Michal Schmidt [Tue, 6 Oct 2015 12:56:33 +0000 (14:56 +0200)]
test: hashmap - increase number of entries for crippled hash test
The purpose of testing with the crippled hash function is to cover
the otherwise very unlikely codepath in bucket_calculate_dib() where
it has to fall back to recomputing the hash value.
This unlikely path was not covered by test-hashmap anymore after 57217c8f "test: hashmap - cripple the hash function by truncating the
input rather than the output".
Restore the test coverage by increasing the number of entries in the test.
The number was determined empirically by checking with lcov.
unit: give systemd-networkd.socket a better description
Usually we try to properly uppercase first characters in the
description, do so here, too. Also, keep it close to the string used in
systemd-networkd.service.
core: add support for naming file descriptors passed using socket activation
This adds support for naming file descriptors passed using socket
activation. The names are passed in a new $LISTEN_FDNAMES= environment
variable, that matches the existign $LISTEN_FDS= one and contains a
colon-separated list of names.
This also adds support for naming fds submitted to the per-service fd
store using FDNAME= in the sd_notify() message.
This also adds a new FileDescriptorName= setting for socket unit files
to set the name for fds created by socket units.
This also adds a new call sd_listen_fds_with_names(), that is similar to
sd_listen_fds(), but also returns the names of the fds.
systemd-activate gained the new --fdname= switch to specify a name for
testing socket activation.
Jacob Keller [Mon, 5 Oct 2015 21:14:45 +0000 (14:14 -0700)]
document ability to disable MACAddressPolicy
While it is currently possible to either not set MACAddressPolicy or set
it to a value different from "persistent" or "random", it is not obvious
that a user can do so. Add a policy, "none", which simply retains kernel
MAC addresses (same as not filling in the policy at all) and document it
so that users are aware of this setting.
Signed-off-by: Jacob Keller <jacob.keller@gmail.com>
Kay Sievers [Mon, 5 Oct 2015 22:41:32 +0000 (00:41 +0200)]
libsystemd: sd-device - translate / vs. ! in sysname
The kernel replaces '/' in device names with '!', we translate that back
to '/' in sysname, when taking sysname as input, we should translate it
back again.
Tom Gundersen [Sat, 3 Oct 2015 22:22:41 +0000 (00:22 +0200)]
hashmap: refactor hash_func
All our hash functions are based on siphash24(), factor out
siphash_init() and siphash24_finalize() and pass the siphash
state to the hash functions rather than the hash key.
This simplifies the hash functions, and in particular makes
composition simpler as calling siphash24_compress() repeatedly
on separate chunks of input has the same effect as first
concatenating the input and then calling siphash23_compress()
on the result.
Tom Gundersen [Sat, 3 Oct 2015 23:14:41 +0000 (01:14 +0200)]
test: hashmap - cripple the hash function by truncating the input rather than the output
The reason for the crippled hash function is to reduce the distribution
of the hash function, do this by truncating the domain rather than the
range. This does introduce a change in behavoir as the range is no longer
contiguous, which greatly reduces collisions.
This is needed as a follow-up patch will no longer allow individual hash
functions to alter the output directly.
Tom Gundersen [Sat, 3 Oct 2015 22:24:23 +0000 (00:24 +0200)]
test: siphash24 - verify internal state and composability
Verify the state of the hash-function according to the reference paper,
also verify that we can decompose the input and hash the chunks one
by one and still get the same result.
Tom Gundersen [Sat, 3 Oct 2015 19:27:25 +0000 (21:27 +0200)]
siphash24: move last compression iteration from compression step to finalization step
The last compression is special as it deals with the length byte, and padding. Move
it to the finalization step in preparation for making compression decomposable.
core: don't unset reload result, unless we begin a start or reload operation
Much like the result of the service itself we should not reset the
reload result unless we actually start from the beginning, so that
clients can query it at any time.
Specifically, let's reset the result states only when we begin with a
start operation (for both the main result, and the reload result), when
we begin with a reload operation (only for the load result), or when the
use explicitly asks for that via "systemctl reset-failed".
Implement a maximum limit on number of journal files to keep around.
Enforcing a limit is useful on this since our performance when viewing
pays a heavy penalty for each journal file to interleve. This setting is
turned on now by default, and set to 100.
Also, actully implement what 348ced909724a1331b85d57aede80a102a00e428
promised: use whatever we find on disk at startup as lower bound on how
much disk space we can use. That commit introduced some provisions to
implement this, but actually never did.
This also adds "journalctl --vacuum-files=" to vacuum files on disk by
their number explicitly.
journal: don't affect atime of journal files when vacuuming
Let's try to use O_NOATIME if we can when vacuuming old journal files,
if we have the permissions for it, so that vacuuming doesn't count as
proper journal read access.
tlv_packet_read_bytes() and tlv_packet_read_string() returned the
wrong length when called after other functions which modify the offset
in the container.
In other words, if the TLV data length is X and we do a
tlv_packet_read_u8(), a subsequent tlv_packet_read_bytes() should
return a length of (X - 1).
lldp: move lldp_receive_packet() to lldp-internal.c
In order to implement tests for the LLDP state machine, we need to
mock lldp_network_bind_raw_socket(). Move the other function
lldp_receive_packet() to another file so that we can replace the first
function with a custom one and keep the second one.
It can be useful to know the destination address of a LLDP frame
because it determines the scope of propagation of the frame and thus
this information be used to know whether the neighbor is connected to
the same physical link.
lldp: add public function to export LLDP TLV packets
Add a public function to get a list of current LLDP neighbours' TLV
packets. The function populates an array of pointers to the opaque
type sd_lldp_packet and returns the number of elements found. Callers
must take care of freeing the array and decreasing the refcount of
elements when done.
The internal speaker is usually not available on modern latops that
support suspend, and even if it is available in the hardware, most
distributions turned support for it off in the kernel. And even if it is
enabled, it's probably still a bad idea to make use of it for the
suspend-failures. If anything a proper sound should be played.
Long story short, let's remove support of this anachronism.
With this rework we introduce systemd-rfkill.service as singleton that
is activated via systemd-rfkill.socket that listens on /dev/rfkill. That
way, we get notified each time a new rfkill device shows up or changes
state, in which case we restore and save its current setting to disk.
This is nicer than the previous logic, as this means we save/restore
state even of rfkill devices that are around only intermittently, and
save/restore the state even if the system is shutdown abruptly instead
of cleanly.
This implements what I suggested in #1019 and obsoletes it.
core: add new setting Writable= to ListenSpecial= socket units
Writable= is a new boolean setting. If ture, then ListenSpecial= will
open the specified path in O_RDWR mode, rather than just O_RDONLY.
This is useful for implementing services like rfkill, where /dev/rfkill
is more useful when opened in write mode, if we want to not only save
but also restore its state.