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)2009-2012 D. R. Commander. All Rights Reserved.
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions are met:
|
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
|
|
|
*
|
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
|
|
|
* - Redistributions of source code must retain the above copyright notice,
|
|
|
|
* this list of conditions and the following disclaimer.
|
|
|
|
* - Redistributions in binary form must reproduce the above copyright notice,
|
|
|
|
* this list of conditions and the following disclaimer in the documentation
|
|
|
|
* and/or other materials provided with the distribution.
|
|
|
|
* - Neither the name of the libjpeg-turbo Project nor the names of its
|
|
|
|
* contributors may be used to endorse or promote products derived from this
|
|
|
|
* software without specific prior written permission.
|
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
|
|
|
*
|
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
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS",
|
|
|
|
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE
|
|
|
|
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
|
|
|
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
|
|
|
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
|
|
|
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
|
|
|
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
|
|
|
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
|
|
|
* POSSIBILITY OF SUCH DAMAGE.
|
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
|
|
|
*/
|
|
|
|
|
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
|
|
|
#ifndef __TURBOJPEG_H__
|
|
|
|
#define __TURBOJPEG_H__
|
|
|
|
|
|
|
|
#if defined(_WIN32) && defined(DLLDEFINE)
|
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
|
|
|
#define DLLEXPORT __declspec(dllexport)
|
|
|
|
#else
|
|
|
|
#define DLLEXPORT
|
|
|
|
#endif
|
|
|
|
#define DLLCALL
|
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* @addtogroup TurboJPEG Lite
|
|
|
|
* TurboJPEG API. This API provides an interface for generating and decoding
|
|
|
|
* JPEG images in memory.
|
|
|
|
*
|
|
|
|
* @{
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
* The number of chrominance subsampling options
|
|
|
|
*/
|
|
|
|
#define TJ_NUMSAMP 5
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Chrominance subsampling options.
|
|
|
|
* When an image is converted from the RGB to the YCbCr colorspace as part of
|
|
|
|
* the JPEG compression process, some of the Cb and Cr (chrominance) components
|
|
|
|
* can be discarded or averaged together to produce a smaller image with little
|
|
|
|
* perceptible loss of image clarity (the human eye is more sensitive to small
|
|
|
|
* changes in brightness than small changes in color.) This is called
|
|
|
|
* "chrominance subsampling".
|
|
|
|
*/
|
|
|
|
enum TJSAMP
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
* 4:4:4 chrominance subsampling (no chrominance subsampling). The JPEG or
|
|
|
|
* YUV image will contain one chrominance component for every pixel in the
|
|
|
|
* source image.
|
|
|
|
*/
|
|
|
|
TJSAMP_444=0,
|
|
|
|
/**
|
|
|
|
* 4:2:2 chrominance subsampling. The JPEG or YUV image will contain one
|
|
|
|
* chrominance component for every 2x1 block of pixels in the source image.
|
|
|
|
*/
|
|
|
|
TJSAMP_422,
|
|
|
|
/**
|
|
|
|
* 4:2:0 chrominance subsampling. The JPEG or YUV image will contain one
|
|
|
|
* chrominance component for every 2x2 block of pixels in the source image.
|
|
|
|
*/
|
|
|
|
TJSAMP_420,
|
|
|
|
/**
|
|
|
|
* Grayscale. The JPEG or YUV image will contain no chrominance components.
|
|
|
|
*/
|
|
|
|
TJSAMP_GRAY,
|
|
|
|
/**
|
|
|
|
* 4:4:0 chrominance subsampling. The JPEG or YUV image will contain one
|
|
|
|
* chrominance component for every 1x2 block of pixels in the source image.
|
|
|
|
*/
|
|
|
|
TJSAMP_440
|
|
|
|
};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* MCU block width (in pixels) for a given level of chrominance subsampling.
|
|
|
|
* MCU block sizes:
|
|
|
|
* - 8x8 for no subsampling or grayscale
|
|
|
|
* - 16x8 for 4:2:2
|
|
|
|
* - 8x16 for 4:4:0
|
|
|
|
* - 16x16 for 4:2:0
|
|
|
|
*/
|
|
|
|
static const int tjMCUWidth[TJ_NUMSAMP] = {8, 16, 16, 8, 8};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* MCU block height (in pixels) for a given level of chrominance subsampling.
|
|
|
|
* MCU block sizes:
|
|
|
|
* - 8x8 for no subsampling or grayscale
|
|
|
|
* - 16x8 for 4:2:2
|
|
|
|
* - 8x16 for 4:4:0
|
|
|
|
* - 16x16 for 4:2:0
|
|
|
|
*/
|
|
|
|
static const int tjMCUHeight[TJ_NUMSAMP] = {8, 8, 16, 8, 16};
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
* The number of pixel formats
|
|
|
|
*/
|
|
|
|
#define TJ_NUMPF 11
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Pixel formats
|
|
|
|
*/
|
|
|
|
enum TJPF
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
* RGB pixel format. The red, green, and blue components in the image are
|
|
|
|
* stored in 3-byte pixels in the order R, G, B from lowest to highest byte
|
|
|
|
* address within each pixel.
|
|
|
|
*/
|
|
|
|
TJPF_RGB=0,
|
|
|
|
/**
|
|
|
|
* BGR pixel format. The red, green, and blue components in the image are
|
|
|
|
* stored in 3-byte pixels in the order B, G, R from lowest to highest byte
|
|
|
|
* address within each pixel.
|
|
|
|
*/
|
|
|
|
TJPF_BGR,
|
|
|
|
/**
|
|
|
|
* RGBX pixel format. The red, green, and blue components in the image are
|
|
|
|
* stored in 4-byte pixels in the order R, G, B from lowest to highest byte
|
|
|
|
* address within each pixel. The X component is ignored when compressing
|
|
|
|
* and undefined when decompressing.
|
|
|
|
*/
|
|
|
|
TJPF_RGBX,
|
|
|
|
/**
|
|
|
|
* BGRX pixel format. The red, green, and blue components in the image are
|
|
|
|
* stored in 4-byte pixels in the order B, G, R from lowest to highest byte
|
|
|
|
* address within each pixel. The X component is ignored when compressing
|
|
|
|
* and undefined when decompressing.
|
|
|
|
*/
|
|
|
|
TJPF_BGRX,
|
|
|
|
/**
|
|
|
|
* XBGR pixel format. The red, green, and blue components in the image are
|
|
|
|
* stored in 4-byte pixels in the order R, G, B from highest to lowest byte
|
|
|
|
* address within each pixel. The X component is ignored when compressing
|
|
|
|
* and undefined when decompressing.
|
|
|
|
*/
|
|
|
|
TJPF_XBGR,
|
|
|
|
/**
|
|
|
|
* XRGB pixel format. The red, green, and blue components in the image are
|
|
|
|
* stored in 4-byte pixels in the order B, G, R from highest to lowest byte
|
|
|
|
* address within each pixel. The X component is ignored when compressing
|
|
|
|
* and undefined when decompressing.
|
|
|
|
*/
|
|
|
|
TJPF_XRGB,
|
|
|
|
/**
|
|
|
|
* Grayscale pixel format. Each 1-byte pixel represents a luminance
|
|
|
|
* (brightness) level from 0 to 255.
|
|
|
|
*/
|
|
|
|
TJPF_GRAY,
|
|
|
|
/**
|
|
|
|
* RGBA pixel format. This is the same as @ref TJPF_RGBX, except that when
|
|
|
|
* decompressing, the X component is guaranteed to be 0xFF, which can be
|
|
|
|
* interpreted as an opaque alpha channel.
|
|
|
|
*/
|
|
|
|
TJPF_RGBA,
|
|
|
|
/**
|
|
|
|
* BGRA pixel format. This is the same as @ref TJPF_BGRX, except that when
|
|
|
|
* decompressing, the X component is guaranteed to be 0xFF, which can be
|
|
|
|
* interpreted as an opaque alpha channel.
|
|
|
|
*/
|
|
|
|
TJPF_BGRA,
|
|
|
|
/**
|
|
|
|
* ABGR pixel format. This is the same as @ref TJPF_XBGR, except that when
|
|
|
|
* decompressing, the X component is guaranteed to be 0xFF, which can be
|
|
|
|
* interpreted as an opaque alpha channel.
|
|
|
|
*/
|
|
|
|
TJPF_ABGR,
|
|
|
|
/**
|
|
|
|
* ARGB pixel format. This is the same as @ref TJPF_XRGB, except that when
|
|
|
|
* decompressing, the X component is guaranteed to be 0xFF, which can be
|
|
|
|
* interpreted as an opaque alpha channel.
|
|
|
|
*/
|
|
|
|
TJPF_ARGB
|
|
|
|
};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Red offset (in bytes) for a given pixel format. This specifies the number
|
|
|
|
* of bytes that the red component is offset from the start of the pixel. For
|
|
|
|
* instance, if a pixel of format TJ_BGRX is stored in <tt>char pixel[]</tt>,
|
|
|
|
* then the red component will be <tt>pixel[tjRedOffset[TJ_BGRX]]</tt>.
|
|
|
|
*/
|
|
|
|
static const int tjRedOffset[TJ_NUMPF] = {0, 2, 0, 2, 3, 1, 0, 0, 2, 3, 1};
|
|
|
|
/**
|
|
|
|
* Green offset (in bytes) for a given pixel format. This specifies the number
|
|
|
|
* of bytes that the green component is offset from the start of the pixel.
|
|
|
|
* For instance, if a pixel of format TJ_BGRX is stored in
|
|
|
|
* <tt>char pixel[]</tt>, then the green component will be
|
|
|
|
* <tt>pixel[tjGreenOffset[TJ_BGRX]]</tt>.
|
|
|
|
*/
|
|
|
|
static const int tjGreenOffset[TJ_NUMPF] = {1, 1, 1, 1, 2, 2, 0, 1, 1, 2, 2};
|
|
|
|
/**
|
|
|
|
* Blue offset (in bytes) for a given pixel format. This specifies the number
|
|
|
|
* of bytes that the Blue component is offset from the start of the pixel. For
|
|
|
|
* instance, if a pixel of format TJ_BGRX is stored in <tt>char pixel[]</tt>,
|
|
|
|
* then the blue component will be <tt>pixel[tjBlueOffset[TJ_BGRX]]</tt>.
|
|
|
|
*/
|
|
|
|
static const int tjBlueOffset[TJ_NUMPF] = {2, 0, 2, 0, 1, 3, 0, 2, 0, 1, 3};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Pixel size (in bytes) for a given pixel format.
|
|
|
|
*/
|
|
|
|
static const int tjPixelSize[TJ_NUMPF] = {3, 3, 4, 4, 4, 4, 1, 4, 4, 4, 4};
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* The uncompressed source/destination image is stored in bottom-up (Windows,
|
|
|
|
* OpenGL) order, not top-down (X11) order.
|
|
|
|
*/
|
|
|
|
#define TJFLAG_BOTTOMUP 2
|
|
|
|
/**
|
|
|
|
* Turn off CPU auto-detection and force TurboJPEG to use MMX code (IPP and
|
|
|
|
* 32-bit libjpeg-turbo versions only.)
|
|
|
|
*/
|
|
|
|
#define TJFLAG_FORCEMMX 8
|
|
|
|
/**
|
|
|
|
* Turn off CPU auto-detection and force TurboJPEG to use SSE code (32-bit IPP
|
|
|
|
* and 32-bit libjpeg-turbo versions only)
|
|
|
|
*/
|
|
|
|
#define TJFLAG_FORCESSE 16
|
|
|
|
/**
|
|
|
|
* Turn off CPU auto-detection and force TurboJPEG to use SSE2 code (32-bit IPP
|
|
|
|
* and 32-bit libjpeg-turbo versions only)
|
|
|
|
*/
|
|
|
|
#define TJFLAG_FORCESSE2 32
|
|
|
|
/**
|
|
|
|
* Turn off CPU auto-detection and force TurboJPEG to use SSE3 code (64-bit IPP
|
|
|
|
* version only)
|
|
|
|
*/
|
|
|
|
#define TJFLAG_FORCESSE3 128
|
|
|
|
/**
|
|
|
|
* Use fast, inaccurate chrominance upsampling routines in the JPEG
|
|
|
|
* decompressor (libjpeg and libjpeg-turbo versions only)
|
|
|
|
*/
|
|
|
|
#define TJFLAG_FASTUPSAMPLE 256
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Scaling factor
|
|
|
|
*/
|
|
|
|
typedef struct
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
* Numerator
|
|
|
|
*/
|
|
|
|
int num;
|
|
|
|
/**
|
|
|
|
* Denominator
|
|
|
|
*/
|
|
|
|
int denom;
|
|
|
|
} tjscalingfactor;
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
* TurboJPEG instance handle
|
|
|
|
*/
|
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
|
|
|
typedef void* tjhandle;
|
|
|
|
|
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
|
|
|
|
|
|
|
/**
|
|
|
|
* Pad the given width to the nearest 32-bit boundary
|
|
|
|
*/
|
|
|
|
#define TJPAD(width) (((width)+3)&(~3))
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Compute the scaled value of <tt>dimension</tt> using the given scaling
|
|
|
|
* factor. This macro performs the integer equivalent of <tt>ceil(dimension *
|
|
|
|
* scalingFactor)</tt>.
|
|
|
|
*/
|
|
|
|
#define TJSCALED(dimension, scalingFactor) ((dimension * scalingFactor.num \
|
|
|
|
+ scalingFactor.denom - 1) / scalingFactor.denom)
|
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
|
|
|
|
|
|
|
|
|
|
|
#ifdef __cplusplus
|
|
|
|
extern "C" {
|
|
|
|
#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
|
|
|
/**
|
|
|
|
* Create a TurboJPEG compressor instance.
|
|
|
|
*
|
|
|
|
* @return a handle to the newly-created instance, or NULL if an error
|
|
|
|
* occurred (see #tjGetErrorStr().)
|
|
|
|
*/
|
|
|
|
DLLEXPORT tjhandle DLLCALL tjInitCompress(void);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Compress an RGB or grayscale image into a JPEG image.
|
|
|
|
*
|
|
|
|
* @param handle a handle to a TurboJPEG compressor or transformer instance
|
|
|
|
* @param srcBuf pointer to an image buffer containing RGB or grayscale pixels
|
|
|
|
* to be compressed
|
|
|
|
* @param width width (in pixels) of the source image
|
|
|
|
* @param pitch bytes per line of the source image. Normally, this should be
|
|
|
|
* <tt>width * #tjPixelSize[pixelFormat]</tt> if the image is unpadded,
|
|
|
|
* or <tt>#TJPAD(width * #tjPixelSize[pixelFormat])</tt> if each line of
|
|
|
|
* the image is padded to the nearest 32-bit boundary, as is the case
|
|
|
|
* for Windows bitmaps. You can also be clever and use this parameter
|
|
|
|
* to skip lines, etc. Setting this parameter to 0 is the equivalent of
|
|
|
|
* setting it to <tt>width * #tjPixelSize[pixelFormat]</tt>.
|
|
|
|
* @param height height (in pixels) of the source image
|
|
|
|
* @param pixelFormat pixel format of the source image (see @ref TJPF
|
|
|
|
* "Pixel formats".)
|
|
|
|
* @param jpegBuf address of a pointer to an image buffer that will receive the
|
|
|
|
* JPEG image. TurboJPEG has the ability to reallocate the JPEG buffer
|
|
|
|
* to accommodate the size of the JPEG image. Thus, you can choose to:
|
|
|
|
* -# pre-allocate the JPEG buffer with an arbitrary size using
|
|
|
|
* #tjAlloc() and let TurboJPEG grow the buffer as needed,
|
|
|
|
* -# set <tt>*jpegBuf</tt> to NULL to tell TurboJPEG to allocate the
|
|
|
|
* buffer for you, or
|
|
|
|
* -# pre-allocate the buffer to a "worst case" size determined by
|
|
|
|
* calling #tjBufSize(). This should ensure that the buffer never has
|
|
|
|
* to be re-allocated (setting #TJFLAG_NOREALLOC guarantees this.)
|
|
|
|
* .
|
|
|
|
* If you choose option 1, <tt>*jpegSize</tt> should be set to the
|
|
|
|
* size of your pre-allocated buffer. In any case, unless you have
|
|
|
|
* set #TJFLAG_NOREALLOC, you should always check <tt>*jpegBuf</tt> upon
|
|
|
|
* return from this function, as it may have changed.
|
|
|
|
* @param jpegSize pointer to an unsigned long variable that holds the size of
|
|
|
|
* the JPEG image buffer. If <tt>*jpegBuf</tt> points to a
|
|
|
|
* pre-allocated buffer, then <tt>*jpegSize</tt> should be set to the
|
|
|
|
* size of the buffer. Upon return, <tt>*jpegSize</tt> will contain the
|
|
|
|
* size of the JPEG image (in bytes.)
|
|
|
|
* @param jpegSubsamp the level of chrominance subsampling to be used when
|
|
|
|
* generating the JPEG image (see @ref TJSAMP
|
|
|
|
* "Chrominance subsampling options".)
|
|
|
|
* @param jpegQual the image quality of the generated JPEG image (1 = worst,
|
|
|
|
100 = best)
|
|
|
|
* @param flags the bitwise OR of one or more of the @ref TJFLAG_BOTTOMUP
|
|
|
|
* "flags".
|
|
|
|
*
|
|
|
|
* @return 0 if successful, or -1 if an error occurred (see #tjGetErrorStr().)
|
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
|
|
|
*/
|
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
|
|
|
DLLEXPORT int DLLCALL tjCompress2(tjhandle handle, unsigned char *srcBuf,
|
|
|
|
int width, int pitch, int height, int pixelFormat, unsigned char **jpegBuf,
|
|
|
|
unsigned long *jpegSize, int jpegSubsamp, int jpegQual, int flags);
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
* The maximum size of the buffer (in bytes) required to hold a JPEG image with
|
|
|
|
* the given parameters. The number of bytes returned by this function is
|
|
|
|
* larger than the size of the uncompressed source image. The reason for this
|
|
|
|
* is that the JPEG format uses 16-bit coefficients, and it is thus possible
|
|
|
|
* for a very high-quality JPEG image with very high frequency content to
|
|
|
|
* expand rather than compress when converted to the JPEG format. Such images
|
|
|
|
* represent a very rare corner case, but since there is no way to predict the
|
|
|
|
* size of a JPEG image prior to compression, the corner case has to be
|
|
|
|
* handled.
|
|
|
|
*
|
|
|
|
* @param width width of the image (in pixels)
|
|
|
|
* @param height height of the image (in pixels)
|
|
|
|
* @param jpegSubsamp the level of chrominance subsampling to be used when
|
|
|
|
* generating the JPEG image (see @ref TJSAMP
|
|
|
|
* "Chrominance subsampling options".)
|
|
|
|
*
|
|
|
|
* @return the maximum size of the buffer (in bytes) required to hold the
|
|
|
|
* image, or -1 if the arguments are out of bounds.
|
|
|
|
*/
|
|
|
|
DLLEXPORT unsigned long DLLCALL tjBufSize(int width, int height,
|
|
|
|
int jpegSubsamp);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Create a TurboJPEG decompressor instance.
|
|
|
|
*
|
|
|
|
* @return a handle to the newly-created instance, or NULL if an error
|
|
|
|
* occurred (see #tjGetErrorStr().)
|
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
|
|
|
*/
|
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
|
|
|
DLLEXPORT tjhandle DLLCALL tjInitDecompress(void);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Retrieve information about a JPEG image without decompressing it.
|
|
|
|
*
|
|
|
|
* @param handle a handle to a TurboJPEG decompressor or transformer instance
|
|
|
|
* @param jpegBuf pointer to a buffer containing a JPEG image
|
|
|
|
* @param jpegSize size of the JPEG image (in bytes)
|
|
|
|
* @param width pointer to an integer variable that will receive the width (in
|
|
|
|
* pixels) of the JPEG image
|
|
|
|
* @param height pointer to an integer variable that will receive the height
|
|
|
|
* (in pixels) of the JPEG image
|
|
|
|
* @param jpegSubsamp pointer to an integer variable that will receive the
|
|
|
|
* level of chrominance subsampling used when compressing the JPEG image
|
|
|
|
* (see @ref TJSAMP "Chrominance subsampling options".)
|
|
|
|
*
|
|
|
|
* @return 0 if successful, or -1 if an error occurred (see #tjGetErrorStr().)
|
|
|
|
*/
|
|
|
|
DLLEXPORT int DLLCALL tjDecompressHeader2(tjhandle handle,
|
|
|
|
unsigned char *jpegBuf, unsigned long jpegSize, int *width, int *height,
|
|
|
|
int *jpegSubsamp);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Returns a list of fractional scaling factors that the JPEG decompressor in
|
|
|
|
* this implementation of TurboJPEG supports.
|
|
|
|
*
|
|
|
|
* @param numscalingfactors pointer to an integer variable that will receive
|
|
|
|
* the number of elements in the list
|
|
|
|
*
|
|
|
|
* @return a pointer to a list of fractional scaling factors, or NULL if an
|
|
|
|
* error is encountered (see #tjGetErrorStr().)
|
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
|
|
|
*/
|
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
|
|
|
DLLEXPORT tjscalingfactor* DLLCALL tjGetScalingFactors(int *numscalingfactors);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Decompress a JPEG image to an RGB or grayscale image.
|
|
|
|
*
|
|
|
|
* @param handle a handle to a TurboJPEG decompressor or transformer instance
|
|
|
|
* @param jpegBuf pointer to a buffer containing the JPEG image to decompress
|
|
|
|
* @param jpegSize size of the JPEG image (in bytes)
|
|
|
|
* @param dstBuf pointer to an image buffer that will receive the decompressed
|
|
|
|
* image. This buffer should normally be <tt>pitch * scaledHeight</tt>
|
|
|
|
* bytes in size, where <tt>scaledHeight</tt> can be determined by
|
|
|
|
* calling #TJSCALED() with the JPEG image height and one of the scaling
|
|
|
|
* factors returned by #tjGetScalingFactors(). The dstBuf pointer may
|
|
|
|
* also be used to decompress into a specific region of a larger buffer.
|
|
|
|
* @param width desired width (in pixels) of the destination image. If this is
|
|
|
|
* smaller than the width of the JPEG image being decompressed, then
|
|
|
|
* TurboJPEG will use scaling in the JPEG decompressor to generate the
|
|
|
|
* largest possible image that will fit within the desired width. If
|
|
|
|
* width is set to 0, then only the height will be considered when
|
|
|
|
* determining the scaled image size.
|
|
|
|
* @param pitch bytes per line of the destination image. Normally, this is
|
|
|
|
* <tt>scaledWidth * #tjPixelSize[pixelFormat]</tt> if the decompressed
|
|
|
|
* image is unpadded, else <tt>#TJPAD(scaledWidth *
|
|
|
|
* #tjPixelSize[pixelFormat])</tt> if each line of the decompressed
|
|
|
|
* image is padded to the nearest 32-bit boundary, as is the case for
|
|
|
|
* Windows bitmaps. (NOTE: <tt>scaledWidth</tt> can be determined by
|
|
|
|
* calling #TJSCALED() with the JPEG image width and one of the scaling
|
|
|
|
* factors returned by #tjGetScalingFactors().) You can also be clever
|
|
|
|
* and use the pitch parameter to skip lines, etc. Setting this
|
|
|
|
* parameter to 0 is the equivalent of setting it to <tt>scaledWidth
|
|
|
|
* * #tjPixelSize[pixelFormat]</tt>.
|
|
|
|
* @param height desired height (in pixels) of the destination image. If this
|
|
|
|
* is smaller than the height of the JPEG image being decompressed, then
|
|
|
|
* TurboJPEG will use scaling in the JPEG decompressor to generate the
|
|
|
|
* largest possible image that will fit within the desired height. If
|
|
|
|
* height is set to 0, then only the width will be considered when
|
|
|
|
* determining the scaled image size.
|
|
|
|
* @param pixelFormat pixel format of the destination image (see @ref
|
|
|
|
* TJPF "Pixel formats".)
|
|
|
|
* @param flags the bitwise OR of one or more of the @ref TJFLAG_BOTTOMUP
|
|
|
|
* "flags".
|
|
|
|
*
|
|
|
|
* @return 0 if successful, or -1 if an error occurred (see #tjGetErrorStr().)
|
|
|
|
*/
|
|
|
|
DLLEXPORT int DLLCALL tjDecompress2(tjhandle handle,
|
|
|
|
unsigned char *jpegBuf, unsigned long jpegSize, unsigned char *dstBuf,
|
|
|
|
int width, int pitch, int height, int pixelFormat, int flags);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Destroy a TurboJPEG compressor, decompressor, or transformer instance.
|
|
|
|
*
|
|
|
|
* @param handle a handle to a TurboJPEG compressor, decompressor or
|
|
|
|
* transformer instance
|
|
|
|
*
|
|
|
|
* @return 0 if successful, or -1 if an error occurred (see #tjGetErrorStr().)
|
|
|
|
*/
|
|
|
|
DLLEXPORT int DLLCALL tjDestroy(tjhandle handle);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* Returns a descriptive error message explaining why the last command failed.
|
|
|
|
*
|
|
|
|
* @return a descriptive error message explaining why the last command failed.
|
|
|
|
*/
|
|
|
|
DLLEXPORT char* DLLCALL tjGetErrorStr(void);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/* Backward compatibility functions and macros (nothing to see here) */
|
|
|
|
#define NUMSUBOPT TJ_NUMSAMP
|
|
|
|
#define TJ_444 TJSAMP_444
|
|
|
|
#define TJ_422 TJSAMP_422
|
|
|
|
#define TJ_420 TJSAMP_420
|
|
|
|
#define TJ_411 TJSAMP_420
|
|
|
|
#define TJ_GRAYSCALE TJSAMP_GRAY
|
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
|
|
|
|
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
|
|
|
#define TJ_BGR 1
|
|
|
|
#define TJ_BOTTOMUP TJFLAG_BOTTOMUP
|
|
|
|
#define TJ_FORCEMMX TJFLAG_FORCEMMX
|
|
|
|
#define TJ_FORCESSE TJFLAG_FORCESSE
|
|
|
|
#define TJ_FORCESSE2 TJFLAG_FORCESSE2
|
|
|
|
#define TJ_ALPHAFIRST 64
|
|
|
|
#define TJ_FORCESSE3 TJFLAG_FORCESSE3
|
|
|
|
#define TJ_FASTUPSAMPLE TJFLAG_FASTUPSAMPLE
|
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
|
|
|
|
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
|
|
|
DLLEXPORT unsigned long DLLCALL TJBUFSIZE(int width, int height);
|
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
|
|
|
|
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
|
|
|
DLLEXPORT int DLLCALL tjCompress(tjhandle handle, unsigned char *srcBuf,
|
|
|
|
int width, int pitch, int height, int pixelSize, unsigned char *dstBuf,
|
|
|
|
unsigned long *compressedSize, int jpegSubsamp, int jpegQual, int flags);
|
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
|
|
|
|
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
|
|
|
DLLEXPORT int DLLCALL tjDecompressHeader(tjhandle handle,
|
|
|
|
unsigned char *jpegBuf, unsigned long jpegSize, int *width, int *height);
|
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
|
|
|
|
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
|
|
|
DLLEXPORT int DLLCALL tjDecompress(tjhandle handle,
|
|
|
|
unsigned char *jpegBuf, unsigned long jpegSize, unsigned char *dstBuf,
|
|
|
|
int width, int pitch, int height, int pixelSize, int flags);
|
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
|
|
|
|
|
|
|
|
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
|
|
|
/**
|
|
|
|
* @}
|
|
|
|
*/
|
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
|
|
|
|
|
|
|
#ifdef __cplusplus
|
|
|
|
}
|
|
|
|
#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
|
|
|
|
|
|
|
#endif
|