|
|
|
|
|
|
if (cl->state == RFB_INITIALISATION_SHARED) {
|
|
|
|
/* In this case behave as though an implicit ClientInit message has
|
|
|
|
* already been received with a shared-flag of true. */
|
|
|
|
ci.shared = 1;
|
|
|
|
/* Avoid the possibility of exposing the RFB_INITIALISATION_SHARED
|
|
|
|
* state to calling software. */
|
|
|
|
cl->state = RFB_INITIALISATION;
|
|
|
|
} else {
|
|
|
|
if ((n = rfbReadExact(cl, (char *)&ci,sz_rfbClientInitMsg)) <= 0) {
|
|
|
|
if (n == 0)
|
|
|
|
rfbLog("rfbProcessClientInitMessage: client gone\n");
|
|
|
|
else
|
|
|
|
rfbLogPerror("rfbProcessClientInitMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
memset(u.buf,0,sizeof(u.buf));
|
|
|
|
|
|
|
|
u.si.framebufferWidth = Swap16IfLE(cl->screen->width);
|
|
|
|
u.si.framebufferHeight = Swap16IfLE(cl->screen->height);
|
|
|
|
u.si.format = cl->screen->serverFormat;
|
|
|
|
u.si.format.redMax = Swap16IfLE(u.si.format.redMax);
|
|
|
|
u.si.format.greenMax = Swap16IfLE(u.si.format.greenMax);
|
|
|
|
u.si.format.blueMax = Swap16IfLE(u.si.format.blueMax);
|
|
|
|
|
|
|
|
strncpy(u.buf + sz_rfbServerInitMsg, cl->screen->desktopName, 127);
|
|
|
|
len = strlen(u.buf + sz_rfbServerInitMsg);
|
|
|
|
u.si.nameLength = Swap32IfLE(len);
|
|
|
|
|
|
|
|
if (rfbWriteExact(cl, u.buf, sz_rfbServerInitMsg + len) < 0) {
|
|
|
|
rfbLogPerror("rfbProcessClientInitMessage: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
for(extension = cl->extensions; extension;) {
|
|
|
|
rfbExtensionData* next = extension->next;
|
|
|
|
if(extension->extension->init &&
|
|
|
|
!extension->extension->init(cl, extension->data))
|
|
|
|
/* extension requested that it be removed */
|
|
|
|
rfbDisableExtension(cl, extension->extension);
|
|
|
|
extension = next;
|
|
|
|
}
|
|
|
|
|
|
|
|
cl->state = RFB_NORMAL;
|
|
|
|
|
|
|
|
if (!cl->reverseConnection &&
|
|
|
|
(cl->screen->neverShared || (!cl->screen->alwaysShared && !ci.shared))) {
|
|
|
|
|
|
|
|
if (cl->screen->dontDisconnect) {
|
|
|
|
iterator = rfbGetClientIterator(cl->screen);
|
|
|
|
while ((otherCl = rfbClientIteratorNext(iterator)) != NULL) {
|
|
|
|
if ((otherCl != cl) && (otherCl->state == RFB_NORMAL)) {
|
|
|
|
rfbLog("-dontdisconnect: Not shared & existing client\n");
|
|
|
|
rfbLog(" refusing new client %s\n", cl->host);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
rfbReleaseClientIterator(iterator);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
rfbReleaseClientIterator(iterator);
|
|
|
|
} else {
|
|
|
|
iterator = rfbGetClientIterator(cl->screen);
|
|
|
|
while ((otherCl = rfbClientIteratorNext(iterator)) != NULL) {
|
|
|
|
if ((otherCl != cl) && (otherCl->state == RFB_NORMAL)) {
|
|
|
|
rfbLog("Not shared - closing connection to client %s\n",
|
|
|
|
otherCl->host);
|
|
|
|
rfbCloseClient(otherCl);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
rfbReleaseClientIterator(iterator);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The values come in based on the scaled screen, we need to convert them to
|
|
|
|
* values based on the man screen's coordinate system
|
|
|
|
*/
|
|
|
|
static rfbBool rectSwapIfLEAndClip(uint16_t* x,uint16_t* y,uint16_t* w,uint16_t* h,
|
|
|
|
rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
int x1=Swap16IfLE(*x);
|
|
|
|
int y1=Swap16IfLE(*y);
|
|
|
|
int w1=Swap16IfLE(*w);
|
|
|
|
int h1=Swap16IfLE(*h);
|
|
|
|
|
|
|
|
rfbScaledCorrection(cl->scaledScreen, cl->screen, &x1, &y1, &w1, &h1, "rectSwapIfLEAndClip");
|
|
|
|
*x = x1;
|
|
|
|
*y = y1;
|
|
|
|
*w = w1;
|
|
|
|
*h = h1;
|
|
|
|
|
|
|
|
if(*w>cl->screen->width-*x)
|
|
|
|
*w=cl->screen->width-*x;
|
|
|
|
/* possible underflow */
|
|
|
|
if(*w>cl->screen->width-*x)
|
|
|
|
return FALSE;
|
|
|
|
if(*h>cl->screen->height-*y)
|
|
|
|
*h=cl->screen->height-*y;
|
|
|
|
if(*h>cl->screen->height-*y)
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send keyboard state (PointerPos pseudo-encoding).
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendKeyboardLedState(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbFramebufferUpdateRectHeader > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingKeyboardLedState);
|
|
|
|
rect.r.x = Swap16IfLE(cl->lastKeyboardLedState);
|
|
|
|
rect.r.y = 0;
|
|
|
|
rect.r.w = 0;
|
|
|
|
rect.r.h = 0;
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingKeyboardLedState, sz_rfbFramebufferUpdateRectHeader, sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
#define rfbSetBit(buffer, position) (buffer[(position & 255) / 8] |= (1 << (position % 8)))
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send rfbEncodingSupportedMessages.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendSupportedMessages(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
rfbSupportedMessages msgs;
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbFramebufferUpdateRectHeader
|
|
|
|
+ sz_rfbSupportedMessages > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingSupportedMessages);
|
|
|
|
rect.r.x = 0;
|
|
|
|
rect.r.y = 0;
|
|
|
|
rect.r.w = Swap16IfLE(sz_rfbSupportedMessages);
|
|
|
|
rect.r.h = 0;
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
memset((char *)&msgs, 0, sz_rfbSupportedMessages);
|
|
|
|
rfbSetBit(msgs.client2server, rfbSetPixelFormat);
|
|
|
|
rfbSetBit(msgs.client2server, rfbFixColourMapEntries);
|
|
|
|
rfbSetBit(msgs.client2server, rfbSetEncodings);
|
|
|
|
rfbSetBit(msgs.client2server, rfbFramebufferUpdateRequest);
|
|
|
|
rfbSetBit(msgs.client2server, rfbKeyEvent);
|
|
|
|
rfbSetBit(msgs.client2server, rfbPointerEvent);
|
|
|
|
rfbSetBit(msgs.client2server, rfbClientCutText);
|
|
|
|
rfbSetBit(msgs.client2server, rfbFileTransfer);
|
|
|
|
rfbSetBit(msgs.client2server, rfbSetScale);
|
|
|
|
/*rfbSetBit(msgs.client2server, rfbSetServerInput); */
|
|
|
|
/*rfbSetBit(msgs.client2server, rfbSetSW); */
|
|
|
|
/*rfbSetBit(msgs.client2server, rfbTextChat); */
|
|
|
|
rfbSetBit(msgs.client2server, rfbPalmVNCSetScaleFactor);
|
|
|
|
rfbSetBit(msgs.client2server, rfbXvp);
|
|
|
|
|
|
|
|
rfbSetBit(msgs.server2client, rfbFramebufferUpdate);
|
|
|
|
rfbSetBit(msgs.server2client, rfbSetColourMapEntries);
|
|
|
|
rfbSetBit(msgs.server2client, rfbBell);
|
|
|
|
rfbSetBit(msgs.server2client, rfbServerCutText);
|
|
|
|
rfbSetBit(msgs.server2client, rfbResizeFrameBuffer);
|
|
|
|
rfbSetBit(msgs.server2client, rfbPalmVNCReSizeFrameBuffer);
|
|
|
|
rfbSetBit(msgs.server2client, rfbXvp);
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&msgs, sz_rfbSupportedMessages);
|
|
|
|
cl->ublen += sz_rfbSupportedMessages;
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingSupportedMessages,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader+sz_rfbSupportedMessages,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader+sz_rfbSupportedMessages);
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send rfbEncodingSupportedEncodings.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendSupportedEncodings(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
static uint32_t supported[] = {
|
|
|
|
rfbEncodingRaw,
|
|
|
|
rfbEncodingCopyRect,
|
|
|
|
rfbEncodingRRE,
|
|
|
|
rfbEncodingCoRRE,
|
|
|
|
rfbEncodingHextile,
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
rfbEncodingZlib,
|
|
|
|
rfbEncodingZRLE,
|
|
|
|
rfbEncodingZYWRLE,
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
rfbEncodingTight,
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPNG
|
|
|
|
rfbEncodingTightPng,
|
|
|
|
#endif
|
|
|
|
rfbEncodingUltra,
|
|
|
|
rfbEncodingUltraZip,
|
|
|
|
rfbEncodingXCursor,
|
|
|
|
rfbEncodingRichCursor,
|
|
|
|
rfbEncodingPointerPos,
|
|
|
|
rfbEncodingLastRect,
|
|
|
|
rfbEncodingNewFBSize,
|
|
|
|
rfbEncodingKeyboardLedState,
|
|
|
|
rfbEncodingSupportedMessages,
|
|
|
|
rfbEncodingSupportedEncodings,
|
|
|
|
rfbEncodingServerIdentity,
|
|
|
|
};
|
|
|
|
uint32_t nEncodings = sizeof(supported) / sizeof(supported[0]), i;
|
|
|
|
|
|
|
|
/* think rfbSetEncodingsMsg */
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbFramebufferUpdateRectHeader
|
|
|
|
+ (nEncodings * sizeof(uint32_t)) > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingSupportedEncodings);
|
|
|
|
rect.r.x = 0;
|
|
|
|
rect.r.y = 0;
|
|
|
|
rect.r.w = Swap16IfLE(nEncodings * sizeof(uint32_t));
|
|
|
|
rect.r.h = Swap16IfLE(nEncodings);
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
for (i = 0; i < nEncodings; i++) {
|
|
|
|
uint32_t encoding = Swap32IfLE(supported[i]);
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&encoding, sizeof(encoding));
|
|
|
|
cl->ublen += sizeof(encoding);
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingSupportedEncodings,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader+(nEncodings * sizeof(uint32_t)),
|
|
|
|
sz_rfbFramebufferUpdateRectHeader+(nEncodings * sizeof(uint32_t)));
|
|
|
|
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbSetServerVersionIdentity(rfbScreenInfoPtr screen, char *fmt, ...)
|
|
|
|
{
|
|
|
|
char buffer[256];
|
|
|
|
va_list ap;
|
|
|
|
|
|
|
|
va_start(ap, fmt);
|
|
|
|
vsnprintf(buffer, sizeof(buffer)-1, fmt, ap);
|
|
|
|
va_end(ap);
|
|
|
|
|
|
|
|
if (screen->versionString!=NULL) free(screen->versionString);
|
|
|
|
screen->versionString = strdup(buffer);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send rfbEncodingServerIdentity.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendServerIdentity(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
char buffer[512];
|
|
|
|
|
|
|
|
/* tack on our library version */
|
|
|
|
snprintf(buffer,sizeof(buffer)-1, "%s (%s)",
|
|
|
|
(cl->screen->versionString==NULL ? "unknown" : cl->screen->versionString),
|
|
|
|
LIBVNCSERVER_PACKAGE_STRING);
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbFramebufferUpdateRectHeader
|
|
|
|
+ (strlen(buffer)+1) > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingServerIdentity);
|
|
|
|
rect.r.x = 0;
|
|
|
|
rect.r.y = 0;
|
|
|
|
rect.r.w = Swap16IfLE(strlen(buffer)+1);
|
|
|
|
rect.r.h = 0;
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], buffer, strlen(buffer)+1);
|
|
|
|
cl->ublen += strlen(buffer)+1;
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingServerIdentity,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader+strlen(buffer)+1,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader+strlen(buffer)+1);
|
|
|
|
|
|
|
|
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send an xvp server message
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendXvp(rfbClientPtr cl, uint8_t version, uint8_t code)
|
|
|
|
{
|
|
|
|
rfbXvpMsg xvp;
|
|
|
|
|
|
|
|
xvp.type = rfbXvp;
|
|
|
|
xvp.pad = 0;
|
|
|
|
xvp.version = version;
|
|
|
|
xvp.code = code;
|
|
|
|
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
if (rfbWriteExact(cl, (char *)&xvp, sz_rfbXvpMsg) < 0) {
|
|
|
|
rfbLogPerror("rfbSendXvp: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
}
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
|
|
|
|
rfbStatRecordMessageSent(cl, rfbXvp, sz_rfbXvpMsg, sz_rfbXvpMsg);
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
rfbBool rfbSendTextChatMessage(rfbClientPtr cl, uint32_t length, char *buffer)
|
|
|
|
{
|
|
|
|
rfbTextChatMsg tc;
|
|
|
|
int bytesToSend=0;
|
|
|
|
|
|
|
|
memset((char *)&tc, 0, sizeof(tc));
|
|
|
|
tc.type = rfbTextChat;
|
|
|
|
tc.length = Swap32IfLE(length);
|
|
|
|
|
|
|
|
switch(length) {
|
|
|
|
case rfbTextChatOpen:
|
|
|
|
case rfbTextChatClose:
|
|
|
|
case rfbTextChatFinished:
|
|
|
|
bytesToSend=0;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
bytesToSend=length;
|
|
|
|
if (bytesToSend>rfbTextMaxSize)
|
|
|
|
bytesToSend=rfbTextMaxSize;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbTextChatMsg + bytesToSend > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&tc, sz_rfbTextChatMsg);
|
|
|
|
cl->ublen += sz_rfbTextChatMsg;
|
|
|
|
if (bytesToSend>0) {
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], buffer, bytesToSend);
|
|
|
|
cl->ublen += bytesToSend;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageSent(cl, rfbTextChat, sz_rfbTextChatMsg+bytesToSend, sz_rfbTextChatMsg+bytesToSend);
|
|
|
|
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
#define FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN(msg, cl, ret) \
|
|
|
|
if ((cl->screen->getFileTransferPermission != NULL \
|
|
|
|
&& cl->screen->getFileTransferPermission(cl) != TRUE) \
|
|
|
|
|| cl->screen->permitFileTransfer != TRUE) { \
|
|
|
|
rfbLog("%sUltra File Transfer is disabled, dropping client: %s\n", msg, cl->host); \
|
|
|
|
rfbCloseClient(cl); \
|
|
|
|
return ret; \
|
|
|
|
}
|
|
|
|
|
|
|
|
int DB = 1;
|
|
|
|
|
|
|
|
rfbBool rfbSendFileTransferMessage(rfbClientPtr cl, uint8_t contentType, uint8_t contentParam, uint32_t size, uint32_t length, const char *buffer)
|
|
|
|
{
|
|
|
|
rfbFileTransferMsg ft;
|
|
|
|
ft.type = rfbFileTransfer;
|
|
|
|
ft.contentType = contentType;
|
|
|
|
ft.contentParam = contentParam;
|
|
|
|
ft.pad = 0; /* UltraVNC did not Swap16LE(ft.contentParam) (Looks like it might be BigEndian) */
|
|
|
|
ft.size = Swap32IfLE(size);
|
|
|
|
ft.length = Swap32IfLE(length);
|
|
|
|
|
|
|
|
FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN("", cl, FALSE);
|
|
|
|
/*
|
|
|
|
rfbLog("rfbSendFileTransferMessage( %dtype, %dparam, %dsize, %dlen, %p)\n", contentType, contentParam, size, length, buffer);
|
|
|
|
*/
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
if (rfbWriteExact(cl, (char *)&ft, sz_rfbFileTransferMsg) < 0) {
|
|
|
|
rfbLogPerror("rfbSendFileTransferMessage: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (length>0)
|
|
|
|
{
|
|
|
|
if (rfbWriteExact(cl, buffer, length) < 0) {
|
|
|
|
rfbLogPerror("rfbSendFileTransferMessage: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
|
|
|
|
rfbStatRecordMessageSent(cl, rfbFileTransfer, sz_rfbFileTransferMsg+length, sz_rfbFileTransferMsg+length);
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* UltraVNC uses Windows Structures
|
|
|
|
*/
|
|
|
|
#define MAX_PATH 260
|
|
|
|
|
|
|
|
typedef struct {
|
|
|
|
uint32_t dwLowDateTime;
|
|
|
|
uint32_t dwHighDateTime;
|
|
|
|
} RFB_FILETIME;
|
|
|
|
|
|
|
|
typedef struct {
|
|
|
|
uint32_t dwFileAttributes;
|
|
|
|
RFB_FILETIME ftCreationTime;
|
|
|
|
RFB_FILETIME ftLastAccessTime;
|
|
|
|
RFB_FILETIME ftLastWriteTime;
|
|
|
|
uint32_t nFileSizeHigh;
|
|
|
|
uint32_t nFileSizeLow;
|
|
|
|
uint32_t dwReserved0;
|
|
|
|
uint32_t dwReserved1;
|
|
|
|
uint8_t cFileName[ MAX_PATH ];
|
|
|
|
uint8_t cAlternateFileName[ 14 ];
|
|
|
|
} RFB_FIND_DATA;
|
|
|
|
|
|
|
|
#define RFB_FILE_ATTRIBUTE_READONLY 0x1
|
|
|
|
#define RFB_FILE_ATTRIBUTE_HIDDEN 0x2
|
|
|
|
#define RFB_FILE_ATTRIBUTE_SYSTEM 0x4
|
|
|
|
#define RFB_FILE_ATTRIBUTE_DIRECTORY 0x10
|
|
|
|
#define RFB_FILE_ATTRIBUTE_ARCHIVE 0x20
|
|
|
|
#define RFB_FILE_ATTRIBUTE_NORMAL 0x80
|
|
|
|
#define RFB_FILE_ATTRIBUTE_TEMPORARY 0x100
|
|
|
|
#define RFB_FILE_ATTRIBUTE_COMPRESSED 0x800
|
|
|
|
|
|
|
|
rfbBool rfbFilenameTranslate2UNIX(rfbClientPtr cl, char *path, char *unixPath)
|
|
|
|
{
|
|
|
|
int x;
|
|
|
|
char *home=NULL;
|
|
|
|
|
|
|
|
FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN("", cl, FALSE);
|
|
|
|
|
|
|
|
/* C: */
|
|
|
|
if (path[0]=='C' && path[1]==':')
|
|
|
|
strcpy(unixPath, &path[2]);
|
|
|
|
else
|
|
|
|
{
|
|
|
|
home = getenv("HOME");
|
|
|
|
if (home!=NULL)
|
|
|
|
{
|
|
|
|
strcpy(unixPath, home);
|
|
|
|
strcat(unixPath,"/");
|
|
|
|
strcat(unixPath, path);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
strcpy(unixPath, path);
|
|
|
|
}
|
|
|
|
for (x=0;x<strlen(unixPath);x++)
|
|
|
|
if (unixPath[x]=='\\') unixPath[x]='/';
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbBool rfbFilenameTranslate2DOS(rfbClientPtr cl, char *unixPath, char *path)
|
|
|
|
{
|
|
|
|
int x;
|
|
|
|
|
|
|
|
FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN("", cl, FALSE);
|
|
|
|
|
|
|
|
sprintf(path,"C:%s", unixPath);
|
|
|
|
for (x=2;x<strlen(path);x++)
|
|
|
|
if (path[x]=='/') path[x]='\\';
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbBool rfbSendDirContent(rfbClientPtr cl, int length, char *buffer)
|
|
|
|
{
|
|
|
|
char retfilename[MAX_PATH];
|
|
|
|
char path[MAX_PATH];
|
|
|
|
struct stat statbuf;
|
|
|
|
RFB_FIND_DATA win32filename;
|
|
|
|
int nOptLen = 0, retval=0;
|
|
|
|
DIR *dirp=NULL;
|
|
|
|
struct dirent *direntp=NULL;
|
|
|
|
|
|
|
|
FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN("", cl, FALSE);
|
|
|
|
|
|
|
|
/* Client thinks we are Winblows */
|
|
|
|
rfbFilenameTranslate2UNIX(cl, buffer, path);
|
|
|
|
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbDirContentRequest: rfbRDirContent: \"%s\"->\"%s\"\n",buffer, path);
|
|
|
|
|
|
|
|
dirp=opendir(path);
|
|
|
|
if (dirp==NULL)
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbDirPacket, rfbADirectory, 0, 0, NULL);
|
|
|
|
/* send back the path name (necessary for links) */
|
|
|
|
if (rfbSendFileTransferMessage(cl, rfbDirPacket, rfbADirectory, 0, length, buffer)==FALSE) return FALSE;
|
|
|
|
for (direntp=readdir(dirp); direntp!=NULL; direntp=readdir(dirp))
|
|
|
|
{
|
|
|
|
/* get stats */
|
|
|
|
snprintf(retfilename,sizeof(retfilename),"%s/%s", path, direntp->d_name);
|
|
|
|
retval = stat(retfilename, &statbuf);
|
|
|
|
|
|
|
|
if (retval==0)
|
|
|
|
{
|
|
|
|
memset((char *)&win32filename, 0, sizeof(win32filename));
|
|
|
|
win32filename.dwFileAttributes = Swap32IfBE(RFB_FILE_ATTRIBUTE_NORMAL);
|
|
|
|
if (S_ISDIR(statbuf.st_mode))
|
|
|
|
win32filename.dwFileAttributes = Swap32IfBE(RFB_FILE_ATTRIBUTE_DIRECTORY);
|
|
|
|
win32filename.ftCreationTime.dwLowDateTime = Swap32IfBE(statbuf.st_ctime); /* Intel Order */
|
|
|
|
win32filename.ftCreationTime.dwHighDateTime = 0;
|
|
|
|
win32filename.ftLastAccessTime.dwLowDateTime = Swap32IfBE(statbuf.st_atime); /* Intel Order */
|
|
|
|
win32filename.ftLastAccessTime.dwHighDateTime = 0;
|
|
|
|
win32filename.ftLastWriteTime.dwLowDateTime = Swap32IfBE(statbuf.st_mtime); /* Intel Order */
|
|
|
|
win32filename.ftLastWriteTime.dwHighDateTime = 0;
|
|
|
|
win32filename.nFileSizeLow = Swap32IfBE(statbuf.st_size); /* Intel Order */
|
|
|
|
win32filename.nFileSizeHigh = 0;
|
|
|
|
win32filename.dwReserved0 = 0;
|
|
|
|
win32filename.dwReserved1 = 0;
|
|
|
|
|
|
|
|
/* If this had the full path, we would need to translate to DOS format ("C:\") */
|
|
|
|
/* rfbFilenameTranslate2DOS(cl, retfilename, win32filename.cFileName); */
|
|
|
|
strcpy((char *)win32filename.cFileName, direntp->d_name);
|
|
|
|
|
|
|
|
/* Do not show hidden files (but show how to move up the tree) */
|
|
|
|
if ((strcmp(direntp->d_name, "..")==0) || (direntp->d_name[0]!='.'))
|
|
|
|
{
|
|
|
|
nOptLen = sizeof(RFB_FIND_DATA) - MAX_PATH - 14 + strlen((char *)win32filename.cFileName);
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbDirContentRequest: rfbRDirContent: Sending \"%s\"\n", (char *)win32filename.cFileName);
|
|
|
|
*/
|
|
|
|
if (rfbSendFileTransferMessage(cl, rfbDirPacket, rfbADirectory, 0, nOptLen, (char *)&win32filename)==FALSE)
|
|
|
|
{
|
|
|
|
closedir(dirp);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
closedir(dirp);
|
|
|
|
/* End of the transfer */
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbDirPacket, 0, 0, 0, NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
char *rfbProcessFileTransferReadBuffer(rfbClientPtr cl, uint32_t length)
|
|
|
|
{
|
|
|
|
char *buffer=NULL;
|
|
|
|
int n=0;
|
|
|
|
|
|
|
|
FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN("", cl, NULL);
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransferReadBuffer(%dlen)\n", length);
|
|
|
|
*/
|
|
|
|
if (length>0) {
|
|
|
|
buffer=malloc(length+1);
|
|
|
|
if (buffer!=NULL) {
|
|
|
|
if ((n = rfbReadExact(cl, (char *)buffer, length)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessFileTransferReadBuffer: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
/* NOTE: don't forget to free(buffer) if you return early! */
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
/* Null Terminate */
|
|
|
|
buffer[length]=0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return buffer;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
rfbBool rfbSendFileTransferChunk(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
/* Allocate buffer for compression */
|
|
|
|
unsigned char readBuf[sz_rfbBlockSize];
|
|
|
|
int bytesRead=0;
|
|
|
|
int retval=0;
|
|
|
|
fd_set wfds;
|
|
|
|
struct timeval tv;
|
|
|
|
int n;
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
unsigned char compBuf[sz_rfbBlockSize + 1024];
|
|
|
|
unsigned long nMaxCompSize = sizeof(compBuf);
|
|
|
|
int nRetC = 0;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Don't close the client if we get into this one because
|
|
|
|
* it is called from many places to service file transfers.
|
|
|
|
* Note that permitFileTransfer is checked first.
|
|
|
|
*/
|
|
|
|
if (cl->screen->permitFileTransfer != TRUE ||
|
|
|
|
(cl->screen->getFileTransferPermission != NULL
|
|
|
|
&& cl->screen->getFileTransferPermission(cl) != TRUE)) {
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If not sending, or no file open... Return as if we sent something! */
|
|
|
|
if ((cl->fileTransfer.fd!=-1) && (cl->fileTransfer.sending==1))
|
|
|
|
{
|
|
|
|
FD_ZERO(&wfds);
|
|
|
|
FD_SET(cl->sock, &wfds);
|
|
|
|
|
|
|
|
/* return immediately */
|
|
|
|
tv.tv_sec = 0;
|
|
|
|
tv.tv_usec = 0;
|
|
|
|
n = select(cl->sock + 1, NULL, &wfds, NULL, &tv);
|
|
|
|
|
|
|
|
if (n<0) {
|
|
|
|
#ifdef WIN32
|
|
|
|
errno=WSAGetLastError();
|
|
|
|
#endif
|
|
|
|
rfbLog("rfbSendFileTransferChunk() select failed: %s\n", strerror(errno));
|
|
|
|
}
|
|
|
|
/* We have space on the transmit queue */
|
|
|
|
if (n > 0)
|
|
|
|
{
|
|
|
|
bytesRead = read(cl->fileTransfer.fd, readBuf, sz_rfbBlockSize);
|
|
|
|
switch (bytesRead) {
|
|
|
|
case 0:
|
|
|
|
/*
|
|
|
|
rfbLog("rfbSendFileTransferChunk(): End-Of-File Encountered\n");
|
|
|
|
*/
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbEndOfFile, 0, 0, 0, NULL);
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd = -1;
|
|
|
|
cl->fileTransfer.sending = 0;
|
|
|
|
cl->fileTransfer.receiving = 0;
|
|
|
|
return retval;
|
|
|
|
case -1:
|
|
|
|
/* TODO : send an error msg to the client... */
|
|
|
|
#ifdef WIN32
|
|
|
|
errno=WSAGetLastError();
|
|
|
|
#endif
|
|
|
|
rfbLog("rfbSendFileTransferChunk(): %s\n",strerror(errno));
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbAbortFileTransfer, 0, 0, 0, NULL);
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd = -1;
|
|
|
|
cl->fileTransfer.sending = 0;
|
|
|
|
cl->fileTransfer.receiving = 0;
|
|
|
|
return retval;
|
|
|
|
default:
|
|
|
|
/*
|
|
|
|
rfbLog("rfbSendFileTransferChunk(): Read %d bytes\n", bytesRead);
|
|
|
|
*/
|
|
|
|
if (!cl->fileTransfer.compressionEnabled)
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFilePacket, 0, 0, bytesRead, (char *)readBuf);
|
|
|
|
else
|
|
|
|
{
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
nRetC = compress(compBuf, &nMaxCompSize, readBuf, bytesRead);
|
|
|
|
/*
|
|
|
|
rfbLog("Compressed the packet from %d -> %d bytes\n", nMaxCompSize, bytesRead);
|
|
|
|
*/
|
|
|
|
|
|
|
|
if ((nRetC==0) && (nMaxCompSize<bytesRead))
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFilePacket, 0, 1, nMaxCompSize, (char *)compBuf);
|
|
|
|
else
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFilePacket, 0, 0, bytesRead, (char *)readBuf);
|
|
|
|
#else
|
|
|
|
/* We do not support compression of the data stream */
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFilePacket, 0, 0, bytesRead, (char *)readBuf);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbBool rfbProcessFileTransfer(rfbClientPtr cl, uint8_t contentType, uint8_t contentParam, uint32_t size, uint32_t length)
|
|
|
|
{
|
|
|
|
char *buffer=NULL, *p=NULL;
|
|
|
|
int retval=0;
|
|
|
|
char filename1[MAX_PATH];
|
|
|
|
char filename2[MAX_PATH];
|
|
|
|
char szFileTime[MAX_PATH];
|
|
|
|
struct stat statbuf;
|
|
|
|
uint32_t sizeHtmp=0;
|
|
|
|
int n=0;
|
|
|
|
char timespec[64];
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
unsigned char compBuff[sz_rfbBlockSize];
|
|
|
|
unsigned long nRawBytes = sz_rfbBlockSize;
|
|
|
|
int nRet = 0;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
FILEXFER_ALLOWED_OR_CLOSE_AND_RETURN("", cl, FALSE);
|
|
|
|
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer(%dtype, %dparam, %dsize, %dlen)\n", contentType, contentParam, size, length);
|
|
|
|
*/
|
|
|
|
|
|
|
|
switch (contentType) {
|
|
|
|
case rfbDirContentRequest:
|
|
|
|
switch (contentParam) {
|
|
|
|
case rfbRDrivesList: /* Client requests the List of Local Drives */
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbDirContentRequest: rfbRDrivesList:\n");
|
|
|
|
*/
|
|
|
|
/* Format when filled : "C:\<NULL>D:\<NULL>....Z:\<NULL><NULL>
|
|
|
|
*
|
|
|
|
* We replace the "\" char following the drive letter and ":"
|
|
|
|
* with a char corresponding to the type of drive
|
|
|
|
* We obtain something like "C:l<NULL>D:c<NULL>....Z:n\<NULL><NULL>"
|
|
|
|
* Isn't it ugly ?
|
|
|
|
* DRIVE_FIXED = 'l' (local?)
|
|
|
|
* DRIVE_REMOVABLE = 'f' (floppy?)
|
|
|
|
* DRIVE_CDROM = 'c'
|
|
|
|
* DRIVE_REMOTE = 'n'
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* in unix, there are no 'drives' (We could list mount points though)
|
|
|
|
* We fake the root as a "C:" for the Winblows users
|
|
|
|
*/
|
|
|
|
filename2[0]='C';
|
|
|
|
filename2[1]=':';
|
|
|
|
filename2[2]='l';
|
|
|
|
filename2[3]=0;
|
|
|
|
filename2[4]=0;
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbDirPacket, rfbADrivesList, 0, 5, filename2);
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return retval;
|
|
|
|
break;
|
|
|
|
case rfbRDirContent: /* Client requests the content of a directory */
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbDirContentRequest: rfbRDirContent\n");
|
|
|
|
*/
|
|
|
|
if ((buffer = rfbProcessFileTransferReadBuffer(cl, length))==NULL) return FALSE;
|
|
|
|
retval = rfbSendDirContent(cl, length, buffer);
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case rfbDirPacket:
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbDirPacket\n");
|
|
|
|
break;
|
|
|
|
case rfbFileAcceptHeader:
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileAcceptHeader\n");
|
|
|
|
break;
|
|
|
|
case rfbCommandReturn:
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbCommandReturn\n");
|
|
|
|
break;
|
|
|
|
case rfbFileChecksums:
|
|
|
|
/* Destination file already exists - the viewer sends the checksums */
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileChecksums\n");
|
|
|
|
break;
|
|
|
|
case rfbFileTransferAccess:
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileTransferAccess\n");
|
|
|
|
break;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* sending from the server to the viewer
|
|
|
|
*/
|
|
|
|
|
|
|
|
case rfbFileTransferRequest:
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileTransferRequest:\n");
|
|
|
|
*/
|
|
|
|
/* add some space to the end of the buffer as we will be adding a timespec to it */
|
|
|
|
if ((buffer = rfbProcessFileTransferReadBuffer(cl, length))==NULL) return FALSE;
|
|
|
|
/* The client requests a File */
|
|
|
|
rfbFilenameTranslate2UNIX(cl, buffer, filename1);
|
|
|
|
cl->fileTransfer.fd=open(filename1, O_RDONLY, 0744);
|
|
|
|
|
|
|
|
/*
|
|
|
|
*/
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbFileTransferRequest(\"%s\"->\"%s\") Open: %s fd=%d\n", buffer, filename1, (cl->fileTransfer.fd==-1?"Failed":"Success"), cl->fileTransfer.fd);
|
|
|
|
|
|
|
|
if (cl->fileTransfer.fd!=-1) {
|
|
|
|
if (fstat(cl->fileTransfer.fd, &statbuf)!=0) {
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd=-1;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
/* Add the File Time Stamp to the filename */
|
|
|
|
strftime(timespec, sizeof(timespec), "%m/%d/%Y %H:%M",gmtime(&statbuf.st_ctime));
|
|
|
|
buffer=realloc(buffer, length + strlen(timespec) + 2); /* comma, and Null term */
|
|
|
|
if (buffer==NULL) {
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileTransferRequest: Failed to malloc %d bytes\n", length + strlen(timespec) + 2);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
strcat(buffer,",");
|
|
|
|
strcat(buffer, timespec);
|
|
|
|
length = strlen(buffer);
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() buffer is now: \"%s\"\n", buffer);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The viewer supports compression if size==1 */
|
|
|
|
cl->fileTransfer.compressionEnabled = (size==1);
|
|
|
|
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileTransferRequest(\"%s\"->\"%s\")%s\n", buffer, filename1, (size==1?" <Compression Enabled>":""));
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* File Size in bytes, 0xFFFFFFFF (-1) means error */
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbFileHeader, 0, (cl->fileTransfer.fd==-1 ? -1 : statbuf.st_size), length, buffer);
|
|
|
|
|
|
|
|
if (cl->fileTransfer.fd==-1)
|
|
|
|
{
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
/* setup filetransfer stuff */
|
|
|
|
cl->fileTransfer.fileSize = statbuf.st_size;
|
|
|
|
cl->fileTransfer.numPackets = statbuf.st_size / sz_rfbBlockSize;
|
|
|
|
cl->fileTransfer.receiving = 0;
|
|
|
|
cl->fileTransfer.sending = 0; /* set when we receive a rfbFileHeader: */
|
|
|
|
|
|
|
|
/* TODO: finish 64-bit file size support */
|
|
|
|
sizeHtmp = 0;
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
if (rfbWriteExact(cl, (char *)&sizeHtmp, 4) < 0) {
|
|
|
|
rfbLogPerror("rfbProcessFileTransfer: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case rfbFileHeader:
|
|
|
|
/* Destination file (viewer side) is ready for reception (size > 0) or not (size = -1) */
|
|
|
|
if (size==-1) {
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileHeader (error, aborting)\n");
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd=-1;
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileHeader (%d bytes of a file)\n", size);
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* Starts the transfer! */
|
|
|
|
cl->fileTransfer.sending=1;
|
|
|
|
return rfbSendFileTransferChunk(cl);
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* sending from the viewer to the server
|
|
|
|
*/
|
|
|
|
|
|
|
|
case rfbFileTransferOffer:
|
|
|
|
/* client is sending a file to us */
|
|
|
|
/* buffer contains full path name (plus FileTime) */
|
|
|
|
/* size contains size of the file */
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFileTransferOffer:\n");
|
|
|
|
*/
|
|
|
|
if ((buffer = rfbProcessFileTransferReadBuffer(cl, length))==NULL) return FALSE;
|
|
|
|
|
|
|
|
/* Parse the FileTime */
|
|
|
|
p = strrchr(buffer, ',');
|
|
|
|
if (p!=NULL) {
|
|
|
|
*p = '\0';
|
|
|
|
strcpy(szFileTime, p+1);
|
|
|
|
} else
|
|
|
|
szFileTime[0]=0;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/* Need to read in sizeHtmp */
|
|
|
|
if ((n = rfbReadExact(cl, (char *)&sizeHtmp, 4)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessFileTransfer: read sizeHtmp");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
/* NOTE: don't forget to free(buffer) if you return early! */
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
sizeHtmp = Swap32IfLE(sizeHtmp);
|
|
|
|
|
|
|
|
rfbFilenameTranslate2UNIX(cl, buffer, filename1);
|
|
|
|
|
|
|
|
/* If the file exists... We can send a rfbFileChecksums back to the client before we send an rfbFileAcceptHeader */
|
|
|
|
/* TODO: Delta Transfer */
|
|
|
|
|
|
|
|
cl->fileTransfer.fd=open(filename1, O_CREAT|O_WRONLY|O_TRUNC, 0744);
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbFileTransferOffer(\"%s\"->\"%s\") %s %s fd=%d\n", buffer, filename1, (cl->fileTransfer.fd==-1?"Failed":"Success"), (cl->fileTransfer.fd==-1?strerror(errno):""), cl->fileTransfer.fd);
|
|
|
|
/*
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* File Size in bytes, 0xFFFFFFFF (-1) means error */
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbFileAcceptHeader, 0, (cl->fileTransfer.fd==-1 ? -1 : 0), length, buffer);
|
|
|
|
if (cl->fileTransfer.fd==-1) {
|
|
|
|
free(buffer);
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* setup filetransfer stuff */
|
|
|
|
cl->fileTransfer.fileSize = size;
|
|
|
|
cl->fileTransfer.numPackets = size / sz_rfbBlockSize;
|
|
|
|
cl->fileTransfer.receiving = 1;
|
|
|
|
cl->fileTransfer.sending = 0;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case rfbFilePacket:
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbFilePacket:\n");
|
|
|
|
*/
|
|
|
|
if ((buffer = rfbProcessFileTransferReadBuffer(cl, length))==NULL) return FALSE;
|
|
|
|
if (cl->fileTransfer.fd!=-1) {
|
|
|
|
/* buffer contains the contents of the file */
|
|
|
|
if (size==0)
|
|
|
|
retval=write(cl->fileTransfer.fd, buffer, length);
|
|
|
|
else
|
|
|
|
{
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
/* compressed packet */
|
|
|
|
nRet = uncompress(compBuff,&nRawBytes,(const unsigned char*)buffer, length);
|
|
|
|
retval=write(cl->fileTransfer.fd, (char*)compBuff, nRawBytes);
|
|
|
|
#else
|
|
|
|
/* Write the file out as received... */
|
|
|
|
retval=write(cl->fileTransfer.fd, buffer, length);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
if (retval==-1)
|
|
|
|
{
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd=-1;
|
|
|
|
cl->fileTransfer.sending = 0;
|
|
|
|
cl->fileTransfer.receiving = 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case rfbEndOfFile:
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbEndOfFile\n");
|
|
|
|
/*
|
|
|
|
*/
|
|
|
|
if (cl->fileTransfer.fd!=-1)
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd=-1;
|
|
|
|
cl->fileTransfer.sending = 0;
|
|
|
|
cl->fileTransfer.receiving = 0;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case rfbAbortFileTransfer:
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbAbortFileTransfer\n");
|
|
|
|
/*
|
|
|
|
*/
|
|
|
|
if (cl->fileTransfer.fd!=-1)
|
|
|
|
{
|
|
|
|
close(cl->fileTransfer.fd);
|
|
|
|
cl->fileTransfer.fd=-1;
|
|
|
|
cl->fileTransfer.sending = 0;
|
|
|
|
cl->fileTransfer.receiving = 0;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
/* We use this message for FileTransfer rights (<=RC18 versions)
|
|
|
|
* The client asks for FileTransfer permission
|
|
|
|
*/
|
|
|
|
if (contentParam == 0)
|
|
|
|
{
|
|
|
|
rfbLog("rfbProcessFileTransfer() File Transfer Permission DENIED! (Client Version <=RC18)\n");
|
|
|
|
/* Old method for FileTransfer handshake perimssion (<=RC18) (Deny it)*/
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbAbortFileTransfer, 0, -1, 0, "");
|
|
|
|
}
|
|
|
|
/* New method is allowed */
|
|
|
|
if (cl->screen->getFileTransferPermission!=NULL)
|
|
|
|
{
|
|
|
|
if (cl->screen->getFileTransferPermission(cl)==TRUE)
|
|
|
|
{
|
|
|
|
rfbLog("rfbProcessFileTransfer() File Transfer Permission Granted!\n");
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFileTransferAccess, 0, 1 , 0, ""); /* Permit */
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
rfbLog("rfbProcessFileTransfer() File Transfer Permission DENIED!\n");
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFileTransferAccess, 0, -1 , 0, ""); /* Deny */
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
if (cl->screen->permitFileTransfer)
|
|
|
|
{
|
|
|
|
rfbLog("rfbProcessFileTransfer() File Transfer Permission Granted!\n");
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFileTransferAccess, 0, 1 , 0, ""); /* Permit */
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
rfbLog("rfbProcessFileTransfer() File Transfer Permission DENIED by default!\n");
|
|
|
|
return rfbSendFileTransferMessage(cl, rfbFileTransferAccess, 0, -1 , 0, ""); /* DEFAULT: DENY (for security) */
|
|
|
|
}
|
|
|
|
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
|
|
case rfbCommand:
|
|
|
|
/*
|
|
|
|
rfbLog("rfbProcessFileTransfer() rfbCommand:\n");
|
|
|
|
*/
|
|
|
|
if ((buffer = rfbProcessFileTransferReadBuffer(cl, length))==NULL) return FALSE;
|
|
|
|
switch (contentParam) {
|
|
|
|
case rfbCDirCreate: /* Client requests the creation of a directory */
|
|
|
|
rfbFilenameTranslate2UNIX(cl, buffer, filename1);
|
|
|
|
retval = mkdir(filename1, 0755);
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbCommand: rfbCDirCreate(\"%s\"->\"%s\") %s\n", buffer, filename1, (retval==-1?"Failed":"Success"));
|
|
|
|
/*
|
|
|
|
*/
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbCommandReturn, rfbADirCreate, retval, length, buffer);
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return retval;
|
|
|
|
case rfbCFileDelete: /* Client requests the deletion of a file */
|
|
|
|
rfbFilenameTranslate2UNIX(cl, buffer, filename1);
|
|
|
|
if (stat(filename1,&statbuf)==0)
|
|
|
|
{
|
|
|
|
if (S_ISDIR(statbuf.st_mode))
|
|
|
|
retval = rmdir(filename1);
|
|
|
|
else
|
|
|
|
retval = unlink(filename1);
|
|
|
|
}
|
|
|
|
else retval=-1;
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbCommandReturn, rfbAFileDelete, retval, length, buffer);
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return retval;
|
|
|
|
case rfbCFileRename: /* Client requests the Renaming of a file/directory */
|
|
|
|
p = strrchr(buffer, '*');
|
|
|
|
if (p != NULL)
|
|
|
|
{
|
|
|
|
/* Split into 2 filenames ('*' is a seperator) */
|
|
|
|
*p = '\0';
|
|
|
|
rfbFilenameTranslate2UNIX(cl, buffer, filename1);
|
|
|
|
rfbFilenameTranslate2UNIX(cl, p+1, filename2);
|
|
|
|
retval = rename(filename1,filename2);
|
|
|
|
if (DB) rfbLog("rfbProcessFileTransfer() rfbCommand: rfbCFileRename(\"%s\"->\"%s\" -->> \"%s\"->\"%s\") %s\n", buffer, filename1, p+1, filename2, (retval==-1?"Failed":"Success"));
|
|
|
|
/*
|
|
|
|
*/
|
|
|
|
/* Restore the buffer so the reply is good */
|
|
|
|
*p = '*';
|
|
|
|
retval = rfbSendFileTransferMessage(cl, rfbCommandReturn, rfbAFileRename, retval, length, buffer);
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* NOTE: don't forget to free(buffer) if you return early! */
|
|
|
|
if (buffer!=NULL) free(buffer);
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbProcessClientNormalMessage is called when the client has sent a normal
|
|
|
|
* protocol message.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void
|
|
|
|
rfbProcessClientNormalMessage(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
int n=0;
|
|
|
|
rfbClientToServerMsg msg;
|
|
|
|
char *str;
|
|
|
|
int i;
|
|
|
|
uint32_t enc=0;
|
|
|
|
uint32_t lastPreferredEncoding = -1;
|
|
|
|
char encBuf[64];
|
|
|
|
char encBuf2[64];
|
|
|
|
|
|
|
|
#ifdef LIBVNCSERVER_WITH_WEBSOCKETS
|
|
|
|
if (cl->wsctx && webSocketCheckDisconnect(cl))
|
|
|
|
return;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, (char *)&msg, 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (msg.type) {
|
|
|
|
|
|
|
|
case rfbSetPixelFormat:
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbSetPixelFormatMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
cl->format.bitsPerPixel = msg.spf.format.bitsPerPixel;
|
|
|
|
cl->format.depth = msg.spf.format.depth;
|
|
|
|
cl->format.bigEndian = (msg.spf.format.bigEndian ? TRUE : FALSE);
|
|
|
|
cl->format.trueColour = (msg.spf.format.trueColour ? TRUE : FALSE);
|
|
|
|
cl->format.redMax = Swap16IfLE(msg.spf.format.redMax);
|
|
|
|
cl->format.greenMax = Swap16IfLE(msg.spf.format.greenMax);
|
|
|
|
cl->format.blueMax = Swap16IfLE(msg.spf.format.blueMax);
|
|
|
|
cl->format.redShift = msg.spf.format.redShift;
|
|
|
|
cl->format.greenShift = msg.spf.format.greenShift;
|
|
|
|
cl->format.blueShift = msg.spf.format.blueShift;
|
|
|
|
|
|
|
|
cl->readyForSetColourMapEntries = TRUE;
|
|
|
|
cl->screen->setTranslateFunction(cl);
|
|
|
|
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetPixelFormatMsg, sz_rfbSetPixelFormatMsg);
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
|
|
case rfbFixColourMapEntries:
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbFixColourMapEntriesMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetPixelFormatMsg, sz_rfbSetPixelFormatMsg);
|
|
|
|
rfbLog("rfbProcessClientNormalMessage: %s",
|
|
|
|
"FixColourMapEntries unsupported\n");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
|
|
/* NOTE: Some clients send us a set of encodings (ie: PointerPos) designed to enable/disable features...
|
|
|
|
* We may want to look into this...
|
|
|
|
* Example:
|
|
|
|
* case rfbEncodingXCursor:
|
|
|
|
* cl->enableCursorShapeUpdates = TRUE;
|
|
|
|
*
|
|
|
|
* Currently: cl->enableCursorShapeUpdates can *never* be turned off...
|
|
|
|
*/
|
|
|
|
case rfbSetEncodings:
|
|
|
|
{
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbSetEncodingsMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
msg.se.nEncodings = Swap16IfLE(msg.se.nEncodings);
|
|
|
|
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetEncodingsMsg+(msg.se.nEncodings*4),sz_rfbSetEncodingsMsg+(msg.se.nEncodings*4));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* UltraVNC Client has the ability to adapt to changing network environments
|
|
|
|
* So, let's give it a change to tell us what it wants now!
|
|
|
|
*/
|
|
|
|
if (cl->preferredEncoding!=-1)
|
|
|
|
lastPreferredEncoding = cl->preferredEncoding;
|
|
|
|
|
|
|
|
/* Reset all flags to defaults (allows us to switch between PointerPos and Server Drawn Cursors) */
|
|
|
|
cl->preferredEncoding=-1;
|
|
|
|
cl->useCopyRect = FALSE;
|
|
|
|
cl->useNewFBSize = FALSE;
|
|
|
|
cl->cursorWasChanged = FALSE;
|
|
|
|
cl->useRichCursorEncoding = FALSE;
|
|
|
|
cl->enableCursorPosUpdates = FALSE;
|
|
|
|
cl->enableCursorShapeUpdates = FALSE;
|
|
|
|
cl->enableCursorShapeUpdates = FALSE;
|
|
|
|
cl->enableLastRectEncoding = FALSE;
|
|
|
|
cl->enableKeyboardLedState = FALSE;
|
|
|
|
cl->enableSupportedMessages = FALSE;
|
|
|
|
cl->enableSupportedEncodings = FALSE;
|
|
|
|
cl->enableServerIdentity = FALSE;
|
|
|
|
#if defined(LIBVNCSERVER_HAVE_LIBZ) || defined(LIBVNCSERVER_HAVE_LIBPNG)
|
|
|
|
cl->tightQualityLevel = -1;
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
#if defined(LIBVNCSERVER_HAVE_LIBJPEG) || defined(LIBVNCSERVER_HAVE_LIBPNG)
|
|
|
|
cl->tightCompressLevel = TIGHT_DEFAULT_COMPRESSION;
|
|
|
|
#endif
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
cl->turboSubsampLevel = TURBO_DEFAULT_SUBSAMP;
|
|
|
|
cl->turboQualityLevel = -1;
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < msg.se.nEncodings; i++) {
|
|
|
|
if ((n = rfbReadExact(cl, (char *)&enc, 4)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
enc = Swap32IfLE(enc);
|
|
|
|
|
|
|
|
switch (enc) {
|
|
|
|
|
|
|
|
case rfbEncodingCopyRect:
|
|
|
|
cl->useCopyRect = TRUE;
|
|
|
|
break;
|
|
|
|
case rfbEncodingRaw:
|
|
|
|
case rfbEncodingRRE:
|
|
|
|
case rfbEncodingCoRRE:
|
|
|
|
case rfbEncodingHextile:
|
|
|
|
case rfbEncodingUltra:
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
case rfbEncodingZlib:
|
|
|
|
case rfbEncodingZRLE:
|
|
|
|
case rfbEncodingZYWRLE:
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
case rfbEncodingTight:
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPNG
|
|
|
|
case rfbEncodingTightPng:
|
|
|
|
#endif
|
|
|
|
/* The first supported encoding is the 'preferred' encoding */
|
|
|
|
if (cl->preferredEncoding == -1)
|
|
|
|
cl->preferredEncoding = enc;
|
|
|
|
|
|
|
|
|
|
|
|
break;
|
|
|
|
case rfbEncodingXCursor:
|
|
|
|
if(!cl->screen->dontConvertRichCursorToXCursor) {
|
|
|
|
rfbLog("Enabling X-style cursor updates for client %s\n",
|
|
|
|
cl->host);
|
|
|
|
/* if cursor was drawn, hide the cursor */
|
|
|
|
if(!cl->enableCursorShapeUpdates)
|
|
|
|
rfbRedrawAfterHideCursor(cl,NULL);
|
|
|
|
|
|
|
|
cl->enableCursorShapeUpdates = TRUE;
|
|
|
|
cl->cursorWasChanged = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingRichCursor:
|
|
|
|
rfbLog("Enabling full-color cursor updates for client %s\n",
|
|
|
|
cl->host);
|
|
|
|
/* if cursor was drawn, hide the cursor */
|
|
|
|
if(!cl->enableCursorShapeUpdates)
|
|
|
|
rfbRedrawAfterHideCursor(cl,NULL);
|
|
|
|
|
|
|
|
cl->enableCursorShapeUpdates = TRUE;
|
|
|
|
cl->useRichCursorEncoding = TRUE;
|
|
|
|
cl->cursorWasChanged = TRUE;
|
|
|
|
break;
|
|
|
|
case rfbEncodingPointerPos:
|
|
|
|
if (!cl->enableCursorPosUpdates) {
|
|
|
|
rfbLog("Enabling cursor position updates for client %s\n",
|
|
|
|
cl->host);
|
|
|
|
cl->enableCursorPosUpdates = TRUE;
|
|
|
|
cl->cursorWasMoved = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingLastRect:
|
|
|
|
if (!cl->enableLastRectEncoding) {
|
|
|
|
rfbLog("Enabling LastRect protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
cl->enableLastRectEncoding = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingNewFBSize:
|
|
|
|
if (!cl->useNewFBSize) {
|
|
|
|
rfbLog("Enabling NewFBSize protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
cl->useNewFBSize = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingKeyboardLedState:
|
|
|
|
if (!cl->enableKeyboardLedState) {
|
|
|
|
rfbLog("Enabling KeyboardLedState protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
cl->enableKeyboardLedState = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingSupportedMessages:
|
|
|
|
if (!cl->enableSupportedMessages) {
|
|
|
|
rfbLog("Enabling SupportedMessages protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
cl->enableSupportedMessages = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingSupportedEncodings:
|
|
|
|
if (!cl->enableSupportedEncodings) {
|
|
|
|
rfbLog("Enabling SupportedEncodings protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
cl->enableSupportedEncodings = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingServerIdentity:
|
|
|
|
if (!cl->enableServerIdentity) {
|
|
|
|
rfbLog("Enabling ServerIdentity protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
cl->enableServerIdentity = TRUE;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case rfbEncodingXvp:
|
|
|
|
rfbLog("Enabling Xvp protocol extension for client "
|
|
|
|
"%s\n", cl->host);
|
|
|
|
if (!rfbSendXvp(cl, 1, rfbXvp_Init)) {
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
#if defined(LIBVNCSERVER_HAVE_LIBZ) || defined(LIBVNCSERVER_HAVE_LIBPNG)
|
|
|
|
if ( enc >= (uint32_t)rfbEncodingCompressLevel0 &&
|
|
|
|
enc <= (uint32_t)rfbEncodingCompressLevel9 ) {
|
|
|
|
cl->zlibCompressLevel = enc & 0x0F;
|
|
|
|
#if defined(LIBVNCSERVER_HAVE_LIBJPEG) || defined(LIBVNCSERVER_HAVE_LIBPNG)
|
|
|
|
cl->tightCompressLevel = enc & 0x0F;
|
|
|
|
rfbLog("Using compression level %d for client %s\n",
|
|
|
|
cl->tightCompressLevel, cl->host);
|
|
|
|
#endif
|
|
|
|
} else if ( enc >= (uint32_t)rfbEncodingQualityLevel0 &&
|
|
|
|
enc <= (uint32_t)rfbEncodingQualityLevel9 ) {
|
|
|
|
cl->tightQualityLevel = enc & 0x0F;
|
|
|
|
rfbLog("Using image quality level %d for client %s\n",
|
|
|
|
cl->tightQualityLevel, cl->host);
|
Replace TightVNC encoder with TurboVNC encoder. This patch is the result of further research and discussion that revealed the following:
-- TightPng encoding and the rfbTightNoZlib extension need not conflict. Since
TightPng is a separate encoding type, not supported by TurboVNC-compatible
viewers, then the rfbTightNoZlib extension can be used solely whenever the
encoding type is Tight and disabled with the encoding type is TightPng.
-- In the TightVNC encoder, compression levels above 5 are basically useless.
On the set of 20 low-level datasets that were used to design the TurboVNC
encoder (these include the eight 2D application captures that were also used
when designing the TightVNC encoder, as well as 12 3D application captures
provided by the VirtualGL Project--
see http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf), moving
from Compression Level (CL) 5 to CL 9 in the TightVNC encoder did not
increase the compression ratio of any datasets more than 10%, and the
compression ratio only increased by more than 5% on four of them. The
compression ratio actually decreased a few percent on five of them. In
exchange for this paltry increase in compression ratio, the CPU usage, on
average, went up by a factor of 5. Thus, for all intents and purposes,
TightVNC CL 5 provides the "best useful compression" for that encoder.
-- TurboVNC's best compression level (CL 2) compresses 3D and video workloads
significantly more "tightly" than TightVNC CL 5 (~70% better, in the
aggregate) but does not quite achieve the same level of compression with 2D
workloads (~20% worse, in the aggregate.) This decrease in compression ratio
may or may not be noticeable, since many of the datasets it affects are not
performance-critical (such as the console output of a compilation, etc.)
However, for peace of mind, it was still desirable to have a mode that
compressed with equal "tightness" to TightVNC CL 5, since we proposed to
replace that encoder entirely.
-- A new mode was discovered in the TurboVNC encoder that produces, in the
aggregate, similar compression ratios on 2D datasets as TightVNC CL 5. That
new mode involves using Zlib level 7 (the same level used by TightVNC CL 5)
but setting the "palette threshold" to 256, so that indexed color encoding
is used whenever possible. This mode reduces bandwidth only marginally
(typically 10-20%) relative to TurboVNC CL 2 on low-color workloads, in
exchange for nearly doubling CPU usage, and it does not benefit high-color
workloads at all (since those are usually encoded with JPEG.) However, it
provides a means of reproducing the same "tightness" as the TightVNC
encoder on 2D workloads without sacrificing any compression for 3D/video
workloads, and without using any more CPU time than necessary.
-- The TurboVNC encoder still performs as well or better than the TightVNC
encoder when plain libjpeg is used instead of libjpeg-turbo.
Specific notes follow:
common/turbojpeg.c common/turbojpeg.h:
Added code to emulate the libjpeg-turbo colorspace extensions, so that the
TurboJPEG wrapper can be used with plain libjpeg as well. This required
updating the TurboJPEG wrapper to the latest code from libjpeg-turbo 1.2.0,
mainly because the TurboJPEG 1.2 API handles pixel formats in a much cleaner
way, which made the conversion code easier to write. It also eases the
maintenance to have the wrapper synced as much as possible with the upstream
code base (so I can merge any relevant bug fixes that are discovered upstream.)
The libvncserver version of the TurboJPEG wrapper is a "lite" version,
containing only the JPEG compression/decompression code and not the lossless
transform, YUV encoding/decoding, and dynamic buffer allocation features from
TurboJPEG 1.2.
configure.ac:
Removed the --with-turbovnc option. configure still checks for the presence of
libjpeg-turbo, but only for the purposes of printing a performance warning if
it isn't available.
rfb/rfb.h:
Fix a bug introduced with the initial TurboVNC encoder patch. We cannot use
tightQualityLevel for the TurboVNC 1-100 quality level, because
tightQualityLevel is also used by ZRLE. Thus, a new parameter
(turboQualityLevel) was created.
rfb/rfbproto.h:
Remove TurboVNC-specific #ifdefs and language
libvncserver/rfbserver.c:
Remove TurboVNC-specific #ifdefs. Fix afore-mentioned tightQualityLevel bug.
libvncserver/tight.c:
Replaced the TightVNC encoder with the TurboVNC encoder. Relative to the
initial TurboVNC encoder patch, this patch also:
-- Adds TightPng support to the TurboVNC encoder
-- Adds the afore-mentioned low-bandwidth mode, which is mapped externally to
Compression Level 9
test/*:
Included TJUnitTest (a regression test for the TurboJPEG wrapper) as well as
TJBench (a benchmark for same.) These are useful for ensuring that the wrapper
still functions correctly and performantly if it needs to be modified for
whatever reason. Both of these programs are derived from libjpeg-turbo 1.2.0.
As with the TurboJPEG wrapper, they do not contain the more advanced features
of TurboJPEG 1.2, such as YUV encoding/decoding and lossless transforms.
13 years ago
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
cl->turboQualityLevel = tight2turbo_qual[enc & 0x0F];
|
|
|
|
cl->turboSubsampLevel = tight2turbo_subsamp[enc & 0x0F];
|
|
|
|
rfbLog("Using JPEG subsampling %d, Q%d for client %s\n",
|
|
|
|
cl->turboSubsampLevel, cl->turboQualityLevel, cl->host);
|
|
|
|
} else if ( enc >= (uint32_t)rfbEncodingFineQualityLevel0 + 1 &&
|
|
|
|
enc <= (uint32_t)rfbEncodingFineQualityLevel100 ) {
|
|
|
|
cl->turboQualityLevel = enc & 0xFF;
|
|
|
|
rfbLog("Using fine quality level %d for client %s\n",
|
|
|
|
cl->turboQualityLevel, cl->host);
|
|
|
|
} else if ( enc >= (uint32_t)rfbEncodingSubsamp1X &&
|
|
|
|
enc <= (uint32_t)rfbEncodingSubsampGray ) {
|
|
|
|
cl->turboSubsampLevel = enc & 0xFF;
|
|
|
|
rfbLog("Using subsampling level %d for client %s\n",
|
|
|
|
cl->turboSubsampLevel, cl->host);
|
Add TurboVNC encoding support.
TurboVNC is a variant of TightVNC that uses the same client/server protocol (RFB version 3.8t),
and thus it is fully cross-compatible with TightVNC and TigerVNC (with one exception, which is noted below.)
Both the TightVNC and TurboVNC encoders analyze each rectangle, pick out regions of solid color to send
separately, and send the remaining subrectangles using mono, indexed color, JPEG, or raw encoding, depending
on the number of colors in the subrectangle. However, TurboVNC uses a fundamentally different selection
algorithm to determine the appropriate subencoding to use for each subrectangle. Thus, while it sends a
protocol stream that can be decoded by any TightVNC-compatible viewer, the mix of subencoding types in this
protocol stream will be different from those generated by a TightVNC server.
The research that led to TurboVNC is described in the following report:
http://www.virtualgl.org/pmwiki/uploads/About/tighttoturbo.pdf.
In summary: 20 RFB captures, representing "common" 2D and 3D application workloads (the 3D workloads were
run using VirtualGL), were studied using the TightVNC encoder in isolation. Some of the analysis features
in the TightVNC encoder, such as smoothness detection, were found to generate a lot of CPU usage with little
or no benefit in compression, so those features were disabled. JPEG encoding was accelerated using
libjpeg-turbo (which achieves a 2-4x speedup over plain libjpeg on modern x86 or ARM processors.) Finally,
the "palette threshold" (minimum number of colors that the subrectangle must have before it is compressed
using JPEG or raw) was adjusted to account for the fact that JPEG encoding is now quite a bit faster
(meaning that we can now use it more without a CPU penalty.) TurboVNC has additional optimizations,
such as the ability to count colors and encode JPEG images directly from the framebuffer without first
translating the pixels into RGB. The TurboVNC encoder compares quite favorably in terms of compression
ratio with TightVNC and generally encodes a great deal faster (often an order of magnitude or more.)
The version of the TurboVNC encoder included in this patch is roughly equivalent to the one found in version
0.6 of the Unix TurboVNC Server, with a few minor patches integrated from TurboVNC 1.1. TurboVNC 1.0
added multi-threading capabilities, which can be added in later if desired (at the expense of making
libvncserver depend on libpthread.)
Because TurboVNC uses a fundamentally different mix of subencodings than TightVNC, because it uses
the identical protocol (and thus a viewer really has no idea whether it's talking to a TightVNC or
TurboVNC server), and because it doesn't support rfbTightPng (and in fact conflicts with it-- see below),
the TurboVNC and TightVNC encoders cannot be enabled simultaneously.
Compatibility:
In *most* cases, a TurboVNC-enabled viewer is fully compatible with a TightVNC server, and vice versa.
TurboVNC supports pseudo-encodings for specifying a fine-grained (1-100) quality scale and specifying
chrominance subsampling. If a TurboVNC viewer sends those to a TightVNC server, then the TightVNC server
ignores them, so the TurboVNC viewer also sends the quality on a 0-9 scale that the TightVNC server can
understand. Similarly, the TurboVNC server checks first for fine-grained quality and subsampling
pseudo-encodings from the viewer, and failing to receive those, it then checks for the TightVNC 0-9
quality pseudo-encoding.
There is one case in which the two systems are not compatible, and that is when a TightVNC or TigerVNC
viewer requests compression level 0 without JPEG from a TurboVNC server. For performance reasons,
this causes the TurboVNC server to send images directly to the viewer, bypassing Zlib. When the
TurboVNC server does this, it also sets bits 7-4 in the compression control byte to rfbTightNoZlib (0x0A),
which is unfortunately the same value as rfbTightPng. Older TightVNC viewers that don't handle PNG
will assume that the stream is uncompressed but still encapsulated in a Zlib structure, whereas newer
PNG-supporting TightVNC viewers will assume that the stream is PNG. In either case, the viewer will
probably crash. Since most VNC viewers don't expose compression level 0 in the GUI, this is a
relatively rare situation.
Description of changes:
configure.ac
-- Added support for libjpeg-turbo. If passed an argument of --with-turbovnc, configure will now run
(or, if cross-compiling, just link) a test program that determines whether the libjpeg library being
used is libjpeg-turbo. libjpeg-turbo must be used when building the TurboVNC encoder, because the
TurboVNC encoder relies on the libjpeg-turbo colorspace extensions in order to compress images directly
out of the framebuffer (which may be, for instance, BGRA rather than RGB.) libjpeg-turbo can optionally
be used with the TightVNC encoder as well, but the speedup will only be marginal (the report linked
above explains why in more detail, but basically it's because of Amdahl's Law. The TightVNC encoder
was designed with the assumption that JPEG had a very high CPU cost, and thus JPEG is used only sparingly.)
-- Added a new configure variable, JPEG_LDFLAGS. This is necessitated by the fact that libjpeg-turbo
often distributes libjpeg.a and libjpeg.so in /opt/libjpeg-turbo/lib32 or /opt/libjpeg-turbo/lib64,
and many people prefer to statically link with it. Thus, more flexibility is needed than is provided
by --with-jpeg. If JPEG_LDFLAGS is specified, then it overrides the changes to LDFLAGS enacted by
--with-jpeg (but --with-jpeg is still used to set the include path.) The addition of JPEG_LDFLAGS
necessitated replacing AC_CHECK_LIB with AC_LINK_IFELSE (because AC_CHECK_LIB automatically sets
LIBS to -ljpeg, which is not what we want if we're, for instance, linking statically with libjpeg-turbo.)
-- configure does not check for PNG support if TurboVNC encoding is enabled. This prevents the
rfbSendRectEncodingTightPng() function from being compiled in, since the TurboVNC encoder doesn't
(and can't) support it.
common/turbojpeg.c, common/turbojpeg.h
-- TurboJPEG is a simple API used to compress and decompress JPEG images in memory. It was originally
implemented because it was desirable to use different types of underlying technologies to compress
JPEG on different platforms (mediaLib on SPARC, Quicktime on PPC Macs, Intel Performance Primitives, etc.)
These days, however, libjpeg-turbo is the only underlying technology used by TurboVNC, so TurboJPEG's
purpose is largely just code simplicity and flexibility. Thus, since there is no real need for
libvncserver to use any technology other than libjpeg-turbo for compressing JPEG, the TurboJPEG wrapper
for libjpeg-turbo has been included in-tree so that libvncserver can be directly linked with libjpeg-turbo.
This is convenient because many modern Linux distros (Fedora, Ubuntu, etc.) now ship libjpeg-turbo as
their default libjpeg library.
libvncserver/rfbserver.c
-- Added logic to check for the TurboVNC fine-grained quality level and subsampling encodings and to
map Tight (0-9) quality levels to appropriate fine-grained quality level and subsampling values if
communicating with a TightVNC/TigerVNC viewer.
libvncserver/turbo.c
-- TurboVNC encoder (compiled instead of libvncserver/tight.c)
rfb/rfb.h
-- Added support for the TurboVNC subsampling level
rfb/rfbproto.h
-- Added constants for the TurboVNC fine quality level and subsampling encodings as well as the rfbTightNoZlib
constant and notes on its usage.
13 years ago
|
|
|
#endif
|
|
|
|
} else
|
|
|
|
#endif
|
|
|
|
{
|
|
|
|
rfbExtensionData* e;
|
|
|
|
for(e = cl->extensions; e;) {
|
|
|
|
rfbExtensionData* next = e->next;
|
|
|
|
if(e->extension->enablePseudoEncoding &&
|
|
|
|
e->extension->enablePseudoEncoding(cl,
|
The PseudoEncoding extension code was getting silly:
If the client asked for an encoding, and no enabled extension handled it,
LibVNCServer would walk through all extensions, and if they promised to handle
the encoding, execute the extension's newClient() if it was not NULL.
However, if newClient is not NULL, it will be called when a client connects,
and if it returns TRUE, the extension will be enabled. Since all the state of
the extension should be in the client data, there is no good reason why
newClient should return FALSE the first time (thus not enabling the extension),
but TRUE when called just before calling enablePseudoEncoding().
So in effect, the extension got enabled all the time, even if that was not
necessary.
The resolution is to pass a void** to enablePseudoEncoding. This has the
further advantage that enablePseudoEncoding can remalloc() or free() the
data without problems. Though keep in mind that if enablePseudoEncoding()
is called on a not-yet-enabled extension, the passed data points to NULL.
19 years ago
|
|
|
&e->data, (int)enc))
|
|
|
|
/* ext handles this encoding */
|
|
|
|
break;
|
|
|
|
e = next;
|
|
|
|
}
|
|
|
|
if(e == NULL) {
|
|
|
|
rfbBool handled = FALSE;
|
|
|
|
/* if the pseudo encoding is not handled by the
|
|
|
|
enabled extensions, search through all
|
|
|
|
extensions. */
|
|
|
|
rfbProtocolExtension* e;
|
|
|
|
|
|
|
|
for(e = rfbGetExtensionIterator(); e;) {
|
|
|
|
int* encs = e->pseudoEncodings;
|
|
|
|
while(encs && *encs!=0) {
|
|
|
|
if(*encs==(int)enc) {
|
|
|
|
void* data = NULL;
|
The PseudoEncoding extension code was getting silly:
If the client asked for an encoding, and no enabled extension handled it,
LibVNCServer would walk through all extensions, and if they promised to handle
the encoding, execute the extension's newClient() if it was not NULL.
However, if newClient is not NULL, it will be called when a client connects,
and if it returns TRUE, the extension will be enabled. Since all the state of
the extension should be in the client data, there is no good reason why
newClient should return FALSE the first time (thus not enabling the extension),
but TRUE when called just before calling enablePseudoEncoding().
So in effect, the extension got enabled all the time, even if that was not
necessary.
The resolution is to pass a void** to enablePseudoEncoding. This has the
further advantage that enablePseudoEncoding can remalloc() or free() the
data without problems. Though keep in mind that if enablePseudoEncoding()
is called on a not-yet-enabled extension, the passed data points to NULL.
19 years ago
|
|
|
if(!e->enablePseudoEncoding(cl, &data, (int)enc)) {
|
|
|
|
rfbLog("Installed extension pretends to handle pseudo encoding 0x%x, but does not!\n",(int)enc);
|
|
|
|
} else {
|
|
|
|
rfbEnableExtension(cl, e, data);
|
|
|
|
handled = TRUE;
|
|
|
|
e = NULL;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
encs++;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(e)
|
|
|
|
e = e->next;
|
|
|
|
}
|
|
|
|
rfbReleaseExtensionIterator();
|
|
|
|
|
|
|
|
if(!handled)
|
|
|
|
rfbLog("rfbProcessClientNormalMessage: "
|
|
|
|
"ignoring unsupported encoding type %s\n",
|
|
|
|
encodingName(enc,encBuf,sizeof(encBuf)));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
if (cl->preferredEncoding == -1) {
|
|
|
|
if (lastPreferredEncoding==-1) {
|
|
|
|
cl->preferredEncoding = rfbEncodingRaw;
|
|
|
|
rfbLog("Defaulting to %s encoding for client %s\n", encodingName(cl->preferredEncoding,encBuf,sizeof(encBuf)),cl->host);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
cl->preferredEncoding = lastPreferredEncoding;
|
|
|
|
rfbLog("Sticking with %s encoding for client %s\n", encodingName(cl->preferredEncoding,encBuf,sizeof(encBuf)),cl->host);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
if (lastPreferredEncoding==-1) {
|
|
|
|
rfbLog("Using %s encoding for client %s\n", encodingName(cl->preferredEncoding,encBuf,sizeof(encBuf)),cl->host);
|
|
|
|
} else {
|
|
|
|
rfbLog("Switching from %s to %s Encoding for client %s\n",
|
|
|
|
encodingName(lastPreferredEncoding,encBuf2,sizeof(encBuf2)),
|
|
|
|
encodingName(cl->preferredEncoding,encBuf,sizeof(encBuf)), cl->host);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (cl->enableCursorPosUpdates && !cl->enableCursorShapeUpdates) {
|
|
|
|
rfbLog("Disabling cursor position updates for client %s\n",
|
|
|
|
cl->host);
|
|
|
|
cl->enableCursorPosUpdates = FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
case rfbFramebufferUpdateRequest:
|
|
|
|
{
|
|
|
|
sraRegionPtr tmpRegion;
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbFramebufferUpdateRequestMsg-1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbFramebufferUpdateRequestMsg,sz_rfbFramebufferUpdateRequestMsg);
|
|
|
|
|
|
|
|
/* The values come in based on the scaled screen, we need to convert them to
|
|
|
|
* values based on the main screen's coordinate system
|
|
|
|
*/
|
|
|
|
if(!rectSwapIfLEAndClip(&msg.fur.x,&msg.fur.y,&msg.fur.w,&msg.fur.h,cl))
|
|
|
|
{
|
|
|
|
rfbLog("Warning, ignoring rfbFramebufferUpdateRequest: %dXx%dY-%dWx%dH\n",msg.fur.x, msg.fur.y, msg.fur.w, msg.fur.h);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
tmpRegion =
|
|
|
|
sraRgnCreateRect(msg.fur.x,
|
|
|
|
msg.fur.y,
|
|
|
|
msg.fur.x+msg.fur.w,
|
|
|
|
msg.fur.y+msg.fur.h);
|
|
|
|
|
|
|
|
LOCK(cl->updateMutex);
|
|
|
|
sraRgnOr(cl->requestedRegion,tmpRegion);
|
|
|
|
|
|
|
|
if (!cl->readyForSetColourMapEntries) {
|
|
|
|
/* client hasn't sent a SetPixelFormat so is using server's */
|
|
|
|
cl->readyForSetColourMapEntries = TRUE;
|
|
|
|
if (!cl->format.trueColour) {
|
|
|
|
if (!rfbSetClientColourMap(cl, 0, 0)) {
|
|
|
|
sraRgnDestroy(tmpRegion);
|
|
|
|
TSIGNAL(cl->updateCond);
|
|
|
|
UNLOCK(cl->updateMutex);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!msg.fur.incremental) {
|
|
|
|
sraRgnOr(cl->modifiedRegion,tmpRegion);
|
|
|
|
sraRgnSubtract(cl->copyRegion,tmpRegion);
|
|
|
|
}
|
|
|
|
TSIGNAL(cl->updateCond);
|
|
|
|
UNLOCK(cl->updateMutex);
|
|
|
|
|
|
|
|
sraRgnDestroy(tmpRegion);
|
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
case rfbKeyEvent:
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbKeyEventMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbKeyEventMsg, sz_rfbKeyEventMsg);
|
|
|
|
|
|
|
|
if(!cl->viewOnly) {
|
|
|
|
cl->screen->kbdAddEvent(msg.ke.down, (rfbKeySym)Swap32IfLE(msg.ke.key), cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
|
|
case rfbPointerEvent:
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbPointerEventMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbPointerEventMsg, sz_rfbPointerEventMsg);
|
|
|
|
|
|
|
|
if (cl->screen->pointerClient && cl->screen->pointerClient != cl)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (msg.pe.buttonMask == 0)
|
|
|
|
cl->screen->pointerClient = NULL;
|
|
|
|
else
|
|
|
|
cl->screen->pointerClient = cl;
|
|
|
|
|
|
|
|
if(!cl->viewOnly) {
|
|
|
|
if (msg.pe.buttonMask != cl->lastPtrButtons ||
|
|
|
|
cl->screen->deferPtrUpdateTime == 0) {
|
|
|
|
cl->screen->ptrAddEvent(msg.pe.buttonMask,
|
|
|
|
ScaleX(cl->scaledScreen, cl->screen, Swap16IfLE(msg.pe.x)),
|
|
|
|
ScaleY(cl->scaledScreen, cl->screen, Swap16IfLE(msg.pe.y)),
|
|
|
|
cl);
|
|
|
|
cl->lastPtrButtons = msg.pe.buttonMask;
|
|
|
|
} else {
|
|
|
|
cl->lastPtrX = ScaleX(cl->scaledScreen, cl->screen, Swap16IfLE(msg.pe.x));
|
|
|
|
cl->lastPtrY = ScaleY(cl->scaledScreen, cl->screen, Swap16IfLE(msg.pe.y));
|
|
|
|
cl->lastPtrButtons = msg.pe.buttonMask;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
|
|
case rfbFileTransfer:
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbFileTransferMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
msg.ft.size = Swap32IfLE(msg.ft.size);
|
|
|
|
msg.ft.length = Swap32IfLE(msg.ft.length);
|
|
|
|
/* record statistics in rfbProcessFileTransfer as length is filled with garbage when it is not valid */
|
|
|
|
rfbProcessFileTransfer(cl, msg.ft.contentType, msg.ft.contentParam, msg.ft.size, msg.ft.length);
|
|
|
|
return;
|
|
|
|
|
|
|
|
case rfbSetSW:
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbSetSWMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
msg.sw.x = Swap16IfLE(msg.sw.x);
|
|
|
|
msg.sw.y = Swap16IfLE(msg.sw.y);
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetSWMsg, sz_rfbSetSWMsg);
|
|
|
|
/* msg.sw.status is not initialized in the ultraVNC viewer and contains random numbers (why???) */
|
|
|
|
|
|
|
|
rfbLog("Received a rfbSetSingleWindow(%d x, %d y)\n", msg.sw.x, msg.sw.y);
|
|
|
|
if (cl->screen->setSingleWindow!=NULL)
|
|
|
|
cl->screen->setSingleWindow(cl, msg.sw.x, msg.sw.y);
|
|
|
|
return;
|
|
|
|
|
|
|
|
case rfbSetServerInput:
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbSetServerInputMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetServerInputMsg, sz_rfbSetServerInputMsg);
|
|
|
|
|
|
|
|
/* msg.sim.pad is not initialized in the ultraVNC viewer and contains random numbers (why???) */
|
|
|
|
/* msg.sim.pad = Swap16IfLE(msg.sim.pad); */
|
|
|
|
|
|
|
|
rfbLog("Received a rfbSetServerInput(%d status)\n", msg.sim.status);
|
|
|
|
if (cl->screen->setServerInput!=NULL)
|
|
|
|
cl->screen->setServerInput(cl, msg.sim.status);
|
|
|
|
return;
|
|
|
|
|
|
|
|
case rfbTextChat:
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbTextChatMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
msg.tc.pad2 = Swap16IfLE(msg.tc.pad2);
|
|
|
|
msg.tc.length = Swap32IfLE(msg.tc.length);
|
|
|
|
|
|
|
|
switch (msg.tc.length) {
|
|
|
|
case rfbTextChatOpen:
|
|
|
|
case rfbTextChatClose:
|
|
|
|
case rfbTextChatFinished:
|
|
|
|
/* commands do not have text following */
|
|
|
|
/* Why couldn't they have used the pad byte??? */
|
|
|
|
str=NULL;
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbTextChatMsg, sz_rfbTextChatMsg);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
if ((msg.tc.length>0) && (msg.tc.length<rfbTextMaxSize))
|
|
|
|
{
|
|
|
|
str = (char *)malloc(msg.tc.length);
|
|
|
|
if (str==NULL)
|
|
|
|
{
|
|
|
|
rfbLog("Unable to malloc %d bytes for a TextChat Message\n", msg.tc.length);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if ((n = rfbReadExact(cl, str, msg.tc.length)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
free(str);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbTextChatMsg+msg.tc.length, sz_rfbTextChatMsg+msg.tc.length);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
/* This should never happen */
|
|
|
|
rfbLog("client sent us a Text Message that is too big %d>%d\n", msg.tc.length, rfbTextMaxSize);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Note: length can be commands: rfbTextChatOpen, rfbTextChatClose, and rfbTextChatFinished
|
|
|
|
* at which point, the str is NULL (as it is not sent)
|
|
|
|
*/
|
|
|
|
if (cl->screen->setTextChat!=NULL)
|
|
|
|
cl->screen->setTextChat(cl, msg.tc.length, str);
|
|
|
|
|
|
|
|
free(str);
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
|
|
case rfbClientCutText:
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbClientCutTextMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
msg.cct.length = Swap32IfLE(msg.cct.length);
|
|
|
|
|
|
|
|
str = (char *)malloc(msg.cct.length);
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, str, msg.cct.length)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
free(str);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbClientCutTextMsg+msg.cct.length, sz_rfbClientCutTextMsg+msg.cct.length);
|
|
|
|
if(!cl->viewOnly) {
|
|
|
|
cl->screen->setXCutText(str, msg.cct.length, cl);
|
|
|
|
}
|
|
|
|
free(str);
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
case rfbPalmVNCSetScaleFactor:
|
|
|
|
cl->PalmVNC = TRUE;
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbSetScaleMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetScaleMsg, sz_rfbSetScaleMsg);
|
|
|
|
rfbLog("rfbSetScale(%d)\n", msg.ssc.scale);
|
|
|
|
rfbScalingSetup(cl,cl->screen->width/msg.ssc.scale, cl->screen->height/msg.ssc.scale);
|
|
|
|
|
|
|
|
rfbSendNewScaleSize(cl);
|
|
|
|
return;
|
|
|
|
|
|
|
|
case rfbSetScale:
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbSetScaleMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbSetScaleMsg, sz_rfbSetScaleMsg);
|
|
|
|
rfbLog("rfbSetScale(%d)\n", msg.ssc.scale);
|
|
|
|
rfbScalingSetup(cl,cl->screen->width/msg.ssc.scale, cl->screen->height/msg.ssc.scale);
|
|
|
|
|
|
|
|
rfbSendNewScaleSize(cl);
|
|
|
|
return;
|
|
|
|
|
|
|
|
case rfbXvp:
|
|
|
|
|
|
|
|
if ((n = rfbReadExact(cl, ((char *)&msg) + 1,
|
|
|
|
sz_rfbXvpMsg - 1)) <= 0) {
|
|
|
|
if (n != 0)
|
|
|
|
rfbLogPerror("rfbProcessClientNormalMessage: read");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, sz_rfbXvpMsg, sz_rfbXvpMsg);
|
|
|
|
|
|
|
|
/* only version when is defined, so echo back a fail */
|
|
|
|
if(msg.xvp.version != 1) {
|
|
|
|
rfbSendXvp(cl, msg.xvp.version, rfbXvp_Fail);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* if the hook exists and fails, send a fail msg */
|
|
|
|
if(cl->screen->xvpHook && !cl->screen->xvpHook(cl, msg.xvp.version, msg.xvp.code))
|
|
|
|
rfbSendXvp(cl, 1, rfbXvp_Fail);
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
|
|
|
|
default:
|
|
|
|
{
|
|
|
|
rfbExtensionData *e,*next;
|
|
|
|
|
|
|
|
for(e=cl->extensions; e;) {
|
|
|
|
next = e->next;
|
|
|
|
if(e->extension->handleMessage &&
|
|
|
|
e->extension->handleMessage(cl, e->data, &msg))
|
|
|
|
{
|
|
|
|
rfbStatRecordMessageRcvd(cl, msg.type, 0, 0); /* Extension should handle this */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
e = next;
|
|
|
|
}
|
|
|
|
|
|
|
|
rfbLog("rfbProcessClientNormalMessage: unknown message type %d\n",
|
|
|
|
msg.type);
|
|
|
|
rfbLog(" ... closing connection\n");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbSendFramebufferUpdate - send the currently pending framebuffer update to
|
|
|
|
* the RFB client.
|
|
|
|
* givenUpdateRegion is not changed.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendFramebufferUpdate(rfbClientPtr cl,
|
|
|
|
sraRegionPtr givenUpdateRegion)
|
|
|
|
{
|
|
|
|
sraRectangleIterator* i=NULL;
|
|
|
|
sraRect rect;
|
|
|
|
int nUpdateRegionRects;
|
|
|
|
rfbFramebufferUpdateMsg *fu = (rfbFramebufferUpdateMsg *)cl->updateBuf;
|
|
|
|
sraRegionPtr updateRegion,updateCopyRegion,tmpRegion;
|
|
|
|
int dx, dy;
|
|
|
|
rfbBool sendCursorShape = FALSE;
|
|
|
|
rfbBool sendCursorPos = FALSE;
|
|
|
|
rfbBool sendKeyboardLedState = FALSE;
|
|
|
|
rfbBool sendSupportedMessages = FALSE;
|
|
|
|
rfbBool sendSupportedEncodings = FALSE;
|
|
|
|
rfbBool sendServerIdentity = FALSE;
|
|
|
|
rfbBool result = TRUE;
|
|
|
|
|
|
|
|
|
|
|
|
if(cl->screen->displayHook)
|
|
|
|
cl->screen->displayHook(cl);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If framebuffer size was changed and the client supports NewFBSize
|
|
|
|
* encoding, just send NewFBSize marker and return.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (cl->useNewFBSize && cl->newFBSizePending) {
|
|
|
|
LOCK(cl->updateMutex);
|
|
|
|
cl->newFBSizePending = FALSE;
|
|
|
|
UNLOCK(cl->updateMutex);
|
|
|
|
fu->type = rfbFramebufferUpdate;
|
|
|
|
fu->nRects = Swap16IfLE(1);
|
|
|
|
cl->ublen = sz_rfbFramebufferUpdateMsg;
|
|
|
|
if (!rfbSendNewFBSize(cl, cl->scaledScreen->width, cl->scaledScreen->height)) {
|
|
|
|
if(cl->screen->displayFinishedHook)
|
|
|
|
cl->screen->displayFinishedHook(cl, FALSE);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
result = rfbSendUpdateBuf(cl);
|
|
|
|
if(cl->screen->displayFinishedHook)
|
|
|
|
cl->screen->displayFinishedHook(cl, result);
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If this client understands cursor shape updates, cursor should be
|
|
|
|
* removed from the framebuffer. Otherwise, make sure it's put up.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (cl->enableCursorShapeUpdates) {
|
|
|
|
if (cl->cursorWasChanged && cl->readyForSetColourMapEntries)
|
|
|
|
sendCursorShape = TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Do we plan to send cursor position update?
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (cl->enableCursorPosUpdates && cl->cursorWasMoved)
|
|
|
|
sendCursorPos = TRUE;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Do we plan to send a keyboard state update?
|
|
|
|
*/
|
|
|
|
if ((cl->enableKeyboardLedState) &&
|
|
|
|
(cl->screen->getKeyboardLedStateHook!=NULL))
|
|
|
|
{
|
|
|
|
int x;
|
|
|
|
x=cl->screen->getKeyboardLedStateHook(cl->screen);
|
|
|
|
if (x!=cl->lastKeyboardLedState)
|
|
|
|
{
|
|
|
|
sendKeyboardLedState = TRUE;
|
|
|
|
cl->lastKeyboardLedState=x;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Do we plan to send a rfbEncodingSupportedMessages?
|
|
|
|
*/
|
|
|
|
if (cl->enableSupportedMessages)
|
|
|
|
{
|
|
|
|
sendSupportedMessages = TRUE;
|
|
|
|
/* We only send this message ONCE <per setEncodings message received>
|
|
|
|
* (We disable it here)
|
|
|
|
*/
|
|
|
|
cl->enableSupportedMessages = FALSE;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Do we plan to send a rfbEncodingSupportedEncodings?
|
|
|
|
*/
|
|
|
|
if (cl->enableSupportedEncodings)
|
|
|
|
{
|
|
|
|
sendSupportedEncodings = TRUE;
|
|
|
|
/* We only send this message ONCE <per setEncodings message received>
|
|
|
|
* (We disable it here)
|
|
|
|
*/
|
|
|
|
cl->enableSupportedEncodings = FALSE;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Do we plan to send a rfbEncodingServerIdentity?
|
|
|
|
*/
|
|
|
|
if (cl->enableServerIdentity)
|
|
|
|
{
|
|
|
|
sendServerIdentity = TRUE;
|
|
|
|
/* We only send this message ONCE <per setEncodings message received>
|
|
|
|
* (We disable it here)
|
|
|
|
*/
|
|
|
|
cl->enableServerIdentity = FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
LOCK(cl->updateMutex);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The modifiedRegion may overlap the destination copyRegion. We remove
|
|
|
|
* any overlapping bits from the copyRegion (since they'd only be
|
|
|
|
* overwritten anyway).
|
|
|
|
*/
|
|
|
|
|
|
|
|
sraRgnSubtract(cl->copyRegion,cl->modifiedRegion);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The client is interested in the region requestedRegion. The region
|
|
|
|
* which should be updated now is the intersection of requestedRegion
|
|
|
|
* and the union of modifiedRegion and copyRegion. If it's empty then
|
|
|
|
* no update is needed.
|
|
|
|
*/
|
|
|
|
|
|
|
|
updateRegion = sraRgnCreateRgn(givenUpdateRegion);
|
|
|
|
if(cl->screen->progressiveSliceHeight>0) {
|
|
|
|
int height=cl->screen->progressiveSliceHeight,
|
|
|
|
y=cl->progressiveSliceY;
|
|
|
|
sraRegionPtr bbox=sraRgnBBox(updateRegion);
|
|
|
|
sraRect rect;
|
|
|
|
if(sraRgnPopRect(bbox,&rect,0)) {
|
|
|
|
sraRegionPtr slice;
|
|
|
|
if(y<rect.y1 || y>=rect.y2)
|
|
|
|
y=rect.y1;
|
|
|
|
slice=sraRgnCreateRect(0,y,cl->screen->width,y+height);
|
|
|
|
sraRgnAnd(updateRegion,slice);
|
|
|
|
sraRgnDestroy(slice);
|
|
|
|
}
|
|
|
|
sraRgnDestroy(bbox);
|
|
|
|
y+=height;
|
|
|
|
if(y>=cl->screen->height)
|
|
|
|
y=0;
|
|
|
|
cl->progressiveSliceY=y;
|
|
|
|
}
|
|
|
|
|
|
|
|
sraRgnOr(updateRegion,cl->copyRegion);
|
|
|
|
if(!sraRgnAnd(updateRegion,cl->requestedRegion) &&
|
|
|
|
sraRgnEmpty(updateRegion) &&
|
|
|
|
(cl->enableCursorShapeUpdates ||
|
|
|
|
(cl->cursorX == cl->screen->cursorX && cl->cursorY == cl->screen->cursorY)) &&
|
|
|
|
!sendCursorShape && !sendCursorPos && !sendKeyboardLedState &&
|
|
|
|
!sendSupportedMessages && !sendSupportedEncodings && !sendServerIdentity) {
|
|
|
|
sraRgnDestroy(updateRegion);
|
|
|
|
UNLOCK(cl->updateMutex);
|
|
|
|
if(cl->screen->displayFinishedHook)
|
|
|
|
cl->screen->displayFinishedHook(cl, TRUE);
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We assume that the client doesn't have any pixel data outside the
|
|
|
|
* requestedRegion. In other words, both the source and destination of a
|
|
|
|
* copy must lie within requestedRegion. So the region we can send as a
|
|
|
|
* copy is the intersection of the copyRegion with both the requestedRegion
|
|
|
|
* and the requestedRegion translated by the amount of the copy. We set
|
|
|
|
* updateCopyRegion to this.
|
|
|
|
*/
|
|
|
|
|
|
|
|
updateCopyRegion = sraRgnCreateRgn(cl->copyRegion);
|
|
|
|
sraRgnAnd(updateCopyRegion,cl->requestedRegion);
|
|
|
|
tmpRegion = sraRgnCreateRgn(cl->requestedRegion);
|
|
|
|
sraRgnOffset(tmpRegion,cl->copyDX,cl->copyDY);
|
|
|
|
sraRgnAnd(updateCopyRegion,tmpRegion);
|
|
|
|
sraRgnDestroy(tmpRegion);
|
|
|
|
dx = cl->copyDX;
|
|
|
|
dy = cl->copyDY;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Next we remove updateCopyRegion from updateRegion so that updateRegion
|
|
|
|
* is the part of this update which is sent as ordinary pixel data (i.e not
|
|
|
|
* a copy).
|
|
|
|
*/
|
|
|
|
|
|
|
|
sraRgnSubtract(updateRegion,updateCopyRegion);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Finally we leave modifiedRegion to be the remainder (if any) of parts of
|
|
|
|
* the screen which are modified but outside the requestedRegion. We also
|
|
|
|
* empty both the requestedRegion and the copyRegion - note that we never
|
|
|
|
* carry over a copyRegion for a future update.
|
|
|
|
*/
|
|
|
|
|
|
|
|
sraRgnOr(cl->modifiedRegion,cl->copyRegion);
|
|
|
|
sraRgnSubtract(cl->modifiedRegion,updateRegion);
|
|
|
|
sraRgnSubtract(cl->modifiedRegion,updateCopyRegion);
|
|
|
|
|
|
|
|
sraRgnMakeEmpty(cl->requestedRegion);
|
|
|
|
sraRgnMakeEmpty(cl->copyRegion);
|
|
|
|
cl->copyDX = 0;
|
|
|
|
cl->copyDY = 0;
|
|
|
|
|
|
|
|
UNLOCK(cl->updateMutex);
|
|
|
|
|
|
|
|
if (!cl->enableCursorShapeUpdates) {
|
|
|
|
if(cl->cursorX != cl->screen->cursorX || cl->cursorY != cl->screen->cursorY) {
|
|
|
|
rfbRedrawAfterHideCursor(cl,updateRegion);
|
|
|
|
LOCK(cl->screen->cursorMutex);
|
|
|
|
cl->cursorX = cl->screen->cursorX;
|
|
|
|
cl->cursorY = cl->screen->cursorY;
|
|
|
|
UNLOCK(cl->screen->cursorMutex);
|
|
|
|
rfbRedrawAfterHideCursor(cl,updateRegion);
|
|
|
|
}
|
|
|
|
rfbShowCursor(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Now send the update.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbStatRecordMessageSent(cl, rfbFramebufferUpdate, 0, 0);
|
|
|
|
if (cl->preferredEncoding == rfbEncodingCoRRE) {
|
|
|
|
nUpdateRegionRects = 0;
|
|
|
|
|
|
|
|
for(i = sraRgnGetIterator(updateRegion); sraRgnIteratorNext(i,&rect);){
|
|
|
|
int x = rect.x1;
|
|
|
|
int y = rect.y1;
|
|
|
|
int w = rect.x2 - x;
|
|
|
|
int h = rect.y2 - y;
|
|
|
|
int rectsPerRow, rows;
|
|
|
|
/* We need to count the number of rects in the scaled screen */
|
|
|
|
if (cl->screen!=cl->scaledScreen)
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "rfbSendFramebufferUpdate");
|
|
|
|
rectsPerRow = (w-1)/cl->correMaxWidth+1;
|
|
|
|
rows = (h-1)/cl->correMaxHeight+1;
|
|
|
|
nUpdateRegionRects += rectsPerRow*rows;
|
|
|
|
}
|
|
|
|
sraRgnReleaseIterator(i); i=NULL;
|
|
|
|
} else if (cl->preferredEncoding == rfbEncodingUltra) {
|
|
|
|
nUpdateRegionRects = 0;
|
|
|
|
|
|
|
|
for(i = sraRgnGetIterator(updateRegion); sraRgnIteratorNext(i,&rect);){
|
|
|
|
int x = rect.x1;
|
|
|
|
int y = rect.y1;
|
|
|
|
int w = rect.x2 - x;
|
|
|
|
int h = rect.y2 - y;
|
|
|
|
/* We need to count the number of rects in the scaled screen */
|
|
|
|
if (cl->screen!=cl->scaledScreen)
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "rfbSendFramebufferUpdate");
|
|
|
|
nUpdateRegionRects += (((h-1) / (ULTRA_MAX_SIZE( w ) / w)) + 1);
|
|
|
|
}
|
|
|
|
sraRgnReleaseIterator(i); i=NULL;
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
} else if (cl->preferredEncoding == rfbEncodingZlib) {
|
|
|
|
nUpdateRegionRects = 0;
|
|
|
|
|
|
|
|
for(i = sraRgnGetIterator(updateRegion); sraRgnIteratorNext(i,&rect);){
|
|
|
|
int x = rect.x1;
|
|
|
|
int y = rect.y1;
|
|
|
|
int w = rect.x2 - x;
|
|
|
|
int h = rect.y2 - y;
|
|
|
|
/* We need to count the number of rects in the scaled screen */
|
|
|
|
if (cl->screen!=cl->scaledScreen)
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "rfbSendFramebufferUpdate");
|
|
|
|
nUpdateRegionRects += (((h-1) / (ZLIB_MAX_SIZE( w ) / w)) + 1);
|
|
|
|
}
|
|
|
|
sraRgnReleaseIterator(i); i=NULL;
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
} else if (cl->preferredEncoding == rfbEncodingTight) {
|
|
|
|
nUpdateRegionRects = 0;
|
|
|
|
|
|
|
|
for(i = sraRgnGetIterator(updateRegion); sraRgnIteratorNext(i,&rect);){
|
|
|
|
int x = rect.x1;
|
|
|
|
int y = rect.y1;
|
|
|
|
int w = rect.x2 - x;
|
|
|
|
int h = rect.y2 - y;
|
|
|
|
int n;
|
|
|
|
/* We need to count the number of rects in the scaled screen */
|
|
|
|
if (cl->screen!=cl->scaledScreen)
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "rfbSendFramebufferUpdate");
|
|
|
|
n = rfbNumCodedRectsTight(cl, x, y, w, h);
|
|
|
|
if (n == 0) {
|
|
|
|
nUpdateRegionRects = 0xFFFF;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
nUpdateRegionRects += n;
|
|
|
|
}
|
|
|
|
sraRgnReleaseIterator(i); i=NULL;
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPNG
|
|
|
|
} else if (cl->preferredEncoding == rfbEncodingTightPng) {
|
|
|
|
nUpdateRegionRects = 0;
|
|
|
|
|
|
|
|
for(i = sraRgnGetIterator(updateRegion); sraRgnIteratorNext(i,&rect);){
|
|
|
|
int x = rect.x1;
|
|
|
|
int y = rect.y1;
|
|
|
|
int w = rect.x2 - x;
|
|
|
|
int h = rect.y2 - y;
|
|
|
|
int n;
|
|
|
|
/* We need to count the number of rects in the scaled screen */
|
|
|
|
if (cl->screen!=cl->scaledScreen)
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "rfbSendFramebufferUpdate");
|
|
|
|
n = rfbNumCodedRectsTight(cl, x, y, w, h);
|
|
|
|
if (n == 0) {
|
|
|
|
nUpdateRegionRects = 0xFFFF;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
nUpdateRegionRects += n;
|
|
|
|
}
|
|
|
|
sraRgnReleaseIterator(i); i=NULL;
|
|
|
|
#endif
|
|
|
|
} else {
|
|
|
|
nUpdateRegionRects = sraRgnCountRects(updateRegion);
|
|
|
|
}
|
|
|
|
|
|
|
|
fu->type = rfbFramebufferUpdate;
|
|
|
|
if (nUpdateRegionRects != 0xFFFF) {
|
|
|
|
if(cl->screen->maxRectsPerUpdate>0
|
|
|
|
/* CoRRE splits the screen into smaller squares */
|
|
|
|
&& cl->preferredEncoding != rfbEncodingCoRRE
|
|
|
|
/* Ultra encoding splits rectangles up into smaller chunks */
|
|
|
|
&& cl->preferredEncoding != rfbEncodingUltra
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
/* Zlib encoding splits rectangles up into smaller chunks */
|
|
|
|
&& cl->preferredEncoding != rfbEncodingZlib
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
/* Tight encoding counts the rectangles differently */
|
|
|
|
&& cl->preferredEncoding != rfbEncodingTight
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPNG
|
|
|
|
/* Tight encoding counts the rectangles differently */
|
|
|
|
&& cl->preferredEncoding != rfbEncodingTightPng
|
|
|
|
#endif
|
|
|
|
&& nUpdateRegionRects>cl->screen->maxRectsPerUpdate) {
|
|
|
|
sraRegion* newUpdateRegion = sraRgnBBox(updateRegion);
|
|
|
|
sraRgnDestroy(updateRegion);
|
|
|
|
updateRegion = newUpdateRegion;
|
|
|
|
nUpdateRegionRects = sraRgnCountRects(updateRegion);
|
|
|
|
}
|
|
|
|
fu->nRects = Swap16IfLE((uint16_t)(sraRgnCountRects(updateCopyRegion) +
|
|
|
|
nUpdateRegionRects +
|
|
|
|
!!sendCursorShape + !!sendCursorPos + !!sendKeyboardLedState +
|
|
|
|
!!sendSupportedMessages + !!sendSupportedEncodings + !!sendServerIdentity));
|
|
|
|
} else {
|
|
|
|
fu->nRects = 0xFFFF;
|
|
|
|
}
|
|
|
|
cl->ublen = sz_rfbFramebufferUpdateMsg;
|
|
|
|
|
|
|
|
if (sendCursorShape) {
|
|
|
|
cl->cursorWasChanged = FALSE;
|
|
|
|
if (!rfbSendCursorShape(cl))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (sendCursorPos) {
|
|
|
|
cl->cursorWasMoved = FALSE;
|
|
|
|
if (!rfbSendCursorPos(cl))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (sendKeyboardLedState) {
|
|
|
|
if (!rfbSendKeyboardLedState(cl))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (sendSupportedMessages) {
|
|
|
|
if (!rfbSendSupportedMessages(cl))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
if (sendSupportedEncodings) {
|
|
|
|
if (!rfbSendSupportedEncodings(cl))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
if (sendServerIdentity) {
|
|
|
|
if (!rfbSendServerIdentity(cl))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!sraRgnEmpty(updateCopyRegion)) {
|
|
|
|
if (!rfbSendCopyRegion(cl,updateCopyRegion,dx,dy))
|
|
|
|
goto updateFailed;
|
|
|
|
}
|
|
|
|
|
|
|
|
for(i = sraRgnGetIterator(updateRegion); sraRgnIteratorNext(i,&rect);){
|
|
|
|
int x = rect.x1;
|
|
|
|
int y = rect.y1;
|
|
|
|
int w = rect.x2 - x;
|
|
|
|
int h = rect.y2 - y;
|
|
|
|
|
|
|
|
/* We need to count the number of rects in the scaled screen */
|
|
|
|
if (cl->screen!=cl->scaledScreen)
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "rfbSendFramebufferUpdate");
|
|
|
|
|
|
|
|
switch (cl->preferredEncoding) {
|
|
|
|
case -1:
|
|
|
|
case rfbEncodingRaw:
|
|
|
|
if (!rfbSendRectEncodingRaw(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
case rfbEncodingRRE:
|
|
|
|
if (!rfbSendRectEncodingRRE(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
case rfbEncodingCoRRE:
|
|
|
|
if (!rfbSendRectEncodingCoRRE(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
case rfbEncodingHextile:
|
|
|
|
if (!rfbSendRectEncodingHextile(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
case rfbEncodingUltra:
|
|
|
|
if (!rfbSendRectEncodingUltra(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
case rfbEncodingZlib:
|
|
|
|
if (!rfbSendRectEncodingZlib(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBJPEG
|
|
|
|
case rfbEncodingTight:
|
|
|
|
if (!rfbSendRectEncodingTight(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBPNG
|
|
|
|
case rfbEncodingTightPng:
|
|
|
|
if (!rfbSendRectEncodingTightPng(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
#endif
|
|
|
|
#ifdef LIBVNCSERVER_HAVE_LIBZ
|
|
|
|
case rfbEncodingZRLE:
|
|
|
|
case rfbEncodingZYWRLE:
|
|
|
|
if (!rfbSendRectEncodingZRLE(cl, x, y, w, h))
|
|
|
|
goto updateFailed;
|
|
|
|
break;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (i) {
|
|
|
|
sraRgnReleaseIterator(i);
|
|
|
|
i = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
if ( nUpdateRegionRects == 0xFFFF &&
|
|
|
|
!rfbSendLastRectMarker(cl) )
|
|
|
|
goto updateFailed;
|
|
|
|
|
|
|
|
if (!rfbSendUpdateBuf(cl)) {
|
|
|
|
updateFailed:
|
|
|
|
result = FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!cl->enableCursorShapeUpdates) {
|
|
|
|
rfbHideCursor(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(i)
|
|
|
|
sraRgnReleaseIterator(i);
|
|
|
|
sraRgnDestroy(updateRegion);
|
|
|
|
sraRgnDestroy(updateCopyRegion);
|
|
|
|
|
|
|
|
if(cl->screen->displayFinishedHook)
|
|
|
|
cl->screen->displayFinishedHook(cl, result);
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send the copy region as a string of CopyRect encoded rectangles.
|
|
|
|
* The only slightly tricky thing is that we should send the messages in
|
|
|
|
* the correct order so that an earlier CopyRect will not corrupt the source
|
|
|
|
* of a later one.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendCopyRegion(rfbClientPtr cl,
|
|
|
|
sraRegionPtr reg,
|
|
|
|
int dx,
|
|
|
|
int dy)
|
|
|
|
{
|
|
|
|
int x, y, w, h;
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
rfbCopyRect cr;
|
|
|
|
sraRectangleIterator* i;
|
|
|
|
sraRect rect1;
|
|
|
|
|
|
|
|
/* printf("copyrect: "); sraRgnPrint(reg); putchar('\n');fflush(stdout); */
|
|
|
|
i = sraRgnGetReverseIterator(reg,dx>0,dy>0);
|
|
|
|
|
|
|
|
/* correct for the scale of the screen */
|
|
|
|
dx = ScaleX(cl->screen, cl->scaledScreen, dx);
|
|
|
|
dy = ScaleX(cl->screen, cl->scaledScreen, dy);
|
|
|
|
|
|
|
|
while(sraRgnIteratorNext(i,&rect1)) {
|
|
|
|
x = rect1.x1;
|
|
|
|
y = rect1.y1;
|
|
|
|
w = rect1.x2 - x;
|
|
|
|
h = rect1.y2 - y;
|
|
|
|
|
|
|
|
/* correct for scaling (if necessary) */
|
|
|
|
rfbScaledCorrection(cl->screen, cl->scaledScreen, &x, &y, &w, &h, "copyrect");
|
|
|
|
|
|
|
|
rect.r.x = Swap16IfLE(x);
|
|
|
|
rect.r.y = Swap16IfLE(y);
|
|
|
|
rect.r.w = Swap16IfLE(w);
|
|
|
|
rect.r.h = Swap16IfLE(h);
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingCopyRect);
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
cr.srcX = Swap16IfLE(x - dx);
|
|
|
|
cr.srcY = Swap16IfLE(y - dy);
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&cr, sz_rfbCopyRect);
|
|
|
|
cl->ublen += sz_rfbCopyRect;
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingCopyRect, sz_rfbFramebufferUpdateRectHeader + sz_rfbCopyRect,
|
|
|
|
w * h * (cl->scaledScreen->bitsPerPixel / 8));
|
|
|
|
}
|
|
|
|
sraRgnReleaseIterator(i);
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send a given rectangle in raw encoding (rfbEncodingRaw).
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendRectEncodingRaw(rfbClientPtr cl,
|
|
|
|
int x,
|
|
|
|
int y,
|
|
|
|
int w,
|
|
|
|
int h)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
int nlines;
|
|
|
|
int bytesPerLine = w * (cl->format.bitsPerPixel / 8);
|
|
|
|
char *fbptr = (cl->scaledScreen->frameBuffer + (cl->scaledScreen->paddedWidthInBytes * y)
|
|
|
|
+ (x * (cl->scaledScreen->bitsPerPixel / 8)));
|
|
|
|
|
|
|
|
/* Flush the buffer to guarantee correct alignment for translateFn(). */
|
|
|
|
if (cl->ublen > 0) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rect.r.x = Swap16IfLE(x);
|
|
|
|
rect.r.y = Swap16IfLE(y);
|
|
|
|
rect.r.w = Swap16IfLE(w);
|
|
|
|
rect.r.h = Swap16IfLE(h);
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingRaw);
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingRaw, sz_rfbFramebufferUpdateRectHeader + bytesPerLine * h,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader + bytesPerLine * h);
|
|
|
|
|
|
|
|
nlines = (UPDATE_BUF_SIZE - cl->ublen) / bytesPerLine;
|
|
|
|
|
|
|
|
while (TRUE) {
|
|
|
|
if (nlines > h)
|
|
|
|
nlines = h;
|
|
|
|
|
|
|
|
(*cl->translateFn)(cl->translateLookupTable,
|
|
|
|
&(cl->screen->serverFormat),
|
|
|
|
&cl->format, fbptr, &cl->updateBuf[cl->ublen],
|
|
|
|
cl->scaledScreen->paddedWidthInBytes, w, nlines);
|
|
|
|
|
|
|
|
cl->ublen += nlines * bytesPerLine;
|
|
|
|
h -= nlines;
|
|
|
|
|
|
|
|
if (h == 0) /* rect fitted in buffer, do next one */
|
|
|
|
return TRUE;
|
|
|
|
|
|
|
|
/* buffer full - flush partial rect and do another nlines */
|
|
|
|
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
fbptr += (cl->scaledScreen->paddedWidthInBytes * nlines);
|
|
|
|
|
|
|
|
nlines = (UPDATE_BUF_SIZE - cl->ublen) / bytesPerLine;
|
|
|
|
if (nlines == 0) {
|
|
|
|
rfbErr("rfbSendRectEncodingRaw: send buffer too small for %d "
|
|
|
|
"bytes per line\n", bytesPerLine);
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send an empty rectangle with encoding field set to value of
|
|
|
|
* rfbEncodingLastRect to notify client that this is the last
|
|
|
|
* rectangle in framebuffer update ("LastRect" extension of RFB
|
|
|
|
* protocol).
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendLastRectMarker(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbFramebufferUpdateRectHeader > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingLastRect);
|
|
|
|
rect.r.x = 0;
|
|
|
|
rect.r.y = 0;
|
|
|
|
rect.r.w = 0;
|
|
|
|
rect.r.h = 0;
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingLastRect, sz_rfbFramebufferUpdateRectHeader, sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send NewFBSize pseudo-rectangle. This tells the client to change
|
|
|
|
* its framebuffer size.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendNewFBSize(rfbClientPtr cl,
|
|
|
|
int w,
|
|
|
|
int h)
|
|
|
|
{
|
|
|
|
rfbFramebufferUpdateRectHeader rect;
|
|
|
|
|
|
|
|
if (cl->ublen + sz_rfbFramebufferUpdateRectHeader > UPDATE_BUF_SIZE) {
|
|
|
|
if (!rfbSendUpdateBuf(cl))
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (cl->PalmVNC==TRUE)
|
|
|
|
rfbLog("Sending rfbEncodingNewFBSize in response to a PalmVNC style framebuffer resize (%dx%d)\n", w, h);
|
|
|
|
else
|
|
|
|
rfbLog("Sending rfbEncodingNewFBSize for resize to (%dx%d)\n", w, h);
|
|
|
|
|
|
|
|
rect.encoding = Swap32IfLE(rfbEncodingNewFBSize);
|
|
|
|
rect.r.x = 0;
|
|
|
|
rect.r.y = 0;
|
|
|
|
rect.r.w = Swap16IfLE(w);
|
|
|
|
rect.r.h = Swap16IfLE(h);
|
|
|
|
|
|
|
|
memcpy(&cl->updateBuf[cl->ublen], (char *)&rect,
|
|
|
|
sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
cl->ublen += sz_rfbFramebufferUpdateRectHeader;
|
|
|
|
|
|
|
|
rfbStatRecordEncodingSent(cl, rfbEncodingNewFBSize, sz_rfbFramebufferUpdateRectHeader, sz_rfbFramebufferUpdateRectHeader);
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Send the contents of cl->updateBuf. Returns 1 if successful, -1 if
|
|
|
|
* not (errno should be set).
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendUpdateBuf(rfbClientPtr cl)
|
|
|
|
{
|
|
|
|
if(cl->sock<0)
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
if (rfbWriteExact(cl, cl->updateBuf, cl->ublen) < 0) {
|
|
|
|
rfbLogPerror("rfbSendUpdateBuf: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
cl->ublen = 0;
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbSendSetColourMapEntries sends a SetColourMapEntries message to the
|
|
|
|
* client, using values from the currently installed colormap.
|
|
|
|
*/
|
|
|
|
|
|
|
|
rfbBool
|
|
|
|
rfbSendSetColourMapEntries(rfbClientPtr cl,
|
|
|
|
int firstColour,
|
|
|
|
int nColours)
|
|
|
|
{
|
|
|
|
char buf[sz_rfbSetColourMapEntriesMsg + 256 * 3 * 2];
|
|
|
|
char *wbuf = buf;
|
|
|
|
rfbSetColourMapEntriesMsg *scme;
|
|
|
|
uint16_t *rgb;
|
|
|
|
rfbColourMap* cm = &cl->screen->colourMap;
|
|
|
|
int i, len;
|
|
|
|
|
|
|
|
if (nColours > 256) {
|
|
|
|
/* some rare hardware has, e.g., 4096 colors cells: PseudoColor:12 */
|
|
|
|
wbuf = (char *) malloc(sz_rfbSetColourMapEntriesMsg + nColours * 3 * 2);
|
|
|
|
}
|
|
|
|
|
|
|
|
scme = (rfbSetColourMapEntriesMsg *)wbuf;
|
|
|
|
rgb = (uint16_t *)(&wbuf[sz_rfbSetColourMapEntriesMsg]);
|
|
|
|
|
|
|
|
scme->type = rfbSetColourMapEntries;
|
|
|
|
|
|
|
|
scme->firstColour = Swap16IfLE(firstColour);
|
|
|
|
scme->nColours = Swap16IfLE(nColours);
|
|
|
|
|
|
|
|
len = sz_rfbSetColourMapEntriesMsg;
|
|
|
|
|
|
|
|
for (i = 0; i < nColours; i++) {
|
|
|
|
if(i<(int)cm->count) {
|
|
|
|
if(cm->is16) {
|
|
|
|
rgb[i*3] = Swap16IfLE(cm->data.shorts[i*3]);
|
|
|
|
rgb[i*3+1] = Swap16IfLE(cm->data.shorts[i*3+1]);
|
|
|
|
rgb[i*3+2] = Swap16IfLE(cm->data.shorts[i*3+2]);
|
|
|
|
} else {
|
|
|
|
rgb[i*3] = Swap16IfLE((unsigned short)cm->data.bytes[i*3]);
|
|
|
|
rgb[i*3+1] = Swap16IfLE((unsigned short)cm->data.bytes[i*3+1]);
|
|
|
|
rgb[i*3+2] = Swap16IfLE((unsigned short)cm->data.bytes[i*3+2]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
len += nColours * 3 * 2;
|
|
|
|
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
if (rfbWriteExact(cl, wbuf, len) < 0) {
|
|
|
|
rfbLogPerror("rfbSendSetColourMapEntries: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
if (wbuf != buf) free(wbuf);
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
|
|
|
|
rfbStatRecordMessageSent(cl, rfbSetColourMapEntries, len, len);
|
|
|
|
if (wbuf != buf) free(wbuf);
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbSendBell sends a Bell message to all the clients.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbSendBell(rfbScreenInfoPtr rfbScreen)
|
|
|
|
{
|
|
|
|
rfbClientIteratorPtr i;
|
|
|
|
rfbClientPtr cl;
|
|
|
|
rfbBellMsg b;
|
|
|
|
|
|
|
|
i = rfbGetClientIterator(rfbScreen);
|
|
|
|
while((cl=rfbClientIteratorNext(i))) {
|
|
|
|
b.type = rfbBell;
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
if (rfbWriteExact(cl, (char *)&b, sz_rfbBellMsg) < 0) {
|
|
|
|
rfbLogPerror("rfbSendBell: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
}
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
}
|
|
|
|
rfbStatRecordMessageSent(cl, rfbBell, sz_rfbBellMsg, sz_rfbBellMsg);
|
|
|
|
rfbReleaseClientIterator(i);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* rfbSendServerCutText sends a ServerCutText message to all the clients.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbSendServerCutText(rfbScreenInfoPtr rfbScreen,char *str, int len)
|
|
|
|
{
|
|
|
|
rfbClientPtr cl;
|
|
|
|
rfbServerCutTextMsg sct;
|
|
|
|
rfbClientIteratorPtr iterator;
|
|
|
|
|
|
|
|
iterator = rfbGetClientIterator(rfbScreen);
|
|
|
|
while ((cl = rfbClientIteratorNext(iterator)) != NULL) {
|
|
|
|
sct.type = rfbServerCutText;
|
|
|
|
sct.length = Swap32IfLE(len);
|
|
|
|
LOCK(cl->sendMutex);
|
|
|
|
if (rfbWriteExact(cl, (char *)&sct,
|
|
|
|
sz_rfbServerCutTextMsg) < 0) {
|
|
|
|
rfbLogPerror("rfbSendServerCutText: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (rfbWriteExact(cl, str, len) < 0) {
|
|
|
|
rfbLogPerror("rfbSendServerCutText: write");
|
|
|
|
rfbCloseClient(cl);
|
|
|
|
}
|
|
|
|
UNLOCK(cl->sendMutex);
|
|
|
|
rfbStatRecordMessageSent(cl, rfbServerCutText, sz_rfbServerCutTextMsg+len, sz_rfbServerCutTextMsg+len);
|
|
|
|
}
|
|
|
|
rfbReleaseClientIterator(iterator);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*****************************************************************************
|
|
|
|
*
|
|
|
|
* UDP can be used for keyboard and pointer events when the underlying
|
|
|
|
* network is highly reliable. This is really here to support ORL's
|
|
|
|
* videotile, whose TCP implementation doesn't like sending lots of small
|
|
|
|
* packets (such as 100s of pen readings per second!).
|
|
|
|
*/
|
|
|
|
|
|
|
|
static unsigned char ptrAcceleration = 50;
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbNewUDPConnection(rfbScreenInfoPtr rfbScreen,
|
|
|
|
int sock)
|
|
|
|
{
|
|
|
|
if (write(sock, (char*) &ptrAcceleration, 1) < 0) {
|
|
|
|
rfbLogPerror("rfbNewUDPConnection: write");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Because UDP is a message based service, we can't read the first byte and
|
|
|
|
* then the rest of the packet separately like we do with TCP. We will always
|
|
|
|
* get a whole packet delivered in one go, so we ask read() for the maximum
|
|
|
|
* number of bytes we can possibly get.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void
|
|
|
|
rfbProcessUDPInput(rfbScreenInfoPtr rfbScreen)
|
|
|
|
{
|
|
|
|
int n;
|
|
|
|
rfbClientPtr cl=rfbScreen->udpClient;
|
|
|
|
rfbClientToServerMsg msg;
|
|
|
|
|
|
|
|
if((!cl) || cl->onHold)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if ((n = read(rfbScreen->udpSock, (char *)&msg, sizeof(msg))) <= 0) {
|
|
|
|
if (n < 0) {
|
|
|
|
rfbLogPerror("rfbProcessUDPInput: read");
|
|
|
|
}
|
|
|
|
rfbDisconnectUDPSock(rfbScreen);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (msg.type) {
|
|
|
|
|
|
|
|
case rfbKeyEvent:
|
|
|
|
if (n != sz_rfbKeyEventMsg) {
|
|
|
|
rfbErr("rfbProcessUDPInput: key event incorrect length\n");
|
|
|
|
rfbDisconnectUDPSock(rfbScreen);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
cl->screen->kbdAddEvent(msg.ke.down, (rfbKeySym)Swap32IfLE(msg.ke.key), cl);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case rfbPointerEvent:
|
|
|
|
if (n != sz_rfbPointerEventMsg) {
|
|
|
|
rfbErr("rfbProcessUDPInput: ptr event incorrect length\n");
|
|
|
|
rfbDisconnectUDPSock(rfbScreen);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
cl->screen->ptrAddEvent(msg.pe.buttonMask,
|
|
|
|
Swap16IfLE(msg.pe.x), Swap16IfLE(msg.pe.y), cl);
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
rfbErr("rfbProcessUDPInput: unknown message type %d\n",
|
|
|
|
msg.type);
|
|
|
|
rfbDisconnectUDPSock(rfbScreen);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|