`uri` is always non-`NULL` by the time the `fail` label is reached, so
drop the `NULL` pointer check. Inline the `fail` code since it’s only
used from two places.
Coverity CID: #1430970
Signed-off-by: Philip Withnall <withnall@endlessm.com>
For interoperability with libdbus, we want to use compatible timeouts.
In particular, this fixes a spurious failure of the `gdbus-server-auth`
test caused by libdbus and gdbus choosing to expire the key (cookie) at
different times, as diagnosed by Thiago Macieira. Previously, the libdbus
client would decline to use keys older than 7 minutes, but the GDBus
server would not generate a new key until the old key was 10 minutes old.
For completeness, also adjust the other arbitrary timeouts in the
DBUS_COOKIE_SHA1 mechanism to be the same as in libdbus. To make it
easier to align with libdbus, create internal macros with the same names
and values used in dbus-keyring.c.
* maximum time a key can be in the future due to clock skew between
systems sharing a home directory
- spec says "a reasonable time in the future"
- was 1 day
- now 5 minutes
- MAX_TIME_TRAVEL_SECONDS
* time to generate a new key if the newest is older
- spec says "If no recent keys remain, the server may generate a new
key", but that isn't practical, because in reality we need a grace
period during which an old key will not be used for new authentication
attempts but old authentication attempts can continue (in practice both
libdbus and GDBus implemented this logic)
- was 10 minutes
- now 5 minutes
- NEW_KEY_TIMEOUT_SECONDS
* time to discard old keys
- spec says "the timeout can be fairly short"
- was 15 minutes
- now 7 minutes
- EXPIRE_KEYS_TIMEOUT_SECONDS
* time allowed for a client using an old key to authenticate, before
that key gets deleted
- was at least 5 minutes
- now at least 2 minutes
- at least (EXPIRE_KEYS_TIMEOUT_SECONDS - NEW_KEY_TIMEOUT_SECONDS)
Based on a merge request by Philip Withnall.
Fixes: #2164
Thanks: Philip Withnall
Thanks: Thiago Macieira
Signed-off-by: Simon McVittie <smcv@collabora.com>
If one thread pool thread fails to set its scheduler settings, it’s
likely that all the rest of them will fail for the same reason. Avoid
printing duplicate critical warnings in that case.
Signed-off-by: Philip Withnall <withnall@endlessm.com>
Fixes: #2191
If the calling code adds more option entries than `G_MAXSIZE` then
there’ll be an integer overflow. This seems vanishingly unlikely (given
that all callers use static option entry lists), but add a precondition
anyway.
Signed-off-by: Philip Withnall <withnall@endlessm.com>
Fixes: #2197
This doesn't trigger the cancellation assertion issue when run locally
(the task didn't return yet, so the error is simply overwritten), but
perhaps it ever does in CI. Anyhow, it's good to have a cancellation
test.
After a splice operation is finished, it attempts to 1) close input/output
streams, as per the given flags, and 2) return the operation result (maybe
an error, too).
However, if the operation gets cancelled early and the streams indirectly
closed, the splice operation will try to close both descriptors and return
on the task when both are already closed. The catch here is that getting the
streams closed under its feet is possible, so the completion callback would
find both streams closed after returning on the first close operation and
return the error, but then the second operation could be able to trigger
a second error which would be returned as well.
What happens here is up to further race conditions, if the task didn't
return yet, the returned error will be simply replaced (but the old one not
freed...), if it did already return, it'll result in:
GLib-GIO-FATAL-CRITICAL: g_task_return_error: assertion '!task->ever_returned' failed
Fix this by flagging the close_async() callbacks, and checking that both
close operations did return, instead of checking that both streams are
closed by who knows.
This error triggers a semi-frequent CI failure in tracker, see the summary at
https://gitlab.gnome.org/GNOME/tracker/-/issues/240
Behaviour in that case is implementation-defined and how many bytes were
actually written can't be expressed by the return value anymore.
Instead do a short write of G_MAXSSIZE bytes and let the existing loop
for handling short writes takes care of the remaining length.
glib/gfileutils.c: In function ‘write_to_file’:
glib/gfileutils.c:1176:19: error: comparison of integer expressions of different signedness: ‘gssize’ {aka ‘long int’} and ‘gsize’ {aka ‘long unsigned int’} [-Werror=sign-compare]
1176 | g_assert (s <= length);
| ^~
glib/gmacros.h:939:25: note: in definition of macro ‘G_LIKELY’
939 | #define G_LIKELY(expr) (expr)
| ^~~~
glib/gfileutils.c:1176:7: note: in expansion of macro ‘g_assert’
1176 | g_assert (s <= length);
| ^~~~~~~~
Related to issue #1735 (Get back to a -werror build)
statx does not provide stx_atime when querying a file in a read-only
mounted file system. So call to statx should not expect it to be in
the mask. Otherwise we would fail with ERANGE for querying any file in
a read-only file system.
Fixes#2189.
