]>
source.dussan.org Git - tigervnc.git/log
Brian P. Hinz [Tue, 21 Apr 2015 03:39:13 +0000 (23:39 -0400)]
Added Fltk patch from Pierre to fix "checksum for freed object" segfault to contrib/fltk folder.
Brian P. Hinz [Wed, 18 Mar 2015 23:13:57 +0000 (19:13 -0400)]
Update build requirements in el6 spec file
Make the dependency on g++ explicit.
Now that the native version of cmake in el6 has been bumped to
v2.8 we can remove the dependency on EPEL's cmake28 package.
Brian P. Hinz [Sun, 15 Mar 2015 19:37:10 +0000 (15:37 -0400)]
Remove unnecessary dependency on openmotif
Requirement was an artifact of older version of mesa, and is no longer necessary.
Brian P. Hinz [Sat, 14 Mar 2015 22:06:14 +0000 (18:06 -0400)]
Update package builds to meet new minimum requirements.
el5 and el6 need newer versions of gnutls (and it's prereqs) than are available natively.
Pierre Ossman [Thu, 12 Mar 2015 10:15:32 +0000 (11:15 +0100)]
Merge branch 'upreqs' of https://github.com/CendioOssman/tigervnc
Conflicts:
contrib/packages/rpm/el5/SPECS/tigervnc.spec
vncviewer/Viewport.cxx
Brian P. Hinz [Sun, 8 Mar 2015 03:05:04 +0000 (22:05 -0500)]
Don't disable xinerama (#147)
Xinerama was disabled in the redhat spec file that ours was adapted from. Unsure why it was disabled in the first place.
Pierre Ossman [Wed, 4 Mar 2015 09:25:49 +0000 (10:25 +0100)]
Merge branch 'mac_menu' of https://github.com/CendioOssman/tigervnc
Pierre Ossman [Wed, 4 Mar 2015 08:57:06 +0000 (09:57 +0100)]
Add a OS X system menu so that new viewer windows can be started
Pierre Ossman [Wed, 4 Mar 2015 08:54:27 +0000 (09:54 +0100)]
Hide OS X specific callback when not needed
Pierre Ossman [Wed, 4 Mar 2015 08:53:16 +0000 (09:53 +0100)]
Fix bad long line detection logic in config file parser
Pierre Ossman [Wed, 4 Mar 2015 08:52:48 +0000 (09:52 +0100)]
Change type to fix bad signed/unsigned comparison
Pierre Ossman [Tue, 3 Mar 2015 15:59:51 +0000 (16:59 +0100)]
Merge branch 'warnings' of https://github.com/CendioOssman/tigervnc
Pierre Ossman [Tue, 3 Mar 2015 15:50:47 +0000 (16:50 +0100)]
Make warnings into errors when doing debug builds
Assume debug builds means it's a developer working on TigerVNC
rather than a user just compiling it. At that point we should not
accept any new warnings in the code.
Pierre Ossman [Tue, 3 Mar 2015 15:50:15 +0000 (16:50 +0100)]
Make sure defaultServerName is initialised even on errors
Pierre Ossman [Tue, 3 Mar 2015 15:49:50 +0000 (16:49 +0100)]
Move about text constant so the compiler can analyse it
Pierre Ossman [Tue, 3 Mar 2015 15:48:58 +0000 (16:48 +0100)]
Fix unsafe format strings to the logging functions
Pierre Ossman [Tue, 3 Mar 2015 15:48:36 +0000 (16:48 +0100)]
Don't shove a pointer into a 32-bit variable
Pierre Ossman [Tue, 3 Mar 2015 15:48:01 +0000 (16:48 +0100)]
Fix bad initialiser lists
Pierre Ossman [Tue, 3 Mar 2015 15:47:24 +0000 (16:47 +0100)]
Compiler doesn't like us not checking return values here
Pierre Ossman [Tue, 3 Mar 2015 15:47:01 +0000 (16:47 +0100)]
Compiler doesn't like referencing unset variables
Pierre Ossman [Tue, 3 Mar 2015 15:46:19 +0000 (16:46 +0100)]
LPARAM and WPARAM have different types on Win32 and Win64
Solve it by casting to the larger of the two, i.e. 64-bit.
Pierre Ossman [Tue, 3 Mar 2015 15:45:55 +0000 (16:45 +0100)]
WIN64 should come from the compiler or SDK, not us
Pierre Ossman [Tue, 3 Mar 2015 15:45:18 +0000 (16:45 +0100)]
Handler still needs to return something even with XDamage
Pierre Ossman [Tue, 3 Mar 2015 15:45:02 +0000 (16:45 +0100)]
No need to force modifiable strings here
Pierre Ossman [Tue, 3 Mar 2015 15:44:30 +0000 (16:44 +0100)]
Pointer truncation/expansion needs to be very explicit
Pierre Ossman [Tue, 3 Mar 2015 15:43:20 +0000 (16:43 +0100)]
Wrong array index in CPU usage calculation on Windows
Pierre Ossman [Tue, 3 Mar 2015 15:43:05 +0000 (16:43 +0100)]
Windows doesn't have %lld
Pierre Ossman [Tue, 3 Mar 2015 15:42:45 +0000 (16:42 +0100)]
overrun() needs to return the available space
Pierre Ossman [Tue, 3 Mar 2015 15:42:15 +0000 (16:42 +0100)]
Move things around slightly so the compiler can track cc
Pierre Ossman [Tue, 3 Mar 2015 15:41:29 +0000 (16:41 +0100)]
Give CharArray a printf style method to ease automatic allocation
Pierre Ossman [Tue, 3 Mar 2015 15:40:36 +0000 (16:40 +0100)]
The PAM callback is slightly different on Solaris
Pierre Ossman [Tue, 3 Mar 2015 15:39:37 +0000 (16:39 +0100)]
Add extra parentheses to avoid and/or priority ambiguity
Pierre Ossman [Tue, 3 Mar 2015 15:37:43 +0000 (16:37 +0100)]
Reorder class initialiser list to match what actually happens
Pierre Ossman [Tue, 3 Mar 2015 15:35:28 +0000 (16:35 +0100)]
Make Rect::area() signed, just like Rect::width()/height()
Makes it more sane what type to use when interacting with this class.
Pierre Ossman [Tue, 3 Mar 2015 15:34:56 +0000 (16:34 +0100)]
Fix mismatches between format string and arguments
Pierre Ossman [Tue, 3 Mar 2015 15:30:13 +0000 (16:30 +0100)]
Shorten stats from EncodeManager using SI/IEC prefixes
Also avoids %lld which isn't supported on Windows.
Pierre Ossman [Tue, 3 Mar 2015 15:28:54 +0000 (16:28 +0100)]
A switch() needs to handle every value in an enum
Pierre Ossman [Tue, 3 Mar 2015 15:27:44 +0000 (16:27 +0100)]
Switch to unsigned parameters for ExtendedDesktopSize handler
The constants used here are unsigned so it makes more sense that
the parameters are as well.
Pierre Ossman [Tue, 3 Mar 2015 15:26:03 +0000 (16:26 +0100)]
Fix bad signed/unsigned comparisons
Either by casting, or switching to a more appropriate type
for the variable.
Pierre Ossman [Tue, 3 Mar 2015 15:17:51 +0000 (16:17 +0100)]
Make sure Exceptions do not use unsafe format strings
Pierre Ossman [Tue, 3 Mar 2015 15:11:26 +0000 (16:11 +0100)]
Tag the Exception constructor as a printf function
This allows the compiler to analyse the format strings and
arguments for correctness.
Pierre Ossman [Tue, 3 Mar 2015 15:10:53 +0000 (16:10 +0100)]
Remove unused variables
Pierre Ossman [Tue, 3 Mar 2015 15:03:32 +0000 (16:03 +0100)]
Handle no useful address types from getaddrinfo()
Pierre Ossman [Tue, 3 Mar 2015 15:02:42 +0000 (16:02 +0100)]
The compiler cannot figure out if this is assigned properly
The logic is a bit complex so we're going to have to silence the
warning with this explicit assignment.
Pierre Ossman [Tue, 3 Mar 2015 15:02:03 +0000 (16:02 +0100)]
Turn on more warnings to catch potential bugs
Brian P. Hinz [Tue, 3 Mar 2015 01:03:42 +0000 (20:03 -0500)]
Decode serverCutText to ISO-8859-1
Java viewer was incorrectly attempting to decode incoming clipboard updates to UTF-8 instead of Latin-1.
Brian Hinz [Mon, 2 Mar 2015 03:02:04 +0000 (22:02 -0500)]
Merge pull request #142 from bphinz/master
Updates to package builds
Brian P. Hinz [Mon, 2 Mar 2015 02:55:43 +0000 (21:55 -0500)]
Build local static libpng on el5
The system libpng is too old for fltk 1.3.3, resulting in undefined reference errors when trying to link vncviewer.
Brian P. Hinz [Mon, 2 Mar 2015 02:54:42 +0000 (21:54 -0500)]
Remove obsolete distro patches
Commit
f0fffd4d7f0907e91dfcee02cd7fbe4f4cbe2342 makes ldnow patch unnecessary.
Commit
0dc36a501ee5e69ce1a201935a4a4a903d06cbe1 makes format-security patch unnecessary.
Brian P. Hinz [Mon, 2 Mar 2015 02:52:39 +0000 (21:52 -0500)]
Use autotools to build fltk-1.3.3 on el5 instead of cmake
Static libraries are too hard to get linked correctly with cmake.
Brian P. Hinz [Mon, 2 Mar 2015 02:51:49 +0000 (21:51 -0500)]
Bump version of fltk to 1.3.3 for el5, el6, & ubuntu packages
Brian Hinz [Sun, 1 Mar 2015 19:41:15 +0000 (14:41 -0500)]
Merge pull request #141 from bphinz/java
Don't violate RFB spec by sending UTF-8
Brian P. Hinz [Sun, 1 Mar 2015 19:40:03 +0000 (14:40 -0500)]
Don't violate RFB spec by sending UTF-8
Java viewer was violating the RFB specification by sending UTF-8 encoded clipboard text rather than ISO-8859-1.
Pierre Ossman [Fri, 27 Feb 2015 08:28:06 +0000 (09:28 +0100)]
Merge branch 'perf' of https://github.com/dcommander/tigervnc
Pierre Ossman [Fri, 27 Feb 2015 08:26:26 +0000 (09:26 +0100)]
Merge branches 'format-security' and 'module-ldnow-master' of https://github.com/twaugh/tigervnc
DRC [Thu, 26 Feb 2015 18:24:03 +0000 (12:24 -0600)]
encperf, decperf: catch fopen() errors
This prevents unhandled exceptions if the file isn't found.
Tim Waugh [Tue, 24 Feb 2015 15:59:03 +0000 (15:59 +0000)]
Link libvnc.so with '-z now' to make symbol resolution failures obvious.
Pierre Ossman [Thu, 26 Feb 2015 13:18:49 +0000 (14:18 +0100)]
Merge branch 'cmdline-crash' of https://github.com/twaugh/tigervnc
Pierre Ossman [Thu, 26 Feb 2015 13:14:21 +0000 (14:14 +0100)]
Merge branch 'xserver117-master' of https://github.com/twaugh/tigervnc
Pierre Ossman [Thu, 26 Feb 2015 11:53:37 +0000 (12:53 +0100)]
Merge branch 'perf' of https://github.com/dcommander/tigervnc
DRC [Wed, 25 Feb 2015 20:18:07 +0000 (14:18 -0600)]
encperf: Optionally disable pixel translation
For 16-bit and 8-bit datasets, you can now run encperf with an argument of
-translate=0 to benchmark just the encoder without converting the datasets to
24-bit. This allows the output of encperf to be directly compared with that
of the TurboVNC Benchmark Tools.
DRC [Wed, 25 Feb 2015 20:08:34 +0000 (14:08 -0600)]
encperf: change % to %% in printf format
GCC lets this slide, but other compilers (clang) complain about using % rather
than %% to print a literal % character.
DRC [Sat, 21 Feb 2015 18:01:47 +0000 (12:01 -0600)]
encperf: add compression level to encoding list
Explicitly add the compression level to the list of encodings so it's obvious
how to override it.
DRC [Sat, 21 Feb 2015 17:57:27 +0000 (11:57 -0600)]
encperf: allow # of iterations to be specified
DRC [Sat, 21 Feb 2015 17:44:26 +0000 (11:44 -0600)]
encperf: report encoded & raw-equivalent bytes
Report encoded and raw-equivalent bytes as well as compression ratio (I like
to have these statistics so I can compute overall compression ratios for
groups of datasets in a spreadsheet.)
DRC [Sat, 21 Feb 2015 17:28:37 +0000 (11:28 -0600)]
encperf: Code formatting tweaks
Code formatting tweaks to improve readability & make the code consistent with
the rest of TigerVNC
Tim Waugh [Wed, 25 Feb 2015 12:25:33 +0000 (12:25 +0000)]
Check for out-of-memory when installing color map.
Also remove Xalloc/Xfree definitions, using malloc/free directly
instead. Note that vfbAllocateFramebufferMemory() does not need to
check for failed allocations as it is the caller's responsibility to
do so (and they do).
Pierre Ossman [Wed, 25 Feb 2015 13:43:42 +0000 (14:43 +0100)]
The legacy OS X keyboard API is only 32-bit
Tim Waugh [Wed, 25 Feb 2015 13:40:45 +0000 (13:40 +0000)]
Don't need to log local port when handling -via (#126).
Tim Waugh [Tue, 24 Feb 2015 16:20:41 +0000 (16:20 +0000)]
Updated to build against xorg-x11-server 1.17.x.
Tim Waugh [Wed, 25 Feb 2015 11:27:19 +0000 (11:27 +0000)]
Don't return without value here.
In particular, return 0 as this is a failure.
Pierre Ossman [Wed, 25 Feb 2015 10:27:49 +0000 (11:27 +0100)]
Escape FLTK menu entries
We don't want it automatically creating submenus when least
expected.
Pierre Ossman [Wed, 25 Feb 2015 09:35:03 +0000 (10:35 +0100)]
Revert "Switch to FLTK's copy() method for menus"
This reverts commit
e95412569bd45ec3da43e1b9a244856e0408f6ab
and commit
52a18150a7a8f2802000136c35005bc83b5b0855 .
The copy() method unfortunately has its own problems so it was not
a good replacement.
Tim Waugh [Tue, 24 Feb 2015 12:34:14 +0000 (12:34 +0000)]
Avoid warning when compiling with gcc's -Wformat-security option.
Pierre Ossman [Fri, 20 Feb 2015 16:34:49 +0000 (17:34 +0100)]
Disable entries instead of shuffling memory around
Less risk for bugs this way.
Pierre Ossman [Fri, 20 Feb 2015 16:12:57 +0000 (17:12 +0100)]
Update copyright years
Pierre Ossman [Fri, 20 Feb 2015 16:08:30 +0000 (17:08 +0100)]
Switch to FLTK's copy() method for menus
It has less surprises in terms of magical characters.
Tim Waugh [Wed, 18 Feb 2015 14:55:11 +0000 (14:55 +0000)]
Abort argument processing on error.
This prevents e.g. 'Xvnc -screen x' crashing.
Pierre Ossman [Tue, 17 Feb 2015 12:23:00 +0000 (13:23 +0100)]
Create a console window for help output on Windows
Pierre Ossman [Mon, 16 Feb 2015 15:31:51 +0000 (16:31 +0100)]
Add shortcuts to context menu for quick access
Pierre Ossman [Mon, 16 Feb 2015 15:24:26 +0000 (16:24 +0100)]
Stop mapping Meta to Super
A lot of X11 keyboard layouts send Meta for Shift-Alt so it's not
really a good idea to assume this is the Windows key.
Pierre Ossman [Mon, 20 Oct 2014 15:21:51 +0000 (17:21 +0200)]
Remove no longer functional workaround
This workaround is no longer effective with the new system handler
method of handling keyboard events. The bug has also been fixed
in upstream FLTK, further making it pointless.
Pierre Ossman [Fri, 13 Feb 2015 12:46:16 +0000 (13:46 +0100)]
Merge branch 'perf' of https://github.com/CendioOssman/tigervnc
Pierre Ossman [Thu, 12 Feb 2015 13:57:05 +0000 (14:57 +0100)]
Default compression level should be interpreted as high
Pierre Ossman [Thu, 12 Feb 2015 12:44:22 +0000 (13:44 +0100)]
Be more aggressive in finding solid rectangles
The previous FIXME was incorrect. The old code did in fact continue
looking over the entire area. This commit restores that old behaviour.
Pierre Ossman [Wed, 11 Feb 2015 12:49:04 +0000 (13:49 +0100)]
Add test to measure high level encoding performance
Pierre Ossman [Wed, 11 Feb 2015 16:43:15 +0000 (17:43 +0100)]
Move encoder statistics into EncodeManager from SMsgWriter
It knows more details so it can track things better.
Pierre Ossman [Fri, 6 Feb 2015 12:51:35 +0000 (13:51 +0100)]
Optimise fillRect()
It had regressed in performance compared to 1.3. This brings it back
up to the same speed.
Pierre Ossman [Fri, 6 Feb 2015 12:50:36 +0000 (13:50 +0100)]
Make it possible to measure CPU usage in parellel
Pierre Ossman [Wed, 4 Feb 2015 13:12:04 +0000 (14:12 +0100)]
Add test to measure decoding performance
Pierre Ossman [Fri, 13 Feb 2015 10:12:42 +0000 (11:12 +0100)]
GnuTLS needs socket libraries so include them in the static link
Pierre Ossman [Wed, 11 Feb 2015 12:47:21 +0000 (13:47 +0100)]
rfb::Region uses the Xregion library
Pierre Ossman [Wed, 4 Feb 2015 13:11:42 +0000 (14:11 +0100)]
Add stream class to feed data from a file
Pierre Ossman [Wed, 11 Feb 2015 12:47:58 +0000 (13:47 +0100)]
No need for this array to be writeable
Pierre Ossman [Fri, 6 Feb 2015 13:41:58 +0000 (14:41 +0100)]
Move reverse connection information into VNCSConnectionST
The only class that cares if connections are reversed or not is
VNCSConnectionST so let it keep track of this fact.
Pierre Ossman [Wed, 4 Feb 2015 13:11:11 +0000 (14:11 +0100)]
These methods aren't critical so no need for them to be pure
Pierre Ossman [Wed, 4 Feb 2015 13:10:43 +0000 (14:10 +0100)]
Give subclasses the ability to overwrite the reader and writer
Needed for some tests.
Pierre Ossman [Fri, 6 Feb 2015 13:44:32 +0000 (14:44 +0100)]
SConnection internals should be private, just like for CConnection
Peter Åstrand (astrand) [Tue, 10 Feb 2015 11:21:05 +0000 (12:21 +0100)]
Build fixes for modern versions of binutils.
Need to explicitly list DSOs.
Peter Åstrand (astrand) [Thu, 5 Feb 2015 10:10:32 +0000 (11:10 +0100)]
Use a common header file for Windows errno names.
In earlier Visual Studio and MinGW editions, BSD socket errno:s were
left undefined. This is no longer the case. This may cause build or
runtime errors. To avoid this, we are using a common header file which
corrects all definitions. This header will also be used with other
projects such as sercd, unfs3, PulseAudio etc.