This function enumerates all user-accessible UWP packages
and calls the user-provided callback for each package.
This can be used to make GLib aware of the UWP applications
installed in the system.
The function works by using IPackageManager/IPackage UWP interfaces
and XmlLite COM library to parse package manifests.
The function requires COM, and initializes it to a single-thread
appartment model. To ensure this doesn't break anything, either
only use it in a separate thread (COM is initialized on a per-thread
basis), or make sure that the main thread also uses the same COM
model (it's OK to initialize COM multiple times, as long as the same
model is used and as long as init/uninit calls are paired correctly).
MinGW-w64 lacks the appropriate headers, so we have to add them
here. Note that these only have the C versions (normally these
things come in both C and C++ flavours), since that's what we use.
Also note that some of the functions that we don't use (but must
describe to maintain binary compatibility) were altered to use
IUnknown (basically, an untyped pointer) instead of the appropriate
object types, as adding these types would require other types,
which would pull even more types, forcing us to drag half of the
UWP headers in here. By replacing unused types with IUnknown we
can trim a lot of branches from the dependency graph.
This is a COM object that implements IStream by using a HANDLE
and WinAPI file functions to access the file (only a file; pipes
are not supported). Only supports synchronous access (this is
a feature - the APIs that read from this stream internally will
never return the COM equivalent of EWOULDBLOCK, which greatly
simplifies their use).
Set counters for the number of running tasks and
for the max. threadpool size. These are meant to
get a sense for whether G_TASK_POOL_SIZE and related
constants are still suitable for current gio and
GTask usage patterns.
Previously it was considered a programming error to call these on
subprocesses created without the correct flags, but for bindings this
distinction is difficult to handle automatically.
Returning NULL instead does not cause any inconsistent behaviour and
simplifies the API.
As hidden file caches currently work, every look up on a directory caches
its .hidden file contents, and sets a 5s timeout to prune the directory
from the cache.
This creates a problem for usecases like Tracker Miners, which is in the
business of inspecting as many files as possible from as many directories
as possible in the shortest time possible. One timeout is created for each
directory, which possibly means gobbling thousands of entries in the hidden
file cache. This adds as many GSources to the glib worker thread, with the
involved CPU overhead in iterating those in its main context.
To fix this, use a unique timeout that will keep running until the cache
is empty. This will keep the overhead constant with many files/folders
being queried.
This is a regression from !1686. The tmp_error is no longer valid after
it is "considered" and cannot be used at this point. We should print the
error earlier instead.
Fixes#2233
This incidentally also exercises the intended pattern for sending fds in
a D-Bus message: the fd list is meant to contain exactly those fds that
are referenced by a handle (type 'h') in the body of the message, with
numeric handle value n corresponding to g_unix_fd_list_peek_fds(...)[n].
Being able to send and receive file descriptors that are not referenced by
a handle (as in OpenFile here) is a quirk of the GDBus API, and while it's
entirely possible in the wire protocol, other D-Bus implementations like
libdbus and sd-bus typically don't provide APIs that make this possible.
Reproduces: https://gitlab.gnome.org/GNOME/glib/-/issues/2074
Signed-off-by: Simon McVittie <smcv@collabora.com>
In the D-Bus wire protocol, the handle type (G_VARIANT_TYPE_HANDLE, h)
is intended to be an index/pointer into the implementation's closest
equivalent of GUnixFDList: its numeric value has no semantic meaning
(in the same way that the numeric values of pointers have no semantic
meaning), but a handle with value n acts as a reference to the nth fd
in the fd list.
GDBus provides a fairly direct mapping from the wire protocol to the
C API, which makes it technically possible to attach and use fds
without ever referring to them in the message body, and some
GLib-centric D-Bus APIs rely on this.
However, the other major implementations of D-Bus (libdbus and sd-bus)
transparently replace file descriptors with handles when building
messages, and transparently replace handles with file descriptors when
parsing messages. This means they cannot implement D-Bus APIs that do
not follow the conventional meaning of handles as indexes/pointers into
an equivalent of GUnixFDList.
For interoperability, we should encourage D-Bus API designers to follow
the convention, even though code written against GDBus doesn't strictly
need to do so.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Suppose we are sending a 5K message with fds (so data->blob points
to 5K of data, data->blob_size is 5K, and fd_list is non-null), but
the kernel is only accepting up to 4K with each sendmsg().
The first time we get into write_message_continue_writing(),
data->total_written will be 0. We will try to write the entire message,
plus the attached file descriptors; or if the stream doesn't support
fd-passing (not a socket), we need to fail with
"Tried sending a file descriptor on unsupported stream".
Because the kernel didn't accept the entire message, we come back in.
This time, we won't enter the Unix-specific block that involves sending
fds, because now data->total_written is 4K, and it would be wrong to try
to attach the same fds again. However, we also need to avoid failing
with "Tried sending a file descriptor on unsupported stream" in this
case. We just want to write out the data of the rest of the message,
starting from (blob + total_written) (in this exaple, the last 1K).
Resolves: https://gitlab.gnome.org/GNOME/glib/-/issues/2074
Signed-off-by: Simon McVittie <smcv@collabora.com>
This test ensures that g_socket_client_connect_to_host_async() fails if
it is cancelled, but it's not cancelled until after 1 millisecond. Our
CI testers are hitting that race window, and Milan is able to reproduce
the crash locally as well. Switching it from 1ms to 0ms is enough for
Milan to avoid the crash, but not enough for our CI, so let's move the
cancellation to a GSocketClientEvent callback where the timing is
completely deterministic.
Hopefully fixes#2221
g_has_typeof macro is wrongly in the public g_ namespace, internaly
symbols are usually in the glib_ namespace. This will also allow to
define glib_typeof differently on non-GNUC compilers (e.g. c++11
decltype).
This introduces no functional changes, but makes the refcount handling a
little easier to follow by no longer splitting a ref/unref pair across
three callbacks. Now, the ref/unref pairs are all within function-local
scopes.
Coverity CID: #1430783
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
By default, when using g_subprocess_launcher_take_fd() to pass an
FD to a child, the GSubprocessLauncher object also takes ownership
of the FD in the parent, and closes it during finalize(). This is
a reasonable assumption in the majority of the cases, but sometimes
it isn't a good idea.
An example is when creating a GSubprocessLauncher in JavaScript:
here, the destruction process is managed by the Garbage Collector,
which means that those sockets will remain opened for some time
after all the references to the object has been droped. This means
that it could be not possible to detect when the child has closed
that same FD, because in order to make that work, both FDs
instances (the one in the parent and the one in the children) must
be closed. This can be a problem in, as an example, a process that
launches a child that communicates with Wayland using an specific
socket (like when using the new API MetaWaylandClient).
Of course, it isn't a valid solution to manually call close() in
the parent process just after the call to spawn(), because the FD
number could be reused in the time between it is manually closed,
and when the object is destroyed and closes again that FD. If that
happens, it will close an incorrect FD.
One solution could be to call run_dispose() from Javascript on the
GSubprocessLauncher object, to force freeing the resources.
Unfortunately, the current code frees them in the finalize()
method, not in dispose() (this is fixed in !1670 (merged) ) but it
isn't a very elegant solution.
This proposal adds a new method, g_subprocess_launcher_close(),
that allows to close the FDs passed to the child. To avoid problems,
after closing an FD with this method, no more spawns are allowed.
Fix: https://gitlab.gnome.org/GNOME/glib/-/merge_requests/1677
Originally, GSocketClient returned whatever error occured last. Turns
out this doesn't work well in practice. Consider the following case:
DNS returns an IPv4 and IPv6 address. First we'll connect() to the
IPv4 address, and say that succeeds, but TLS is enabled and the TLS
handshake fails. Then we try the IPv6 address and receive ENETUNREACH
because IPv6 isn't supported. We wind up returning NETWORK_UNREACHABLE
even though the address can be pinged and a TLS error would be more
appropriate. So instead, we now try to return the error corresponding
to the latest attempted GSocketClientEvent in the connection process.
TLS errors take precedence over proxy errors, which take precedence
over connect() errors, which take precedence over DNS errors.
In writing this commit, I made several mistakes that were caught by
proxy-test.c, which tests using GSocketClient to make a proxy
connection. So although adding a new test to ensure we get the
best-possible error would be awkward, at least we have some test
coverage for the code that helped avoid introducing bugs.
Fixes#2211
We should never return unknown errors to the application. This would be
a glib bug.
I don't think it's currently possible to hit these cases, so asserts
should be OK. For this to happen, either (a) a GSocketAddressEnumerator
would have to return NULL on its first enumeration, without returning an
error, or (b) there would have to be a bug in our GSocketClient logic.
Either way, if such a bug were to exist, it would be better to surface
it rather than hide it.
These changes are actually going to be effectively undone in a
subsequent commit, as I'm refactoring the error handling, but the commit
history is a bit nicer with two separate commits, so let's go with two.
GSocketAddressEnumerator encapsulates the details of how DNS happens, so
we don't have to think about it. But we may have taken encapsulation a
bit too far, here. Usually, we resolve a domain name to a list of IPv4
and IPv6 addresses. Then we go through each address in the list and try
to connect to it. Name resolution happens exactly once, at the start.
It doesn't happen each time we enumerate the enumerator. In theory, it
*could*, because we've designed these APIs to be agnostic of underlying
implementation details like DNS and network protocols. But in practice,
we know that's not really what's happening. It's weird to say that we
are RESOLVING what we know to be the same name multiple times. Behind
the scenes, we're not doing that.
This also fixes#1994, where enumeration can end with a RESOLVING event,
even though this is supposed to be the first event rather than the last.
I thought this would be hard to fix, even requiring new public API in
GSocketAddressEnumerator to peek ahead to see if the next enumeration is
going to return NULL. Then I decided we should just fake it: always emit
both RESOLVING and RESOLVED at the same time right after each
enumeration. Finally, I realized we can emit them at the correct time if
we simply assume resolving only happens the first time. This seems like
the most elegant of the possible solutions.
Now, this is a behavior change, and arguably an API break, but it should
align better with reasonable expectations of how GSocketClientEvent
ought to work. I don't expect it to break anything besides tests that
check which order GSocketClientEvent events are emitted in. (Currently,
libsoup has such tests, which will need to be updated.) Ideally we would
have GLib-level tests as well, but in a concession to pragmatism, it's a
lot easier to keep network tests in libsoup.
This isn't an API guarantee, but it's a potentially-surprising
behavior difference between the sync and async functions that is good
to know about, especially because our sync and async functions are
normally identical.
The linux kernel does not know that the socket will be used
for connect or listen and if you bind() to a local address it must
reserve a random port (if port == 0) at bind() time, making very easy
to exhaust the ~32k port range, setting IP_BIND_ADDRESS_NO_PORT tells
the kernel to choose random port at connect() time instead, when the
full 4-tuple is known.
