Otherwise, log_netdev_xyz() does not provide netdev name if it is called
in done(). It is hard to debug.
This should not change any effective behavior, at least with the current
implementation of done() per netdev kind.
When we receive a goodby packet about a host, and we have a cache entry about
the host, we do not immediately remove the cache entry, but update it with TTL 1.
See RFC 6762 section 10.1 and 3755027c2c.
If we receive a request soon after the goodby packet, previously the
entry was included in the known answers section of the reply. But such
information should not be appended.
Follow-up for 3755027c2c.
Currently, link_queue_request_safe(), which is a wrapper of
request_new(), is called with a free function at
- link_request_stacked_netdev() at netdev/netdev.c,
- link_request_address() at networkd-address.c,
- link_request_nexthop() at networkd-nexthop.c,
- link_request_neighbor() at networkd-networkd.c.
For the netdev case, the reference counter of the passed object is increased
only when the function returns 1. So, on failure (with -ENOMEM)
previously we unexpectedly dropped the reference of the NetDev object.
Similarly, for Address and friends, the ownership of the object is moved to the
Request object only when the function returns 1. And on failure, previously
the object was freed twice.
Also, netdev_queue_request(), which is another wrapper of request_new()
potentially leaks memory when the same NetDev object is queued twice.
Fortunately, that should not happen as the function is called only once
per object.
This fixes the above issue, and now the ownership or the reference
counter of the object is changed only when it is succeeded with 1.
Then, replace address_remove_and_drop() with it.
If an address is requested, and the request is already called,
we may not received its reply and notification from the kernel, and
the corresponding address object may not be remmbered. Even in such
case, we need to remove the address, otherwise the address will come
later after the function called.
Otherwise, udev workers cannot detect slow programs invoked by
IMPORT{program}=, PROGRAM=, or RUN=, and whole worker process may be
killed.
Fixes#30436.
Co-authored-by: sushmbha <sushmita.bhattacharya@oracle.com>
Also, calculate the timeout for warning based on the remaining time for
the timeout of the event, rather than the timeout itself.
Currently, udev manager kills the worker if the timeout exceeds. So,
this does not change anything except for the timing of the warning.
Just refactoring and preparation for later commits.
This makes the folloing:
- Each Directory object now has a reference to sd-journal.
- Hence, directory_free(), which is renamed from remove_directory(), can
be called without sd-journal as an argument.
- Introduces hash ops for Directory, so the finalization becomes
slightly simpler.
- Allocate hashmaps that store Directory objects when necessary.
- Split out add_directory_impl().
No functional changes, just refactoring.
The similar check already exists in schedule_post_change().
The function is currently called at two places.
- journal_file_open() in sd-journal:
In this case, if the timer is not set up, then journal_file_post_change()
will be called at the end of journal_file_append_entry(). So, the necessary
task will be done sequentially when an journal entry is stored to the opened
journal file. That is desired when the function is called at outside of the
event loop.
- server_open_journal() in journald:
This is not called after we exit the event loop.
So, we can safely do nothing in the function if the event loop is being
finished or already finished.
Fixes#30644.
Previously, we checked if the address is ready only when it is an IPv6
address. But, IPv4 address never enter tentative state, hence the check
for an IPv4 address is effectively equivalent to the one for IPv6.
So, let's unify the check. This does not change any effective behavior.
Triggering assert_return() should be a bug in general, and we should
really fix that. But, previously, it is hard to notice such bug, as
it was not critical.
This is for making CI or our testing environment fail if we unexpectedly
trigger assert_return(). So, hopefully we can easily find such bugs.
Several test cases intentionally trigger assert_return(). So, to avoid
the entire test fails, this introduces several macros that tentatively
make assert_return() not critical.
- introduce device_should_have_db(),
- split out device_get_db_path(),
- update log messages, especially clarify which stage is failed,
- use _cleanup_(unlink_and_freep) attribute,
- clear existing database file also when failed to create database directory
and when failed to create temporary file.
Setting zero or too short timeout for each uevent is meaningless, and
causes the system fails to boot. Let's refuse such values.
Also, delaying execution of RUN= commands too long also makes many
uevents enter the failed state. So, let's refuse such misconfiguration.
Yu Watanabe
