The return of the get_array_size() is used as a limit in a loop,
being compared with an unsigned index (indexes are always unsigned).
To avoid the -Wsign-compare warning, let's cast the return of the
get_array_size() to unsigned and make GCC happier.
Although the most part of the parameters marked as unused are actually
being used for a few functions, a bunch of warnings can be seen when
the code is compiled with "-Wall -Wextra". As adding the unused attribute
means that the variable/parameter is meant to be *possibly* unused, we're
safe adding it in the generated code, even for used variables/parameters.
Check for MELCSTATES - 1 to get inside the branch, otherwise
(...)->rgb_state.melcstate may be up to MELCSTATES after the
pre-incrementing, which would result in an access to a position
that is out bounds of the array size MELCSTATES.
These 2 files are not build as part of spice-common
build system, but modules using spice-common will build
them with the appropriate options. We need to let automake
know that these are source files so that it can properly
track these files dependencies.
Without this change, when eg spice-gtk switches to use 'subdir-objects'
to build sw_canvas.c as recommended by automake 1.14, the build
will fail because $(top_srcdir)/spice-common/common/.deps/sw_canvas.Plo
will not have been generated.
Since the (de)marshallers are now generated in $builddir and not in
$srcdir, when these generated files include a file located in
$srcdir/common, the compiler will find them thanks to a -I directive, so it
makes more sense to use <> rather than "" when including them.
Now that they are created in $builddir, their includes will need to refer
to files in $srcdir, which can be different. It's cleaner to add
-I $(top_srcdir)/spice-common/ to modules using spice-common rather than
having -I $(top_srcdir)/spice-common/common which would could create header
collisions.
The rules to generate the .c/.h (de)marshalling files have targets based in
$builddir, but the CLIENT_MARSHALLERS/SERVER_MARSHALLERS list refer to
files in $srcdir.
When using a $srcdir != $builddir, these $srcdir files will not exist, and
it will not be possible to generate them, which causes the build to fail.
When using $srcdir == $builddir from git or from tarballs, this will not
make a difference. When building from git, if $srcdir and $builddir are the
same, then the files will be found regardless of if we look for them in
$srcdir or $builddir as they are the same.
In tarballs, the files will be shipped with the tarball and thus available
in $srcdir. As $builddir is the same as $srcdir, the files will already
exist and not be regenerated.
The only change of behaviour will be when using a tarball and doing a
$srcdir != $builddir build. In this case, the files will need to be
regenerated, so the tools needed for that must be installed on the
machine doing the build.
This channel provides a webdav server (rfc4918). This allows various
guest or remote system that support webdav to access a folder shared by
the client (some agent can be used to proxy the requests on a local port
for example). The webdav server may also be accessed by an hypervisor as
a remote filesystem interface, which can then be accessed by the guest
via other means (fs/fat emulation, mtp device, etc)
Due to the usage of a single channel stream and the need for concurrent
requests, webdav clients streams are multiplexed. Each client stream is
framed within 64k max messages (in little-endian)
int64 client_id
uint16 size
char data[size]
A new client_id indicates a new connection. A new communication stream
with the webdav server should be started. A client stream message of
size 0 indicates a disconnection of client_id. This multiplexed
communication happens over the channel "data" message.
Only when the port is opened may the communication be started.
A closed port event should close all currently known multiplexed
connections.
Why WebDAV?
webdav is supported natively by various OS for a long time (circa
Windows XP). It has several open-source implementations and a variety of
tools exist. A webdav implementation can be tested and used without a
Spice server or any virtualization (this also permit sharing the
implementation with other projects in the future, such as GNOME). It is
an IETF open standard and thus thoroughly specified.
The basic requirements for an efficient remote filesystem are provided
by the standard (pipelining, concurrency, caching, copy/move, partial
io, compression, locking ...) While other features are easily possible
via extensions to the protocol (common ones are executable attributes,
or searching for example).
Given the requirements, and the popularity of http/webdav, I believe it
is the best candidate for Spice remote filesystem support.
Other alternatives (adhoc, p9, smb2, sftp) have been studied and
discarded so far since they do not match in term of features or
requirements.
find_model_params() is first doing *nbuckets = 0; and it then checks
nbuckets for NULL. This is redundant as the dereferencing would cause a
segfault if nbuckets was NULL, so the if (nbuckets) test can't be false.
As Uri pointed out, the "/* bucket start */" comment on the same line
probably implies that the test was meant to be 'if (*nbuckets)'
I've ran a few test and I did not observe issues because of it...
If verify_subject() is called with a SpiceOpenSSLVerify struct containing a
non-NULL 'in_subject' member, it would try to use the local 'in_entries'
variable without having initialized it first. This could happen if
verify_subject() was called multiple time with the same SpiceOpenSSLVerify
context, which probably isn't occurring the way we are using it.
However, since verify_subject() is the only method which needs in_subject,
we don't need to have it stored in SpiceOpenSSLVerify, and we can
recreate it as needed locally in that method, which avoids that issue.
Unhandled values call an error callback, and then fall through the default:
case, which will call again the error callback. This commit adds some
break; after these cases to avoid this.
We are mostly likely not running as root, so this call will fail. As we are
supposed to check its return value as it's declared with
warn_unused_result, the current way of using it is wrong, so this commit just
removes the call.
Use same license as spice-gtk and spice modules (LGPL 2.1) since those licenses
applied to the spice-common submodule in the past. This makes it more clear
that if you use spice-common separately, the license is still LGPL. Also
mention license and copyright in generated files.
The wireshark protocol dissector is a bit out-of-date. Several new channel types
and enums have been added. It would be nice if these values (and the
translation between the value and the name) could be automatically generated so
that updating the dissector was a slightly less manual process. This patch adds
a commandline switch which generates both the enums and value-name lists in the
format that wireshark expects.
Currently, SSL verification of the peer certificate checks if
the certificate's subject CN or one of its subjectAltName match
the hostname. If this succeeds, then the verification succeeds.
Otherwise openssl_verify() checks the cert subject if this was set,
which means it checks the certificate's subject (not just its CN) matches
exactly the cert subject string that is set in SpiceSession.
Given that the cert subject is something the user specifies in addition
to the hostname, the cert subject check should have priority over the
hostname check, that is, when we have a cert subject set, the
success/failure of the cert subject cert should determine the
success/failure of openssl_verify(), and the hostname check
should only be carried out when no cert subject was set.
This fixes rhbz#871034
We currently log an error when openssl_verify() is called with
preverify_ok set to 0 for all certificates in the certificate chain
except for the peer certificate (when 'depth' is 0).
This commit logs an error in the latter case as well.
I don't see why the fallback method should be the default.
glCopyTexImage2D and glCopyPixels are equally broken with current Intel
mesa/drivers (I'll be filing a bug). But sw rendering is correct.
We can avoid repetitive computation by using two precomputed array, of
8k each.
before:
1.79% lt-spicy-stats libspice-client-glib-2.0.so.8.4.0 [.]
golomb_code_len_8bpc
after:
0.79% lt-spicy-stats libspice-client-glib-2.0.so.8.4.0 [.]
golomb_code_len_8bpc
Thos function shows up in some profiling results, it seems we can
trivially replace it with a precomputed array of 256bytes.
before:
5.66% 691 lt-spicy-stats
libspice-client-glib-2.0.so.8.4.0 [.] revers_bits
after:
0.53% 64 lt-spicy-stats
libspice-client-glib-2.0.so.8.4.0 [.] revers_bits
We currently use it only on gcc 4.5 or newer, but it was actually
introduced much earlier than that. It's documented in gcc 2.95.3
manual:
http://gcc.gnu.org/onlinedocs/gcc-2.95.3/gcc_4.html#SEC84
and glib uses starting from gcc 2.2.5.
This commit uses the same minimum version as glib.
This was causing warnings on RHEL6 systems which have gcc 4.4.7
SPICE_MSG_PLAYBACK_LATENCY is intended for adjusting the
latency of the audio playback. It is used for synchronizing
the audio and video playback.
The corresponding capability is SPICE_PLAYBACK_CAP_LATENCY.
If the server & client support SPICE_DISPLAY_CAP_STREAM_REPORT,
the server first sends SPICE_MSG_DISPLAY_STREAM_ACTIVATE_REPORT. Then,
the client periodically sends SPICE_MSGC_DISPLAY_STREAM_REPORT
messages that supply the server details about the current quality of
the video streaming on the client side. The server analyses the
report and adjust the stream parameters accordingly.
In addition to Laszlo's commit fixing rhbz#928973, we can add
some compile-time assertion to lz_common.h to help catch similar
bugs in the future.
This uses gnulib's verify.h to make sure at compile-time that
the various constant arrays used in lz_common.h are of the expected
size.
I've checked that before Laszlo's patch, the assert triggers, and
that it's gone after it.
