<varlistentry>
     <term><symbol>SPI_OPT_NONATOMIC</symbol></term>
     <listitem>
      <para>
       Sets the SPI connection to be <firstterm>nonatomic</firstterm>, which
       means that transaction control calls (<function>SPI_commit</function>,
       <function>SPI_rollback</function>) are allowed.  Otherwise,
       calling those functions will result in an immediate error.
      </para>
     </listitem>
    </varlistentry>
   </variablelist>
  </para>

  <para>
   <literal>SPI_connect()</literal> is equivalent to
   <literal>SPI_connect_ext(0)</literal>.
  </para>
 </refsect1>

 <refsect1>
  <title>Return Value</title>

  <variablelist>
   <varlistentry>
    <term><symbol>SPI_OK_CONNECT</symbol></term>
    <listitem>
     <para>
      on success
     </para>
    </listitem>
   </varlistentry>
  </variablelist>

  <para>
   The fact that these functions return <type>int</type>
   not <type>void</type> is historical.  All failure cases are reported
   via <function>ereport</function> or <function>elog</function>.
   (In versions before <productname>PostgreSQL</productname> v10,
   some but not all failures would be reported with a result value
   of <symbol>SPI_ERROR_CONNECT</symbol>.)
  </para>
 </refsect1>
</refentry>

<!-- *********************************************** -->

<refentry id="spi-spi-finish">
 <indexterm><primary>SPI_finish</primary></indexterm>

 <refmeta>
  <refentrytitle>SPI_finish</refentrytitle>
  <manvolnum>3</manvolnum>
 </refmeta>

 <refnamediv>
  <refname>SPI_finish</refname>
  <refpurpose>disconnect a C function from the SPI manager</refpurpose>
 </refnamediv>

 <refsynopsisdiv>
<synopsis>
int SPI_finish(void)
</synopsis>
 </refsynopsisdiv>

 <refsect1>
  <title>Description</title>

  <para>
   <function>SPI_finish</function> closes an existing connection to
   the SPI manager.  You must call this function after completing the
   SPI operations needed during your C function's current invocation.
   You do not need to worry about making this happen, however, if you
   abort the transaction via <literal>elog(ERROR)</literal>.  In that
   case SPI will clean itself up automatically.
  </para>
 </refsect1>

 <refsect1>
  <title>Return Value</title>

  <variablelist>
   <varlistentry>
    <term><symbol>SPI_OK_FINISH</symbol></term>
    <listitem>
     <para>
      if properly disconnected
     </para>
    </listitem>
   </varlistentry>

   <varlistentry>
    <term><symbol>SPI_ERROR_UNCONNECTED</symbol></term>
    <listitem>
     <para>
      if called from an unconnected C function
     </para>
    </listitem>
   </varlistentry>
  </variablelist>
 </refsect1>
</refentry>

<!-- *********************************************** -->

<refentry id="spi-spi-execute">
 <indexterm><primary>SPI_execute</primary></indexterm>

 <refmeta>
  <refentrytitle>SPI_execute</refentrytitle>
  <manvolnum>3</manvolnum>
 </refmeta>

 <refnamediv>
  <refname>SPI_execute</refname>
  <refpurpose>execute a command</refpurpose>
 </refnamediv>

 <refsynopsisdiv>
<synopsis>
int SPI_execute(const char * <parameter>command</parameter>, bool <parameter>read_only</parameter>, long <parameter>count</parameter>)
</synopsis>
 </refsynopsisdiv>

 <refsect1>
  <title>Description</title>

  <para>
   <function>SPI_execute</function> executes the specified SQL command
   for <parameter>count</parameter> rows.  If <parameter>read_only</parameter>
   is <literal>true</literal>, the command must be read-only, and execution overhead
   is somewhat reduced.
  </para>

  <para>
   This function can only be called from a connected C function.
  </para>

  <para>
   If <parameter>count</parameter> is zero then the command is executed
   for all rows that it applies to.  If <parameter>count</parameter>
   is greater than zero, then no more than <parameter>count</parameter> rows
   will be retrieved; execution stops when the count is reached, much like
   adding a <literal>LIMIT</literal> clause to the query. For example,
<programlisting>
SPI_execute("SELECT