systemd/man/sd_journal_get_fd.xml
Lennart Poettering 64a7ef8bc0 man: be more explicit about thread safety of sd_journal
Triggered by https://bugzilla.redhat.com/show_bug.cgi?id=1609349

This adds two generic paragaphs we include via xinclude. One is the
"strict" version, which contains wording saying that we are thread
agnostic and what that means. And the other is the "safe" version, for
the cases we provide fully safety.

Let's then change most man pages to use either of these generic
paragraphs. With one exception: man/sd_journal_get_catalog.xml contains
both kinds of function, we hence use manual wording.
2018-08-03 17:36:11 +02:00

263 lines
12 KiB
XML

<?xml version='1.0'?> <!--*-nxml-*-->
<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">
<!--
SPDX-License-Identifier: LGPL-2.1+
-->
<refentry id="sd_journal_get_fd" xmlns:xi="http://www.w3.org/2001/XInclude">
<refentryinfo>
<title>sd_journal_get_fd</title>
<productname>systemd</productname>
</refentryinfo>
<refmeta>
<refentrytitle>sd_journal_get_fd</refentrytitle>
<manvolnum>3</manvolnum>
</refmeta>
<refnamediv>
<refname>sd_journal_get_fd</refname>
<refname>sd_journal_get_events</refname>
<refname>sd_journal_get_timeout</refname>
<refname>sd_journal_process</refname>
<refname>sd_journal_wait</refname>
<refname>sd_journal_reliable_fd</refname>
<refname>SD_JOURNAL_NOP</refname>
<refname>SD_JOURNAL_APPEND</refname>
<refname>SD_JOURNAL_INVALIDATE</refname>
<refpurpose>Journal change notification
interface</refpurpose>
</refnamediv>
<refsynopsisdiv>
<funcsynopsis>
<funcsynopsisinfo>#include &lt;systemd/sd-journal.h&gt;</funcsynopsisinfo>
<funcprototype>
<funcdef>int <function>sd_journal_get_fd</function></funcdef>
<paramdef>sd_journal *<parameter>j</parameter></paramdef>
</funcprototype>
<funcprototype>
<funcdef>int <function>sd_journal_get_events</function></funcdef>
<paramdef>sd_journal *<parameter>j</parameter></paramdef>
</funcprototype>
<funcprototype>
<funcdef>int <function>sd_journal_get_timeout</function></funcdef>
<paramdef>sd_journal *<parameter>j</parameter></paramdef>
<paramdef>uint64_t *<parameter>timeout_usec</parameter></paramdef>
</funcprototype>
<funcprototype>
<funcdef>int <function>sd_journal_process</function></funcdef>
<paramdef>sd_journal *<parameter>j</parameter></paramdef>
</funcprototype>
<funcprototype>
<funcdef>int <function>sd_journal_wait</function></funcdef>
<paramdef>sd_journal *<parameter>j</parameter></paramdef>
<paramdef>uint64_t <parameter>timeout_usec</parameter></paramdef>
</funcprototype>
<funcprototype>
<funcdef>int <function>sd_journal_reliable_fd</function></funcdef>
<paramdef>sd_journal *<parameter>j</parameter></paramdef>
</funcprototype>
</funcsynopsis>
</refsynopsisdiv>
<refsect1>
<title>Description</title>
<para><function>sd_journal_get_fd()</function> returns a file
descriptor that may be asynchronously polled in an external event
loop and is signaled as soon as the journal changes, because new
entries or files were added, rotation took place, or files have
been deleted, and similar. The file descriptor is suitable for
usage in
<citerefentry><refentrytitle>poll</refentrytitle><manvolnum>2</manvolnum></citerefentry>.
Use <function>sd_journal_get_events()</function> for an events
mask to watch for. The call takes one argument: the journal
context object. Note that not all file systems are capable of
generating the necessary events for wakeups from this file
descriptor for changes to be noticed immediately. In particular
network files systems do not generate suitable file change events
in all cases. Cases like this can be detected with
<function>sd_journal_reliable_fd()</function>, below.
<function>sd_journal_get_timeout()</function> will ensure in these
cases that wake-ups happen frequently enough for changes to be
noticed, although with a certain latency.</para>
<para><function>sd_journal_get_events()</function> will return the
<function>poll()</function> mask to wait for. This function will
return a combination of <constant>POLLIN</constant> and
<constant>POLLOUT</constant> and similar to fill into the
<literal>.events</literal> field of <varname>struct
pollfd</varname>.</para>
<para><function>sd_journal_get_timeout()</function> will return a
timeout value for usage in <function>poll()</function>. This
returns a value in microseconds since the epoch of
<constant>CLOCK_MONOTONIC</constant> for timing out
<function>poll()</function> in <varname>timeout_usec</varname>.
See
<citerefentry><refentrytitle>clock_gettime</refentrytitle><manvolnum>2</manvolnum></citerefentry>
for details about <constant>CLOCK_MONOTONIC</constant>. If there
is no timeout to wait for, this will fill in <constant>(uint64_t)
-1</constant> instead. Note that <function>poll()</function> takes
a relative timeout in milliseconds rather than an absolute timeout
in microseconds. To convert the absolute 'us' timeout into
relative 'ms', use code like the following:</para>
<programlisting>uint64_t t;
int msec;
sd_journal_get_timeout(m, &amp;t);
if (t == (uint64_t) -1)
msec = -1;
else {
struct timespec ts;
uint64_t n;
clock_gettime(CLOCK_MONOTONIC, &amp;ts);
n = (uint64_t) ts.tv_sec * 1000000 + ts.tv_nsec / 1000;
msec = t > n ? (int) ((t - n + 999) / 1000) : 0;
}</programlisting>
<para>The code above does not do any error checking for brevity's
sake. The calculated <varname>msec</varname> integer can be passed
directly as <function>poll()</function>'s timeout
parameter.</para>
<para>After each <function>poll()</function> wake-up
<function>sd_journal_process()</function> needs to be called to
process events. This call will also indicate what kind of change
has been detected (see below; note that spurious wake-ups are
possible).</para>
<para>A synchronous alternative for using
<function>sd_journal_get_fd()</function>,
<function>sd_journal_get_events()</function>,
<function>sd_journal_get_timeout()</function> and
<function>sd_journal_process()</function> is
<function>sd_journal_wait()</function>. It will synchronously wait
until the journal gets changed. The maximum time this call sleeps
may be controlled with the <parameter>timeout_usec</parameter>
parameter. Pass <constant>(uint64_t) -1</constant> to wait
indefinitely. Internally this call simply combines
<function>sd_journal_get_fd()</function>,
<function>sd_journal_get_events()</function>,
<function>sd_journal_get_timeout()</function>,
<function>poll()</function> and
<function>sd_journal_process()</function> into one.</para>
<para><function>sd_journal_reliable_fd()</function> may be used to
check whether the wakeup events from the file descriptor returned
by <function>sd_journal_get_fd()</function> are known to be
immediately triggered. On certain file systems where file change
events from the OS are not available (such as NFS) changes need to
be polled for repeatedly, and hence are detected only with a
certain latency. This call will return a positive value if the
journal changes are detected immediately and zero when they need
to be polled for and hence might be noticed only with a certain
latency. Note that there is usually no need to invoke this function
directly as <function>sd_journal_get_timeout()</function> on these
file systems will ask for timeouts explicitly anyway.</para>
</refsect1>
<refsect1>
<title>Return Value</title>
<para><function>sd_journal_get_fd()</function> returns a valid
file descriptor on success or a negative errno-style error
code.</para>
<para><function>sd_journal_get_events()</function> returns a
combination of <constant>POLLIN</constant>,
<constant>POLLOUT</constant> and suchlike on success or a negative
errno-style error code.</para>
<para><function>sd_journal_reliable_fd()</function> returns a
positive integer if the file descriptor returned by
<function>sd_journal_get_fd()</function> will generate wake-ups
immediately for all journal changes. Returns 0 if there might be a
latency involved.</para>
<para><function>sd_journal_process()</function> and <function>sd_journal_wait()</function> return a negative
errno-style error code, or one of <constant>SD_JOURNAL_NOP</constant>, <constant>SD_JOURNAL_APPEND</constant> or
<constant>SD_JOURNAL_INVALIDATE</constant> on success:</para>
<itemizedlist>
<listitem><para>If <constant>SD_JOURNAL_NOP</constant> is returned, the journal did not change since the last
invocation.</para></listitem>
<listitem><para>If <constant>SD_JOURNAL_APPEND</constant> is returned, new entries have been appended to the end
of the journal. In this case it is sufficient to simply continue reading at the previous end location of the
journal, to read the newly added entries.</para></listitem>
<listitem><para>If <constant>SD_JOURNAL_INVALIDATE</constant>, journal files were added to or removed from the
set of journal files watched (e.g. due to rotation or vacuuming), and thus entries might have appeared or
disappeared at arbitrary places in the log stream, possibly before or after the previous end of the log
stream. If <constant>SD_JOURNAL_INVALIDATE</constant> is returned, live-view UIs that want to reflect on screen
the precise state of the log data on disk should probably refresh their entire display (relative to the cursor of
the log entry on the top of the screen). Programs only interested in a strictly sequential stream of log data may
treat <constant>SD_JOURNAL_INVALIDATE</constant> the same way as <constant>SD_JOURNAL_APPEND</constant>, thus
ignoring any changes to the log view earlier than the old end of the log stream.</para></listitem>
</itemizedlist>
</refsect1>
<refsect1>
<title>Signal safety</title>
<para>In general, <function>sd_journal_get_fd()</function>, <function>sd_journal_get_events()</function>, and
<function>sd_journal_get_timeout()</function> are <emphasis>not</emphasis> "async signal safe" in the meaning of
<citerefentry
project='man-pages'><refentrytitle>signal-safety</refentrytitle><manvolnum>7</manvolnum></citerefentry>.
Nevertheless, only the first call to any of those three functions performs unsafe operations, so subsequent calls
<emphasis>are</emphasis> safe.</para>
<para><function>sd_journal_process()</function> and <function>sd_journal_wait()</function> are not
safe. <function>sd_journal_reliable_fd()</function> is safe.</para>
</refsect1>
<refsect1>
<title>Notes</title>
<xi:include href="threads-aware.xml" xpointer="strict"/>
<xi:include href="libsystemd-pkgconfig.xml" xpointer="pkgconfig-text"/>
</refsect1>
<refsect1>
<title>Examples</title>
<para>Iterating through the journal, in a live view tracking all
changes:</para>
<programlisting><xi:include href="journal-iterate-wait.c" parse="text" /></programlisting>
<para>Waiting with <function>poll()</function> (this
example lacks all error checking for the sake of
simplicity):</para>
<programlisting><xi:include href="journal-iterate-poll.c" parse="text" /></programlisting>
</refsect1>
<refsect1>
<title>See Also</title>
<para>
<citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>,
<citerefentry><refentrytitle>sd-journal</refentrytitle><manvolnum>3</manvolnum></citerefentry>,
<citerefentry><refentrytitle>sd_journal_open</refentrytitle><manvolnum>3</manvolnum></citerefentry>,
<citerefentry><refentrytitle>sd_journal_next</refentrytitle><manvolnum>3</manvolnum></citerefentry>,
<citerefentry><refentrytitle>poll</refentrytitle><manvolnum>2</manvolnum></citerefentry>,
<citerefentry><refentrytitle>clock_gettime</refentrytitle><manvolnum>2</manvolnum></citerefentry>
</para>
</refsect1>
</refentry>