kernel / Documentation / DocBook / drm.tmpl

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN"
	"http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" []>

<book id="drmDevelopersGuide">
  <bookinfo>
    <title>Linux DRM Developer's Guide</title>

    <authorgroup>
      <author>
	<firstname>Jesse</firstname>
	<surname>Barnes</surname>
	<contrib>Initial version</contrib>
	<affiliation>
	  <orgname>Intel Corporation</orgname>
	  <address>
	    <email>[e-mail]</email>
	  </address>
	</affiliation>
      </author>
      <author>
	<firstname>Laurent</firstname>
	<surname>Pinchart</surname>
	<contrib>Driver internals</contrib>
	<affiliation>
	  <orgname>Ideas on board SPRL</orgname>
	  <address>
	    <email>[e-mail]</email>
	  </address>
	</affiliation>
      </author>
    </authorgroup>

    <copyright>
      <year>2008-2009</year>
      <year>2012</year>
      <holder>Intel Corporation</holder>
      <holder>Laurent Pinchart</holder>
    </copyright>

    <legalnotice>
      <para>
	The contents of this file may be used under the terms of the GNU
	General Public License version 2 (the "GPL") as distributed in
	the kernel source COPYING file.
      </para>
    </legalnotice>

    <revhistory>
      <!-- Put document revisions here, newest first. -->
      <revision>
	<revnumber>1.0</revnumber>
	<date>2012-07-13</date>
	<authorinitials>LP</authorinitials>
	<revremark>Added extensive documentation about driver internals.
	</revremark>
      </revision>
    </revhistory>
  </bookinfo>

<toc></toc>

  <!-- Introduction -->

  <chapter id="drmIntroduction">
    <title>Introduction</title>
    <para>
      The Linux DRM layer contains code intended to support the needs
      of complex graphics devices, usually containing programmable
      pipelines well suited to 3D graphics acceleration.  Graphics
      drivers in the kernel may make use of DRM functions to make
      tasks like memory management, interrupt handling and DMA easier,
      and provide a uniform interface to applications.
    </para>
    <para>
      A note on versions: this guide covers features found in the DRM
      tree, including the TTM memory manager, output configuration and
      mode setting, and the new vblank internals, in addition to all
      the regular features found in current kernels.
    </para>
    <para>
      [Insert diagram of typical DRM stack here]
    </para>
  </chapter>

  <!-- Internals -->

  <chapter id="drmInternals">
    <title>DRM Internals</title>
    <para>
      This chapter documents DRM internals relevant to driver authors
      and developers working to add support for the latest features to
      existing drivers.
    </para>
    <para>
      First, we go over some typical driver initialization
      requirements, like setting up command buffers, creating an
      initial output configuration, and initializing core services.
      Subsequent sections cover core internals in more detail,
      providing implementation notes and examples.
    </para>
    <para>
      The DRM layer provides several services to graphics drivers,
      many of them driven by the application interfaces it provides
      through libdrm, the library that wraps most of the DRM ioctls.
      These include vblank event handling, memory
      management, output management, framebuffer management, command
      submission &amp; fencing, suspend/resume support, and DMA
      services.
    </para>

  <!-- Internals: driver init -->

  <sect1>
    <title>Driver Initialization</title>
    <para>
      At the core of every DRM driver is a <structname>drm_driver</structname>
      structure. Drivers typically statically initialize a drm_driver structure,
      and then pass it to one of the <function>drm_*_init()</function> functions
      to register it with the DRM subsystem.
    </para>
    <para>
      The <structname>drm_driver</structname> structure contains static
      information that describes the driver and features it supports, and
      pointers to methods that the DRM core will call to implement the DRM API.
      We will first go through the <structname>drm_driver</structname> static
      information fields, and will then describe individual operations in
      details as they get used in later sections.
    </para>
    <sect2>
      <title>Driver Information</title>
      <sect3>
        <title>Driver Features</title>
        <para>
          Drivers inform the DRM core about their requirements and supported
          features by setting appropriate flags in the
          <structfield>driver_features</structfield> field. Since those flags
          influence the DRM core behaviour since registration time, most of them
          must be set to registering the <structname>drm_driver</structname>
          instance.
        </para>
        <synopsis>u32 driver_features;</synopsis>
        <variablelist>
          <title>Driver Feature Flags</title>
          <varlistentry>
            <term>DRIVER_USE_AGP</term>
            <listitem><para>
              Driver uses AGP interface, the DRM core will manage AGP resources.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_REQUIRE_AGP</term>
            <listitem><para>
              Driver needs AGP interface to function. AGP initialization failure
              will become a fatal error.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_USE_MTRR</term>
            <listitem><para>
              Driver uses MTRR interface for mapping memory, the DRM core will
              manage MTRR resources. Deprecated.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_PCI_DMA</term>
            <listitem><para>
              Driver is capable of PCI DMA, mapping of PCI DMA buffers to
              userspace will be enabled. Deprecated.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_SG</term>
            <listitem><para>
              Driver can perform scatter/gather DMA, allocation and mapping of
              scatter/gather buffers will be enabled. Deprecated.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_HAVE_DMA</term>
            <listitem><para>
              Driver supports DMA, the userspace DMA API will be supported.
              Deprecated.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_HAVE_IRQ</term><term>DRIVER_IRQ_SHARED</term>
            <listitem><para>
              DRIVER_HAVE_IRQ indicates whether the driver has an IRQ handler. The
              DRM core will automatically register an interrupt handler when the
              flag is set. DRIVER_IRQ_SHARED indicates whether the device &amp;
              handler support shared IRQs (note that this is required of PCI
              drivers).
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_IRQ_VBL</term>
            <listitem><para>Unused. Deprecated.</para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_DMA_QUEUE</term>
            <listitem><para>
              Should be set if the driver queues DMA requests and completes them
              asynchronously.  Deprecated.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_FB_DMA</term>
            <listitem><para>
              Driver supports DMA to/from the framebuffer, mapping of frambuffer
              DMA buffers to userspace will be supported. Deprecated.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_IRQ_VBL2</term>
            <listitem><para>Unused. Deprecated.</para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_GEM</term>
            <listitem><para>
              Driver use the GEM memory manager.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_MODESET</term>
            <listitem><para>
              Driver supports mode setting interfaces (KMS).
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term>DRIVER_PRIME</term>
            <listitem><para>
              Driver implements DRM PRIME buffer sharing.
            </para></listitem>
          </varlistentry>
        </variablelist>
      </sect3>
      <sect3>
        <title>Major, Minor and Patchlevel</title>
        <synopsis>int major;
int minor;
int patchlevel;</synopsis>
        <para>
          The DRM core identifies driver versions by a major, minor and patch
          level triplet. The information is printed to the kernel log at
          initialization time and passed to userspace through the
          DRM_IOCTL_VERSION ioctl.
        </para>
        <para>
          The major and minor numbers are also used to verify the requested driver
          API version passed to DRM_IOCTL_SET_VERSION. When the driver API changes
          between minor versions, applications can call DRM_IOCTL_SET_VERSION to
          select a specific version of the API. If the requested major isn't equal
          to the driver major, or the requested minor is larger than the driver
          minor, the DRM_IOCTL_SET_VERSION call will return an error. Otherwise
          the driver's set_version() method will be called with the requested
          version.
        </para>
      </sect3>
      <sect3>
        <title>Name, Description and Date</title>
        <synopsis>char *name;
char *desc;
char *date;</synopsis>
        <para>
          The driver name is printed to the kernel log at initialization time,
          used for IRQ registration and passed to userspace through
          DRM_IOCTL_VERSION.
        </para>
        <para>
          The driver description is a purely informative string passed to
          userspace through the DRM_IOCTL_VERSION ioctl and otherwise unused by
          the kernel.
        </para>
        <para>
          The driver date, formatted as YYYYMMDD, is meant to identify the date of
          the latest modification to the driver. However, as most drivers fail to
          update it, its value is mostly useless. The DRM core prints it to the
          kernel log at initialization time and passes it to userspace through the
          DRM_IOCTL_VERSION ioctl.
        </para>
      </sect3>
    </sect2>
    <sect2>
      <title>Driver Load</title>
      <para>
        The <methodname>load</methodname> method is the driver and device
        initialization entry point. The method is responsible for allocating and
        initializing driver private data, specifying supported performance
        counters, performing resource allocation and mapping (e.g. acquiring
        clocks, mapping registers or allocating command buffers), initializing
        the memory manager (<xref linkend="drm-memory-management"/>), installing
        the IRQ handler (<xref linkend="drm-irq-registration"/>), setting up
        vertical blanking handling (<xref linkend="drm-vertical-blank"/>), mode
	setting (<xref linkend="drm-mode-setting"/>) and initial output
	configuration (<xref linkend="drm-kms-init"/>).
      </para>
      <note><para>
        If compatibility is a concern (e.g. with drivers converted over from
        User Mode Setting to Kernel Mode Setting), care must be taken to prevent
        device initialization and control that is incompatible with currently
        active userspace drivers. For instance, if user level mode setting
        drivers are in use, it would be problematic to perform output discovery
        &amp; configuration at load time. Likewise, if user-level drivers
        unaware of memory management are in use, memory management and command
        buffer setup may need to be omitted. These requirements are
        driver-specific, and care needs to be taken to keep both old and new
        applications and libraries working.
      </para></note>
      <synopsis>int (*load) (struct drm_device *, unsigned long flags);</synopsis>
      <para>
        The method takes two arguments, a pointer to the newly created
	<structname>drm_device</structname> and flags. The flags are used to
	pass the <structfield>driver_data</structfield> field of the device id
	corresponding to the device passed to <function>drm_*_init()</function>.
	Only PCI devices currently use this, USB and platform DRM drivers have
	their <methodname>load</methodname> method called with flags to 0.
      </para>
      <sect3>
        <title>Driver Private &amp; Performance Counters</title>
        <para>
          The driver private hangs off the main
          <structname>drm_device</structname> structure and can be used for
          tracking various device-specific bits of information, like register
          offsets, command buffer status, register state for suspend/resume, etc.
          At load time, a driver may simply allocate one and set
          <structname>drm_device</structname>.<structfield>dev_priv</structfield>
          appropriately; it should be freed and
          <structname>drm_device</structname>.<structfield>dev_priv</structfield>
          set to NULL when the driver is unloaded.
        </para>
        <para>
          DRM supports several counters which were used for rough performance
          characterization. This stat counter system is deprecated and should not
          be used. If performance monitoring is desired, the developer should
          investigate and potentially enhance the kernel perf and tracing
          infrastructure to export GPU related performance information for
          consumption by performance monitoring tools and applications.
        </para>
      </sect3>
      <sect3 id="drm-irq-registration">
        <title>IRQ Registration</title>
        <para>
          The DRM core tries to facilitate IRQ handler registration and
          unregistration by providing <function>drm_irq_install</function> and
          <function>drm_irq_uninstall</function> functions. Those functions only
          support a single interrupt per device.
        </para>
  <!--!Fdrivers/char/drm/drm_irq.c drm_irq_install-->
        <para>
          Both functions get the device IRQ by calling
          <function>drm_dev_to_irq</function>. This inline function will call a
          bus-specific operation to retrieve the IRQ number. For platform devices,
          <function>platform_get_irq</function>(..., 0) is used to retrieve the
          IRQ number.
        </para>
        <para>
          <function>drm_irq_install</function> starts by calling the
          <methodname>irq_preinstall</methodname> driver operation. The operation
          is optional and must make sure that the interrupt will not get fired by
          clearing all pending interrupt flags or disabling the interrupt.
        </para>
        <para>
          The IRQ will then be requested by a call to
          <function>request_irq</function>. If the DRIVER_IRQ_SHARED driver
          feature flag is set, a shared (IRQF_SHARED) IRQ handler will be
          requested.
        </para>
        <para>
          The IRQ handler function must be provided as the mandatory irq_handler
          driver operation. It will get passed directly to
          <function>request_irq</function> and thus has the same prototype as all
          IRQ handlers. It will get called with a pointer to the DRM device as the
          second argument.
        </para>
        <para>
          Finally the function calls the optional
          <methodname>irq_postinstall</methodname> driver operation. The operation
          usually enables interrupts (excluding the vblank interrupt, which is
          enabled separately), but drivers may choose to enable/disable interrupts
          at a different time.
        </para>
        <para>
          <function>drm_irq_uninstall</function> is similarly used to uninstall an
          IRQ handler. It starts by waking up all processes waiting on a vblank
          interrupt to make sure they don't hang, and then calls the optional
          <methodname>irq_uninstall</methodname> driver operation. The operation
          must disable all hardware interrupts. Finally the function frees the IRQ
          by calling <function>free_irq</function>.
        </para>
      </sect3>
      <sect3>
        <title>Memory Manager Initialization</title>
        <para>
          Every DRM driver requires a memory manager which must be initialized at
          load time. DRM currently contains two memory managers, the Translation
          Table Manager (TTM) and the Graphics Execution Manager (GEM).
          This document describes the use of the GEM memory manager only. See
          <xref linkend="drm-memory-management"/> for details.
        </para>
      </sect3>
      <sect3>
        <title>Miscellaneous Device Configuration</title>
        <para>
          Another task that may be necessary for PCI devices during configuration
          is mapping the video BIOS. On many devices, the VBIOS describes device
          configuration, LCD panel timings (if any), and contains flags indicating
          device state. Mapping the BIOS can be done using the pci_map_rom() call,
          a convenience function that takes care of mapping the actual ROM,
          whether it has been shadowed into memory (typically at address 0xc0000)
          or exists on the PCI device in the ROM BAR. Note that after the ROM has
          been mapped and any necessary information has been extracted, it should
          be unmapped; on many devices, the ROM address decoder is shared with
          other BARs, so leaving it mapped could cause undesired behaviour like
          hangs or memory corruption.
  <!--!Fdrivers/pci/rom.c pci_map_rom-->
        </para>
      </sect3>
    </sect2>
  </sect1>

  <!-- Internals: memory management -->

  <sect1 id="drm-memory-management">
    <title>Memory management</title>
    <para>
      Modern Linux systems require large amount of graphics memory to store
      frame buffers, textures, vertices and other graphics-related data. Given
      the very dynamic nature of many of that data, managing graphics memory
      efficiently is thus crucial for the graphics stack and plays a central
      role in the DRM infrastructure.
    </para>
    <para>
      The DRM core includes two memory managers, namely Translation Table Maps
      (TTM) and Graphics Execution Manager (GEM). TTM was the first DRM memory
      manager to be developed and tried to be a one-size-fits-them all
      solution. It provides a single userspace API to accomodate the need of
      all hardware, supporting both Unified Memory Architecture (UMA) devices
      and devices with dedicated video RAM (i.e. most discrete video cards).
      This resulted in a large, complex piece of code that turned out to be
      hard to use for driver development.
    </para>
    <para>
      GEM started as an Intel-sponsored project in reaction to TTM's
      complexity. Its design philosophy is completely different: instead of
      providing a solution to every graphics memory-related problems, GEM
      identified common code between drivers and created a support library to
      share it. GEM has simpler initialization and execution requirements than
      TTM, but has no video RAM management capabitilies and is thus limited to
      UMA devices.
    </para>
    <sect2>
      <title>The Translation Table Manager (TTM)</title>
      <para>
	TTM design background and information belongs here.
      </para>
      <sect3>
	<title>TTM initialization</title>
        <warning><para>This section is outdated.</para></warning>
        <para>
          Drivers wishing to support TTM must fill out a drm_bo_driver
          structure. The structure contains several fields with function
          pointers for initializing the TTM, allocating and freeing memory,
          waiting for command completion and fence synchronization, and memory
          migration. See the radeon_ttm.c file for an example of usage.
	</para>
	<para>
	  The ttm_global_reference structure is made up of several fields:
	</para>
	<programlisting>
	  struct ttm_global_reference {
	  	enum ttm_global_types global_type;
	  	size_t size;
	  	void *object;
	  	int (*init) (struct ttm_global_reference *);
	  	void (*release) (struct ttm_global_reference *);
	  };
	</programlisting>
	<para>
	  There should be one global reference structure for your memory
	  manager as a whole, and there will be others for each object
	  created by the memory manager at runtime.  Your global TTM should
	  have a type of TTM_GLOBAL_TTM_MEM.  The size field for the global
	  object should be sizeof(struct ttm_mem_global), and the init and
	  release hooks should point at your driver-specific init and
	  release routines, which probably eventually call
	  ttm_mem_global_init and ttm_mem_global_release, respectively.
	</para>
	<para>
	  Once your global TTM accounting structure is set up and initialized
	  by calling ttm_global_item_ref() on it,
	  you need to create a buffer object TTM to
	  provide a pool for buffer object allocation by clients and the
	  kernel itself.  The type of this object should be TTM_GLOBAL_TTM_BO,
	  and its size should be sizeof(struct ttm_bo_global).  Again,
	  driver-specific init and release functions may be provided,
	  likely eventually calling ttm_bo_global_init() and
	  ttm_bo_global_release(), respectively.  Also, like the previous
	  object, ttm_global_item_ref() is used to create an initial reference
	  count for the TTM, which will call your initialization function.
	</para>
      </sect3>
    </sect2>
    <sect2 id="drm-gem">
      <title>The Graphics Execution Manager (GEM)</title>
      <para>
        The GEM design approach has resulted in a memory manager that doesn't
        provide full coverage of all (or even all common) use cases in its
        userspace or kernel API. GEM exposes a set of standard memory-related
        operations to userspace and a set of helper functions to drivers, and let
        drivers implement hardware-specific operations with their own private API.
      </para>
      <para>
        The GEM userspace API is described in the
        <ulink url="http://lwn.net/Articles/283798/"><citetitle>GEM - the Graphics
        Execution Manager</citetitle></ulink> article on LWN. While slightly
        outdated, the document provides a good overview of the GEM API principles.
        Buffer allocation and read and write operations, described as part of the
        common GEM API, are currently implemented using driver-specific ioctls.
      </para>
      <para>
        GEM is data-agnostic. It manages abstract buffer objects without knowing
        what individual buffers contain. APIs that require knowledge of buffer
        contents or purpose, such as buffer allocation or synchronization
        primitives, are thus outside of the scope of GEM and must be implemented
        using driver-specific ioctls.
      </para>
      <para>
	On a fundamental level, GEM involves several operations:
	<itemizedlist>
	  <listitem>Memory allocation and freeing</listitem>
	  <listitem>Command execution</listitem>
	  <listitem>Aperture management at command execution time</listitem>
	</itemizedlist>
	Buffer object allocation is relatively straightforward and largely
        provided by Linux's shmem layer, which provides memory to back each
        object.
      </para>
      <para>
        Device-specific operations, such as command execution, pinning, buffer
	read &amp; write, mapping, and domain ownership transfers are left to
        driver-specific ioctls.
      </para>
      <sect3>
        <title>GEM Initialization</title>
        <para>
          Drivers that use GEM must set the DRIVER_GEM bit in the struct
          <structname>drm_driver</structname>
          <structfield>driver_features</structfield> field. The DRM core will
          then automatically initialize the GEM core before calling the
          <methodname>load</methodname> operation. Behind the scene, this will
          create a DRM Memory Manager object which provides an address space
          pool for object allocation.
        </para>
        <para>
          In a KMS configuration, drivers need to allocate and initialize a
          command ring buffer following core GEM initialization if required by
          the hardware. UMA devices usually have what is called a "stolen"
          memory region, which provides space for the initial framebuffer and
          large, contiguous memory regions required by the device. This space is
          typically not managed by GEM, and must be initialized separately into
          its own DRM MM object.
        </para>
      </sect3>
      <sect3>
        <title>GEM Objects Creation</title>
        <para>
          GEM splits creation of GEM objects and allocation of the memory that
          backs them in two distinct operations.
        </para>
        <para>
          GEM objects are represented by an instance of struct
          <structname>drm_gem_object</structname>. Drivers usually need to extend
          GEM objects with private information and thus create a driver-specific
          GEM object structure type that embeds an instance of struct
          <structname>drm_gem_object</structname>.
        </para>
        <para>
          To create a GEM object, a driver allocates memory for an instance of its
          specific GEM object type and initializes the embedded struct
          <structname>drm_gem_object</structname> with a call to
          <function>drm_gem_object_init</function>. The function takes a pointer to
          the DRM device, a pointer to the GEM object and the buffer object size
          in bytes.
        </para>
        <para>
          GEM uses shmem to allocate anonymous pageable memory.
          <function>drm_gem_object_init</function> will create an shmfs file of
          the requested size and store it into the struct
          <structname>drm_gem_object</structname> <structfield>filp</structfield>
          field. The memory is used as either main storage for the object when the
          graphics hardware uses system memory directly or as a backing store
          otherwise.
        </para>
        <para>
          Drivers are responsible for the actual physical pages allocation by
          calling <function>shmem_read_mapping_page_gfp</function> for each page.
          Note that they can decide to allocate pages when initializing the GEM
          object, or to delay allocation until the memory is needed (for instance
          when a page fault occurs as a result of a userspace memory access or
          when the driver needs to start a DMA transfer involving the memory).
        </para>
        <para>
          Anonymous pageable memory allocation is not always desired, for instance
          when the hardware requires physically contiguous system memory as is
          often the case in embedded devices. Drivers can create GEM objects with
          no shmfs backing (called private GEM objects) by initializing them with
          a call to <function>drm_gem_private_object_init</function> instead of
          <function>drm_gem_object_init</function>. Storage for private GEM
          objects must be managed by drivers.
        </para>
        <para>
          Drivers that do not need to extend GEM objects with private information
          can call the <function>drm_gem_object_alloc</function> function to
          allocate and initialize a struct <structname>drm_gem_object</structname>
          instance. The GEM core will call the optional driver
          <methodname>gem_init_object</methodname> operation after initializing
          the GEM object with <function>drm_gem_object_init</function>.
          <synopsis>int (*gem_init_object) (struct drm_gem_object *obj);</synopsis>
        </para>
        <para>
          No alloc-and-init function exists for private GEM objects.
        </para>
      </sect3>
      <sect3>
        <title>GEM Objects Lifetime</title>
        <para>
          All GEM objects are reference-counted by the GEM core. References can be
          acquired and release by <function>calling drm_gem_object_reference</function>
          and <function>drm_gem_object_unreference</function> respectively. The
          caller must hold the <structname>drm_device</structname>
          <structfield>struct_mutex</structfield> lock. As a convenience, GEM
          provides the <function>drm_gem_object_reference_unlocked</function> and
          <function>drm_gem_object_unreference_unlocked</function> functions that
          can be called without holding the lock.
        </para>
        <para>
          When the last reference to a GEM object is released the GEM core calls
          the <structname>drm_driver</structname>
          <methodname>gem_free_object</methodname> operation. That operation is
          mandatory for GEM-enabled drivers and must free the GEM object and all
          associated resources.
        </para>
        <para>
          <synopsis>void (*gem_free_object) (struct drm_gem_object *obj);</synopsis>
          Drivers are responsible for freeing all GEM object resources, including
          the resources created by the GEM core. If an mmap offset has been
          created for the object (in which case
          <structname>drm_gem_object</structname>::<structfield>map_list</structfield>::<structfield>map</structfield>
          is not NULL) it must be freed by a call to
          <function>drm_gem_free_mmap_offset</function>. The shmfs backing store
          must be released by calling <function>drm_gem_object_release</function>
          (that function can safely be called if no shmfs backing store has been
          created).
        </para>
      </sect3>
      <sect3>
        <title>GEM Objects Naming</title>
        <para>
          Communication between userspace and the kernel refers to GEM objects
          using local handles, global names or, more recently, file descriptors.
          All of those are 32-bit integer values; the usual Linux kernel limits
          apply to the file descriptors.
        </para>
        <para>
          GEM handles are local to a DRM file. Applications get a handle to a GEM
          object through a driver-specific ioctl, and can use that handle to refer
          to the GEM object in other standard or driver-specific ioctls. Closing a
          DRM file handle frees all its GEM handles and dereferences the
          associated GEM objects.
        </para>
        <para>
          To create a handle for a GEM object drivers call
          <function>drm_gem_handle_create</function>. The function takes a pointer
          to the DRM file and the GEM object and returns a locally unique handle.
          When the handle is no longer needed drivers delete it with a call to
          <function>drm_gem_handle_delete</function>. Finally the GEM object
          associated with a handle can be retrieved by a call to
          <function>drm_gem_object_lookup</function>.
        </para>
        <para>
          Handles don't take ownership of GEM objects, they only take a reference
          to the object that will be dropped when the handle is destroyed. To
          avoid leaking GEM objects, drivers must make sure they drop the
          reference(s) they own (such as the initial reference taken at object
          creation time) as appropriate, without any special consideration for the
          handle. For example, in the particular case of combined GEM object and
          handle creation in the implementation of the
          <methodname>dumb_create</methodname> operation, drivers must drop the
          initial reference to the GEM object before returning the handle.
        </para>
        <para>
          GEM names are similar in purpose to handles but are not local to DRM
          files. They can be passed between processes to reference a GEM object
          globally. Names can't be used directly to refer to objects in the DRM
          API, applications must convert handles to names and names to handles
          using the DRM_IOCTL_GEM_FLINK and DRM_IOCTL_GEM_OPEN ioctls
          respectively. The conversion is handled by the DRM core without any
          driver-specific support.
        </para>
        <para>
          Similar to global names, GEM file descriptors are also used to share GEM
          objects across processes. They offer additional security: as file
          descriptors must be explictly sent over UNIX domain sockets to be shared
          between applications, they can't be guessed like the globally unique GEM
          names.
        </para>
        <para>
          Drivers that support GEM file descriptors, also known as the DRM PRIME
          API, must set the DRIVER_PRIME bit in the struct
          <structname>drm_driver</structname>
          <structfield>driver_features</structfield> field, and implement the
          <methodname>prime_handle_to_fd</methodname> and
          <methodname>prime_fd_to_handle</methodname> operations.
        </para>
        <para>
          <synopsis>int (*prime_handle_to_fd)(struct drm_device *dev,
                            struct drm_file *file_priv, uint32_t handle,
                            uint32_t flags, int *prime_fd);
  int (*prime_fd_to_handle)(struct drm_device *dev,
                            struct drm_file *file_priv, int prime_fd,
                            uint32_t *handle);</synopsis>
          Those two operations convert a handle to a PRIME file descriptor and
          vice versa. Drivers must use the kernel dma-buf buffer sharing framework
          to manage the PRIME file descriptors.
        </para>
        <para>
          While non-GEM drivers must implement the operations themselves, GEM
          drivers must use the <function>drm_gem_prime_handle_to_fd</function>
          and <function>drm_gem_prime_fd_to_handle</function> helper functions.
          Those helpers rely on the driver
          <methodname>gem_prime_export</methodname> and
          <methodname>gem_prime_import</methodname> operations to create a dma-buf
          instance from a GEM object (dma-buf exporter role) and to create a GEM
          object from a dma-buf instance (dma-buf importer role).
        </para>
        <para>
          <synopsis>struct dma_buf * (*gem_prime_export)(struct drm_device *dev,
                                       struct drm_gem_object *obj,
                                       int flags);
  struct drm_gem_object * (*gem_prime_import)(struct drm_device *dev,
                                              struct dma_buf *dma_buf);</synopsis>
          These two operations are mandatory for GEM drivers that support DRM
          PRIME.
        </para>
      </sect3>
      <sect3 id="drm-gem-objects-mapping">
        <title>GEM Objects Mapping</title>
        <para>
          Because mapping operations are fairly heavyweight GEM favours
          read/write-like access to buffers, implemented through driver-specific
          ioctls, over mapping buffers to userspace. However, when random access
          to the buffer is needed (to perform software rendering for instance),
          direct access to the object can be more efficient.
        </para>
        <para>
          The mmap system call can't be used directly to map GEM objects, as they
          don't have their own file handle. Two alternative methods currently
          co-exist to map GEM objects to userspace. The first method uses a
          driver-specific ioctl to perform the mapping operation, calling
          <function>do_mmap</function> under the hood. This is often considered
          dubious, seems to be discouraged for new GEM-enabled drivers, and will
          thus not be described here.
        </para>
        <para>
          The second method uses the mmap system call on the DRM file handle.
          <synopsis>void *mmap(void *addr, size_t length, int prot, int flags, int fd,
             off_t offset);</synopsis>
          DRM identifies the GEM object to be mapped by a fake offset passed
          through the mmap offset argument. Prior to being mapped, a GEM object
          must thus be associated with a fake offset. To do so, drivers must call
          <function>drm_gem_create_mmap_offset</function> on the object. The
          function allocates a fake offset range from a pool and stores the
          offset divided by PAGE_SIZE in
          <literal>obj-&gt;map_list.hash.key</literal>. Care must be taken not to
          call <function>drm_gem_create_mmap_offset</function> if a fake offset
          has already been allocated for the object. This can be tested by
          <literal>obj-&gt;map_list.map</literal> being non-NULL.
        </para>
        <para>
          Once allocated, the fake offset value
          (<literal>obj-&gt;map_list.hash.key &lt;&lt; PAGE_SHIFT</literal>)
          must be passed to the application in a driver-specific way and can then
          be used as the mmap offset argument.
        </para>
        <para>
          The GEM core provides a helper method <function>drm_gem_mmap</function>
          to handle object mapping. The method can be set directly as the mmap
          file operation handler. It will look up the GEM object based on the
          offset value and set the VMA operations to the
          <structname>drm_driver</structname> <structfield>gem_vm_ops</structfield>
          field. Note that <function>drm_gem_mmap</function> doesn't map memory to
          userspace, but relies on the driver-provided fault handler to map pages
          individually.
        </para>
        <para>
          To use <function>drm_gem_mmap</function>, drivers must fill the struct
          <structname>drm_driver</structname> <structfield>gem_vm_ops</structfield>
          field with a pointer to VM operations.
        </para>
        <para>
          <synopsis>struct vm_operations_struct *gem_vm_ops

  struct vm_operations_struct {
          void (*open)(struct vm_area_struct * area);
          void (*close)(struct vm_area_struct * area);
          int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf);
  };</synopsis>
        </para>
        <para>
          The <methodname>open</methodname> and <methodname>close</methodname>
          operations must update the GEM object reference count. Drivers can use
          the <function>drm_gem_vm_open</function> and
          <function>drm_gem_vm_close</function> helper functions directly as open
          and close handlers.
        </para>
        <para>
          The fault operation handler is responsible for mapping individual pages
          to userspace when a page fault occurs. Depending on the memory
          allocation scheme, drivers can allocate pages at fault time, or can
          decide to allocate memory for the GEM object at the time the object is
          created.
        </para>
        <para>
          Drivers that want to map the GEM object upfront instead of handling page
          faults can implement their own mmap file operation handler.
        </para>
      </sect3>
      <sect3>
        <title>Dumb GEM Objects</title>
        <para>
          The GEM API doesn't standardize GEM objects creation and leaves it to
          driver-specific ioctls. While not an issue for full-fledged graphics
          stacks that include device-specific userspace components (in libdrm for
          instance), this limit makes DRM-based early boot graphics unnecessarily
          complex.
        </para>
        <para>
          Dumb GEM objects partly alleviate the problem by providing a standard
          API to create dumb buffers suitable for scanout, which can then be used
          to create KMS frame buffers.
        </para>
        <para>
          To support dumb GEM objects drivers must implement the
          <methodname>dumb_create</methodname>,
          <methodname>dumb_destroy</methodname> and
          <methodname>dumb_map_offset</methodname> operations.
        </para>
        <itemizedlist>
          <listitem>
            <synopsis>int (*dumb_create)(struct drm_file *file_priv, struct drm_device *dev,
                     struct drm_mode_create_dumb *args);</synopsis>
            <para>
              The <methodname>dumb_create</methodname> operation creates a GEM
              object suitable for scanout based on the width, height and depth
              from the struct <structname>drm_mode_create_dumb</structname>
              argument. It fills the argument's <structfield>handle</structfield>,
              <structfield>pitch</structfield> and <structfield>size</structfield>
              fields with a handle for the newly created GEM object and its line
              pitch and size in bytes.
            </para>
          </listitem>
          <listitem>
            <synopsis>int (*dumb_destroy)(struct drm_file *file_priv, struct drm_device *dev,
                      uint32_t handle);</synopsis>
            <para>
              The <methodname>dumb_destroy</methodname> operation destroys a dumb
              GEM object created by <methodname>dumb_create</methodname>.
            </para>
          </listitem>
          <listitem>
            <synopsis>int (*dumb_map_offset)(struct drm_file *file_priv, struct drm_device *dev,
                         uint32_t handle, uint64_t *offset);</synopsis>
            <para>
              The <methodname>dumb_map_offset</methodname> operation associates an
              mmap fake offset with the GEM object given by the handle and returns
              it. Drivers must use the
              <function>drm_gem_create_mmap_offset</function> function to
              associate the fake offset as described in
              <xref linkend="drm-gem-objects-mapping"/>.
            </para>
          </listitem>
        </itemizedlist>
      </sect3>
      <sect3>
        <title>Memory Coherency</title>
        <para>
          When mapped to the device or used in a command buffer, backing pages
          for an object are flushed to memory and marked write combined so as to
          be coherent with the GPU. Likewise, if the CPU accesses an object
          after the GPU has finished rendering to the object, then the object
          must be made coherent with the CPU's view of memory, usually involving
          GPU cache flushing of various kinds. This core CPU&lt;-&gt;GPU
          coherency management is provided by a device-specific ioctl, which
          evaluates an object's current domain and performs any necessary
          flushing or synchronization to put the object into the desired
          coherency domain (note that the object may be busy, i.e. an active
          render target; in that case, setting the domain blocks the client and
          waits for rendering to complete before performing any necessary
          flushing operations).
        </para>
      </sect3>
      <sect3>
        <title>Command Execution</title>
        <para>
	  Perhaps the most important GEM function for GPU devices is providing a
          command execution interface to clients. Client programs construct
          command buffers containing references to previously allocated memory
          objects, and then submit them to GEM. At that point, GEM takes care to
          bind all the objects into the GTT, execute the buffer, and provide
          necessary synchronization between clients accessing the same buffers.
          This often involves evicting some objects from the GTT and re-binding
          others (a fairly expensive operation), and providing relocation
          support which hides fixed GTT offsets from clients. Clients must take
          care not to submit command buffers that reference more objects than
          can fit in the GTT; otherwise, GEM will reject them and no rendering
          will occur. Similarly, if several objects in the buffer require fence
          registers to be allocated for correct rendering (e.g. 2D blits on
          pre-965 chips), care must be taken not to require more fence registers
          than are available to the client. Such resource management should be
          abstracted from the client in libdrm.
        </para>
      </sect3>
    </sect2>
  </sect1>

  <!-- Internals: mode setting -->

  <sect1 id="drm-mode-setting">
    <title>Mode Setting</title>
    <para>
      Drivers must initialize the mode setting core by calling
      <function>drm_mode_config_init</function> on the DRM device. The function
      initializes the <structname>drm_device</structname>
      <structfield>mode_config</structfield> field and never fails. Once done,
      mode configuration must be setup by initializing the following fields.
    </para>
    <itemizedlist>
      <listitem>
        <synopsis>int min_width, min_height;
int max_width, max_height;</synopsis>
        <para>
	  Minimum and maximum width and height of the frame buffers in pixel
	  units.
	</para>
      </listitem>
      <listitem>
        <synopsis>struct drm_mode_config_funcs *funcs;</synopsis>
	<para>Mode setting functions.</para>
      </listitem>
    </itemizedlist>
    <sect2>
      <title>Frame Buffer Creation</title>
      <synopsis>struct drm_framebuffer *(*fb_create)(struct drm_device *dev,
				     struct drm_file *file_priv,
				     struct drm_mode_fb_cmd2 *mode_cmd);</synopsis>
      <para>
        Frame buffers are abstract memory objects that provide a source of
        pixels to scanout to a CRTC. Applications explicitly request the
        creation of frame buffers through the DRM_IOCTL_MODE_ADDFB(2) ioctls and
        receive an opaque handle that can be passed to the KMS CRTC control,
        plane configuration and page flip functions.
      </para>
      <para>
        Frame buffers rely on the underneath memory manager for low-level memory
        operations. When creating a frame buffer applications pass a memory
        handle (or a list of memory handles for multi-planar formats) through
        the <parameter>drm_mode_fb_cmd2</parameter> argument. This document
        assumes that the driver uses GEM, those handles thus reference GEM
        objects.
      </para>
      <para>
        Drivers must first validate the requested frame buffer parameters passed
        through the mode_cmd argument. In particular this is where invalid
        sizes, pixel formats or pitches can be caught.
      </para>
      <para>
        If the parameters are deemed valid, drivers then create, initialize and
        return an instance of struct <structname>drm_framebuffer</structname>.
        If desired the instance can be embedded in a larger driver-specific
        structure. The new instance is initialized with a call to
        <function>drm_framebuffer_init</function> which takes a pointer to DRM
        frame buffer operations (struct
        <structname>drm_framebuffer_funcs</structname>). Frame buffer operations are
        <itemizedlist>
          <listitem>
            <synopsis>int (*create_handle)(struct drm_framebuffer *fb,
		     struct drm_file *file_priv, unsigned int *handle);</synopsis>
            <para>
              Create a handle to the frame buffer underlying memory object. If
              the frame buffer uses a multi-plane format, the handle will
              reference the memory object associated with the first plane.
            </para>
            <para>
              Drivers call <function>drm_gem_handle_create</function> to create
              the handle.
            </para>
          </listitem>
          <listitem>
            <synopsis>void (*destroy)(struct drm_framebuffer *framebuffer);</synopsis>
            <para>
              Destroy the frame buffer object and frees all associated
              resources. Drivers must call
              <function>drm_framebuffer_cleanup</function> to free resources
              allocated by the DRM core for the frame buffer object, and must
              make sure to unreference all memory objects associated with the
              frame buffer. Handles created by the
              <methodname>create_handle</methodname> operation are released by
              the DRM core.
            </para>
          </listitem>
          <listitem>
            <synopsis>int (*dirty)(struct drm_framebuffer *framebuffer,
	     struct drm_file *file_priv, unsigned flags, unsigned color,
	     struct drm_clip_rect *clips, unsigned num_clips);</synopsis>
            <para>
              This optional operation notifies the driver that a region of the
              frame buffer has changed in response to a DRM_IOCTL_MODE_DIRTYFB
              ioctl call.
            </para>
          </listitem>
        </itemizedlist>
      </para>
      <para>
        After initializing the <structname>drm_framebuffer</structname>
        instance drivers must fill its <structfield>width</structfield>,
        <structfield>height</structfield>, <structfield>pitches</structfield>,
        <structfield>offsets</structfield>, <structfield>depth</structfield>,
        <structfield>bits_per_pixel</structfield> and
        <structfield>pixel_format</structfield> fields from the values passed
        through the <parameter>drm_mode_fb_cmd2</parameter> argument. They
        should call the <function>drm_helper_mode_fill_fb_struct</function>
        helper function to do so.
      </para>
    </sect2>
    <sect2>
      <title>Output Polling</title>
      <synopsis>void (*output_poll_changed)(struct drm_device *dev);</synopsis>
      <para>
        This operation notifies the driver that the status of one or more
        connectors has changed. Drivers that use the fb helper can just call the
        <function>drm_fb_helper_hotplug_event</function> function to handle this
        operation.
      </para>
    </sect2>
  </sect1>

  <!-- Internals: kms initialization and cleanup -->

  <sect1 id="drm-kms-init">
    <title>KMS Initialization and Cleanup</title>
    <para>
      A KMS device is abstracted and exposed as a set of planes, CRTCs, encoders
      and connectors. KMS drivers must thus create and initialize all those
      objects at load time after initializing mode setting.
    </para>
    <sect2>
      <title>CRTCs (struct <structname>drm_crtc</structname>)</title>
      <para>
        A CRTC is an abstraction representing a part of the chip that contains a
	pointer to a scanout buffer. Therefore, the number of CRTCs available
	determines how many independent scanout buffers can be active at any
	given time. The CRTC structure contains several fields to support this:
	a pointer to some video memory (abstracted as a frame buffer object), a
	display mode, and an (x, y) offset into the video memory to support
	panning or configurations where one piece of video memory spans multiple
	CRTCs.
      </para>
      <sect3>
        <title>CRTC Initialization</title>
        <para>
          A KMS device must create and register at least one struct
          <structname>drm_crtc</structname> instance. The instance is allocated
          and zeroed by the driver, possibly as part of a larger structure, and
          registered with a call to <function>drm_crtc_init</function> with a
          pointer to CRTC functions.
        </para>
      </sect3>
      <sect3>
        <title>CRTC Operations</title>
        <sect4>
          <title>Set Configuration</title>
          <synopsis>int (*set_config)(struct drm_mode_set *set);</synopsis>
          <para>
            Apply a new CRTC configuration to the device. The configuration
            specifies a CRTC, a frame buffer to scan out from, a (x,y) position in
            the frame buffer, a display mode and an array of connectors to drive
            with the CRTC if possible.
          </para>
          <para>
            If the frame buffer specified in the configuration is NULL, the driver
            must detach all encoders connected to the CRTC and all connectors
            attached to those encoders and disable them.
          </para>
          <para>
            This operation is called with the mode config lock held.
          </para>
          <note><para>
            FIXME: How should set_config interact with DPMS? If the CRTC is
            suspended, should it be resumed?
          </para></note>
        </sect4>
        <sect4>
          <title>Page Flipping</title>
          <synopsis>int (*page_flip)(struct drm_crtc *crtc, struct drm_framebuffer *fb,
                   struct drm_pending_vblank_event *event);</synopsis>
          <para>
            Schedule a page flip to the given frame buffer for the CRTC. This
            operation is called with the mode config mutex held.
          </para>
          <para>
            Page flipping is a synchronization mechanism that replaces the frame
            buffer being scanned out by the CRTC with a new frame buffer during
            vertical blanking, avoiding tearing. When an application requests a page
            flip the DRM core verifies that the new frame buffer is large enough to
            be scanned out by  the CRTC in the currently configured mode and then
            calls the CRTC <methodname>page_flip</methodname> operation with a
            pointer to the new frame buffer.
          </para>
          <para>
            The <methodname>page_flip</methodname> operation schedules a page flip.
            Once any pending rendering targetting the new frame buffer has
            completed, the CRTC will be reprogrammed to display that frame buffer
            after the next vertical refresh. The operation must return immediately
            without waiting for rendering or page flip to complete and must block
            any new rendering to the frame buffer until the page flip completes.
          </para>
          <para>
            If a page flip is already pending, the
            <methodname>page_flip</methodname> operation must return
            -<errorname>EBUSY</errorname>.
          </para>
          <para>
            To synchronize page flip to vertical blanking the driver will likely
            need to enable vertical blanking interrupts. It should call
            <function>drm_vblank_get</function> for that purpose, and call
            <function>drm_vblank_put</function> after the page flip completes.
          </para>
          <para>
            If the application has requested to be notified when page flip completes
            the <methodname>page_flip</methodname> operation will be called with a
            non-NULL <parameter>event</parameter> argument pointing to a
            <structname>drm_pending_vblank_event</structname> instance. Upon page
            flip completion the driver must fill the
            <parameter>event</parameter>::<structfield>event</structfield>
            <structfield>sequence</structfield>, <structfield>tv_sec</structfield>
            and <structfield>tv_usec</structfield> fields with the associated
            vertical blanking count and timestamp, add the event to the
            <parameter>drm_file</parameter> list of events to be signaled, and wake
            up any waiting process. This can be performed with
            <programlisting><![CDATA[
            struct timeval now;

            event->event.sequence = drm_vblank_count_and_time(..., &now);
            event->event.tv_sec = now.tv_sec;
            event->event.tv_usec = now.tv_usec;

            spin_lock_irqsave(&dev->event_lock, flags);
            list_add_tail(&event->base.link, &event->base.file_priv->event_list);
            wake_up_interruptible(&event->base.file_priv->event_wait);
            spin_unlock_irqrestore(&dev->event_lock, flags);
            ]]></programlisting>
          </para>
          <note><para>
            FIXME: Could drivers that don't need to wait for rendering to complete
            just add the event to <literal>dev-&gt;vblank_event_list</literal> and
            let the DRM core handle everything, as for "normal" vertical blanking
            events?
          </para></note>
          <para>
            While waiting for the page flip to complete, the
            <literal>event-&gt;base.link</literal> list head can be used freely by
            the driver to store the pending event in a driver-specific list.
          </para>
          <para>
            If the file handle is closed before the event is signaled, drivers must
            take care to destroy the event in their
            <methodname>preclose</methodname> operation (and, if needed, call
            <function>drm_vblank_put</function>).
          </para>
        </sect4>
        <sect4>
          <title>Miscellaneous</title>
          <itemizedlist>
            <listitem>
              <synopsis>void (*gamma_set)(struct drm_crtc *crtc, u16 *r, u16 *g, u16 *b,
                        uint32_t start, uint32_t size);</synopsis>
              <para>
                Apply a gamma table to the device. The operation is optional.
              </para>
            </listitem>
            <listitem>
              <synopsis>void (*destroy)(struct drm_crtc *crtc);</synopsis>
              <para>
                Destroy the CRTC when not needed anymore. See
                <xref linkend="drm-kms-init"/>.
              </para>
            </listitem>
          </itemizedlist>
        </sect4>
      </sect3>
    </sect2>
    <sect2>
      <title>Planes (struct <structname>drm_plane</structname>)</title>
      <para>
        A plane represents an image source that can be blended with or overlayed
	on top of a CRTC during the scanout process. Planes are associated with
	a frame buffer to crop a portion of the image memory (source) and
	optionally scale it to a destination size. The result is then blended
	with or overlayed on top of a CRTC.
      </para>
      <sect3>
        <title>Plane Initialization</title>
        <para>
          Planes are optional. To create a plane, a KMS drivers allocates and
          zeroes an instances of struct <structname>drm_plane</structname>
          (possibly as part of a larger structure) and registers it with a call
          to <function>drm_plane_init</function>. The function takes a bitmask
          of the CRTCs that can be associated with the plane, a pointer to the
          plane functions and a list of format supported formats.
        </para>
      </sect3>
      <sect3>
        <title>Plane Operations</title>
        <itemizedlist>
          <listitem>
            <synopsis>int (*update_plane)(struct drm_plane *plane, struct drm_crtc *crtc,
                        struct drm_framebuffer *fb, int crtc_x, int crtc_y,
                        unsigned int crtc_w, unsigned int crtc_h,
                        uint32_t src_x, uint32_t src_y,
                        uint32_t src_w, uint32_t src_h);</synopsis>
            <para>
              Enable and configure the plane to use the given CRTC and frame buffer.
            </para>
            <para>
              The source rectangle in frame buffer memory coordinates is given by
              the <parameter>src_x</parameter>, <parameter>src_y</parameter>,
              <parameter>src_w</parameter> and <parameter>src_h</parameter>
              parameters (as 16.16 fixed point values). Devices that don't support
              subpixel plane coordinates can ignore the fractional part.
            </para>
            <para>
              The destination rectangle in CRTC coordinates is given by the
              <parameter>crtc_x</parameter>, <parameter>crtc_y</parameter>,
              <parameter>crtc_w</parameter> and <parameter>crtc_h</parameter>
              parameters (as integer values). Devices scale the source rectangle to
              the destination rectangle. If scaling is not supported, and the source
              rectangle size doesn't match the destination rectangle size, the
              driver must return a -<errorname>EINVAL</errorname> error.
            </para>
          </listitem>
          <listitem>
            <synopsis>int (*disable_plane)(struct drm_plane *plane);</synopsis>
            <para>
              Disable the plane. The DRM core calls this method in response to a
              DRM_IOCTL_MODE_SETPLANE ioctl call with the frame buffer ID set to 0.
              Disabled planes must not be processed by the CRTC.
            </para>
          </listitem>
          <listitem>
            <synopsis>void (*destroy)(struct drm_plane *plane);</synopsis>
            <para>
              Destroy the plane when not needed anymore. See
              <xref linkend="drm-kms-init"/>.
            </para>
          </listitem>
        </itemizedlist>
      </sect3>
    </sect2>
    <sect2>
      <title>Encoders (struct <structname>drm_encoder</structname>)</title>
      <para>
        An encoder takes pixel data from a CRTC and converts it to a format
	suitable for any attached connectors. On some devices, it may be
	possible to have a CRTC send data to more than one encoder. In that
	case, both encoders would receive data from the same scanout buffer,
	resulting in a "cloned" display configuration across the connectors
	attached to each encoder.
      </para>
      <sect3>
        <title>Encoder Initialization</title>
        <para>
          As for CRTCs, a KMS driver must create, initialize and register at
          least one struct <structname>drm_encoder</structname> instance. The
          instance is allocated and zeroed by the driver, possibly as part of a
          larger structure.
        </para>
        <para>
          Drivers must initialize the struct <structname>drm_encoder</structname>
          <structfield>possible_crtcs</structfield> and
          <structfield>possible_clones</structfield> fields before registering the
          encoder. Both fields are bitmasks of respectively the CRTCs that the
          encoder can be connected to, and sibling encoders candidate for cloning.
        </para>
        <para>
          After being initialized, the encoder must be registered with a call to
          <function>drm_encoder_init</function>. The function takes a pointer to
          the encoder functions and an encoder type. Supported types are
          <itemizedlist>
            <listitem>
              DRM_MODE_ENCODER_DAC for VGA and analog on DVI-I/DVI-A
              </listitem>
            <listitem>
              DRM_MODE_ENCODER_TMDS for DVI, HDMI and (embedded) DisplayPort
            </listitem>
            <listitem>
              DRM_MODE_ENCODER_LVDS for display panels
            </listitem>
            <listitem>
              DRM_MODE_ENCODER_TVDAC for TV output (Composite, S-Video, Component,
              SCART)
            </listitem>
            <listitem>
              DRM_MODE_ENCODER_VIRTUAL for virtual machine displays
            </listitem>
          </itemizedlist>
        </para>
        <para>
          Encoders must be attached to a CRTC to be used. DRM drivers leave
          encoders unattached at initialization time. Applications (or the fbdev
          compatibility layer when implemented) are responsible for attaching the
          encoders they want to use to a CRTC.
        </para>
      </sect3>
      <sect3>
        <title>Encoder Operations</title>
        <itemizedlist>
          <listitem>
            <synopsis>void (*destroy)(struct drm_encoder *encoder);</synopsis>
            <para>
              Called to destroy the encoder when not needed anymore. See
              <xref linkend="drm-kms-init"/>.
            </para>
          </listitem>
        </itemizedlist>
      </sect3>
    </sect2>
    <sect2>
      <title>Connectors (struct <structname>drm_connector</structname>)</title>
      <para>
        A connector is the final destination for pixel data on a device, and
	usually connects directly to an external display device like a monitor
	or laptop panel. A connector can only be attached to one encoder at a
	time. The connector is also the structure where information about the
	attached display is kept, so it contains fields for display data, EDID
	data, DPMS &amp; connection status, and information about modes
	supported on the attached displays.
      </para>
      <sect3>
        <title>Connector Initialization</title>
        <para>
          Finally a KMS driver must create, initialize, register and attach at
          least one struct <structname>drm_connector</structname> instance. The
          instance is created as other KMS objects and initialized by setting the
          following fields.
        </para>
        <variablelist>
          <varlistentry>
            <term><structfield>interlace_allowed</structfield></term>
            <listitem><para>
              Whether the connector can handle interlaced modes.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term><structfield>doublescan_allowed</structfield></term>
            <listitem><para>
              Whether the connector can handle doublescan.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term><structfield>display_info
            </structfield></term>
            <listitem><para>
              Display information is filled from EDID information when a display
              is detected. For non hot-pluggable displays such as flat panels in
              embedded systems, the driver should initialize the
              <structfield>display_info</structfield>.<structfield>width_mm</structfield>
              and
              <structfield>display_info</structfield>.<structfield>height_mm</structfield>
              fields with the physical size of the display.
            </para></listitem>
          </varlistentry>
          <varlistentry>
            <term id="drm-kms-connector-polled"><structfield>polled</structfield></term>
            <listitem><para>
              Connector polling mode, a combination of
              <variablelist>
                <varlistentry>
                  <term>DRM_CONNECTOR_POLL_HPD</term>
                  <listitem><para>
                    The connector generates hotplug events and doesn't need to be
                    periodically polled. The CONNECT and DISCONNECT flags must not
                    be set together with the HPD flag.
                  </para></listitem>
                </varlistentry>
                <varlistentry>
                  <term>DRM_CONNECTOR_POLL_CONNECT</term>
                  <listitem><para>
                    Periodically poll the connector for connection.
                  </para></listitem>
                </varlistentry>
                <varlistentry>
                  <term>DRM_CONNECTOR_POLL_DISCONNECT</term>
                  <listitem><para>
                    Periodically poll the connector for disconnection.
                  </para></listitem>
                </varlistentry>
              </variablelist>
              Set to 0 for connectors that don't support connection status
              discovery.
            </para></listitem>
          </varlistentry>
        </variablelist>
        <para>
          The connector is then registered with a call to
          <function>drm_connector_init</function> with a pointer to the connector
          functions and a connector type, and exposed through sysfs with a call to
          <function>drm_sysfs_connector_add</function>.
        </para>
        <para>
          Supported connector types are
          <itemizedlist>
            <listitem>DRM_MODE_CONNECTOR_VGA</listitem>
            <listitem>DRM_MODE_CONNECTOR_DVII</listitem>
            <listitem>DRM_MODE_CONNECTOR_DVID</listitem>
            <listitem>DRM_MODE_CONNECTOR_DVIA</listitem>
            <listitem>DRM_MODE_CONNECTOR_Composite</listitem>
            <listitem>DRM_MODE_CONNECTOR_SVIDEO</listitem>
            <listitem>DRM_MODE_CONNECTOR_LVDS</listitem>
            <listitem>DRM_MODE_CONNECTOR_Component</listitem>
            <listitem>DRM_MODE_CONNECTOR_9PinDIN</listitem>
            <listitem>DRM_MODE_CONNECTOR_DisplayPort</listitem>
            <listitem>DRM_MODE_CONNECTOR_HDMIA</listitem>
            <listitem>DRM_MODE_CONNECTOR_HDMIB</listitem>
            <listitem>DRM_MODE_CONNECTOR_TV</listitem>
            <listitem>DRM_MODE_CONNECTOR_eDP</listitem>
            <listitem>DRM_MODE_CONNECTOR_VIRTUAL</listitem>
          </itemizedlist>
        </para>
        <para>
          Connectors must be attached to an encoder to be used. For devices that
          map connectors to encoders 1:1, the connector should be attached at
          initialization time with a call to
          <function>drm_mode_connector_attach_encoder</function>. The driver must
          also set the <structname>drm_connector</structname>
          <structfield>encoder</structfield> field to point to the attached
          encoder.
        </para>
        <para>
          Finally, drivers must initialize the connectors state change detection
          with a call to <function>drm_kms_helper_poll_init</function>. If at
          least one connector is pollable but can't generate hotplug interrupts
          (indicated by the DRM_CONNECTOR_POLL_CONNECT and
          DRM_CONNECTOR_POLL_DISCONNECT connector flags), a delayed work will
          automatically be queued to periodically poll for changes. Connectors
          that can generate hotplug interrupts must be marked with the
          DRM_CONNECTOR_POLL_HPD flag instead, and their interrupt handler must
          call <function>drm_helper_hpd_irq_event</function>. The function will
          queue a delayed work to check the state of all connectors, but no
          periodic polling will be done.
        </para>
      </sect3>
      <sect3>
        <title>Connector Operations</title>
        <note><para>
          Unless otherwise state, all operations are mandatory.
        </para></note>
        <sect4>
          <title>DPMS</title>
          <synopsis>void (*dpms)(struct drm_connector *connector, int mode);</synopsis>
          <para>
            The DPMS operation sets the power state of a connector. The mode
            argument is one of
            <itemizedlist>
              <listitem><para>DRM_MODE_DPMS_ON</para></listitem>
              <listitem><para>DRM_MODE_DPMS_STANDBY</para></listitem>
              <listitem><para>DRM_MODE_DPMS_SUSPEND</para></listitem>
              <listitem><para>DRM_MODE_DPMS_OFF</para></listitem>
            </itemizedlist>
          </para>
          <para>
            In all but DPMS_ON mode the encoder to which the connector is attached
            should put the display in low-power mode by driving its signals
            appropriately. If more than one connector is attached to the encoder
            care should be taken not to change the power state of other displays as
            a side effect. Low-power mode should be propagated to the encoders and
            CRTCs when all related connectors are put in low-power mode.
          </para>
        </sect4>
        <sect4>
          <title>Modes</title>
          <synopsis>int (*fill_modes)(struct drm_connector *connector, uint32_t max_width,
                      uint32_t max_height);</synopsis>
          <para>
            Fill the mode list with all supported modes for the connector. If the
            <parameter>max_width</parameter> and <parameter>max_height</parameter>
            arguments are non-zero, the implementation must ignore all modes wider
            than <parameter>max_width</parameter> or higher than
            <parameter>max_height</parameter>.
          </para>
          <para>
            The connector must also fill in this operation its
            <structfield>display_info</structfield>
            <structfield>width_mm</structfield> and
            <structfield>height_mm</structfield> fields with the connected display
            physical size in millimeters. The fields should be set to 0 if the value
            isn't known or is not applicable (for instance for projector devices).
          </para>
        </sect4>
        <sect4>
          <title>Connection Status</title>
          <para>
            The connection status is updated through polling or hotplug events when
            supported (see <xref linkend="drm-kms-connector-polled"/>). The status
            value is reported to userspace through ioctls and must not be used
            inside the driver, as it only gets initialized by a call to
            <function>drm_mode_getconnector</function> from userspace.
          </para>
          <synopsis>enum drm_connector_status (*detect)(struct drm_connector *connector,
                                        bool force);</synopsis>
          <para>
            Check to see if anything is attached to the connector. The
            <parameter>force</parameter> parameter is set to false whilst polling or
            to true when checking the connector due to user request.
            <parameter>force</parameter> can be used by the driver to avoid
            expensive, destructive operations during automated probing.
          </para>
          <para>
            Return connector_status_connected if something is connected to the
            connector, connector_status_disconnected if nothing is connected and
            connector_status_unknown if the connection state isn't known.
          </para>
          <para>
            Drivers should only return connector_status_connected if the connection
            status has really been probed as connected. Connectors that can't detect
            the connection status, or failed connection status probes, should return
            connector_status_unknown.
          </para>
        </sect4>
        <sect4>
          <title>Miscellaneous</title>
          <itemizedlist>
            <listitem>
              <synopsis>void (*destroy)(struct drm_connector *connector);</synopsis>
              <para>
                Destroy the connector when not needed anymore. See
                <xref linkend="drm-kms-init"/>.
              </para>
            </listitem>
          </itemizedlist>
        </sect4>
      </sect3>
    </sect2>
    <sect2>
      <title>Cleanup</title>
      <para>
        The DRM core manages its objects' lifetime. When an object is not needed
	anymore the core calls its destroy function, which must clean up and
	free every resource allocated for the object. Every
	<function>drm_*_init</function> call must be matched with a
	corresponding <function>drm_*_cleanup</function> call to cleanup CRTCs
	(<function>drm_crtc_cleanup</function>), planes
	(<function>drm_plane_cleanup</function>), encoders
	(<function>drm_encoder_cleanup</function>) and connectors
	(<function>drm_connector_cleanup</function>). Furthermore, connectors
	that have been added to sysfs must be removed by a call to
	<function>drm_sysfs_connector_remove</function> before calling
	<function>drm_connector_cleanup</function>.
      </para>
      <para>
        Connectors state change detection must be cleanup up with a call to
	<function>drm_kms_helper_poll_fini</function>.
      </para>
    </sect2>
    <sect2>
      <title>Output discovery and initialization example</title>
      <programlisting><![CDATA[
void intel_crt_init(struct drm_device *dev)
{
	struct drm_connector *connector;
	struct intel_output *intel_output;

	intel_output = kzalloc(sizeof(struct intel_output), GFP_KERNEL);
	if (!intel_output)
		return;

	connector = &intel_output->base;
	drm_connector_init(dev, &intel_output->base,
			   &intel_crt_connector_funcs, DRM_MODE_CONNECTOR_VGA);

	drm_encoder_init(dev, &intel_output->enc, &intel_crt_enc_funcs,
			 DRM_MODE_ENCODER_DAC);

	drm_mode_connector_attach_encoder(&intel_output->base,
					  &intel_output->enc);

	/* Set up the DDC bus. */
	intel_output->ddc_bus = intel_i2c_create(dev, GPIOA, "CRTDDC_A");
	if (!intel_output->ddc_bus) {
		dev_printk(KERN_ERR, &dev->pdev->dev, "DDC bus registration "
			   "failed.\n");
		return;
	}

	intel_output->type = INTEL_OUTPUT_ANALOG;
	connector->interlace_allowed = 0;
	connector->doublescan_allowed = 0;

	drm_encoder_helper_add(&intel_output->enc, &intel_crt_helper_funcs);
	drm_connector_helper_add(connector, &intel_crt_connector_helper_funcs);

	drm_sysfs_connector_add(connector);
}]]></programlisting>
      <para>
        In the example above (taken from the i915 driver), a CRTC, connector and
        encoder combination is created. A device-specific i2c bus is also
        created for fetching EDID data and performing monitor detection. Once
        the process is complete, the new connector is registered with sysfs to
        make its properties available to applications.
      </para>
    </sect2>
  </sect1>

  <!-- Internals: mid-layer helper functions -->

  <sect1>
    <title>Mid-layer Helper Functions</title>
    <para>
      The CRTC, encoder and connector functions provided by the drivers
      implement the DRM API. They're called by the DRM core and ioctl handlers
      to handle device state changes and configuration request. As implementing
      those functions often requires logic not specific to drivers, mid-layer
      helper functions are available to avoid duplicating boilerplate code.
    </para>
    <para>
      The DRM core contains one mid-layer implementation. The mid-layer provides
      implementations of several CRTC, encoder and connector functions (called
      from the top of the mid-layer) that pre-process requests and call
      lower-level functions provided by the driver (at the bottom of the
      mid-layer). For instance, the
      <function>drm_crtc_helper_set_config</function> function can be used to
      fill the struct <structname>drm_crtc_funcs</structname>
      <structfield>set_config</structfield> field. When called, it will split
      the <methodname>set_config</methodname> operation in smaller, simpler
      operations and call the driver to handle them.
    </para>
    <para>
      To use the mid-layer, drivers call <function>drm_crtc_helper_add</function>,
      <function>drm_encoder_helper_add</function> and
      <function>drm_connector_helper_add</function> functions to install their
      mid-layer bottom operations handlers, and fill the
      <structname>drm_crtc_funcs</structname>,
      <structname>drm_encoder_funcs</structname> and
      <structname>drm_connector_funcs</structname> structures with pointers to
      the mid-layer top API functions. Installing the mid-layer bottom operation
      handlers is best done right after registering the corresponding KMS object.
    </para>
    <para>
      The mid-layer is not split between CRTC, encoder and connector operations.
      To use it, a driver must provide bottom functions for all of the three KMS
      entities.
    </para>
    <sect2>
      <title>Helper Functions</title>
      <itemizedlist>
        <listitem>
          <synopsis>int drm_crtc_helper_set_config(struct drm_mode_set *set);</synopsis>
          <para>
            The <function>drm_crtc_helper_set_config</function> helper function
            is a CRTC <methodname>set_config</methodname> implementation. It
            first tries to locate the best encoder for each connector by calling
            the connector <methodname>best_encoder</methodname> helper
            operation.
          </para>
          <para>
            After locating the appropriate encoders, the helper function will
            call the <methodname>mode_fixup</methodname> encoder and CRTC helper
            operations to adjust the requested mode, or reject it completely in
            which case an error will be returned to the application. If the new
            configuration after mode adjustment is identical to the current
            configuration the helper function will return without performing any
            other operation.
          </para>
          <para>
            If the adjusted mode is identical to the current mode but changes to
            the frame buffer need to be applied, the
            <function>drm_crtc_helper_set_config</function> function will call
            the CRTC <methodname>mode_set_base</methodname> helper operation. If
            the adjusted mode differs from the current mode, or if the
            <methodname>mode_set_base</methodname> helper operation is not
            provided, the helper function performs a full mode set sequence by
            calling the <methodname>prepare</methodname>,
            <methodname>mode_set</methodname> and
            <methodname>commit</methodname> CRTC and encoder helper operations,
            in that order.
          </para>
        </listitem>
        <listitem>
          <synopsis>void drm_helper_connector_dpms(struct drm_connector *connector, int mode);</synopsis>
          <para>
            The <function>drm_helper_connector_dpms</function> helper function
            is a connector <methodname>dpms</methodname> implementation that
            tracks power state of connectors. To use the function, drivers must
            provide <methodname>dpms</methodname> helper operations for CRTCs
            and encoders to apply the DPMS state to the device.
          </para>
          <para>
            The mid-layer doesn't track the power state of CRTCs and encoders.
            The <methodname>dpms</methodname> helper operations can thus be
            called with a mode identical to the currently active mode.
          </para>
        </listitem>
        <listitem>
          <synopsis>int drm_helper_probe_single_connector_modes(struct drm_connector *connector,
                                            uint32_t maxX, uint32_t maxY);</synopsis>
          <para>
            The <function>drm_helper_probe_single_connector_modes</function> helper
            function is a connector <methodname>fill_modes</methodname>
            implementation that updates the connection status for the connector
            and then retrieves a list of modes by calling the connector
            <methodname>get_modes</methodname> helper operation.
          </para>
          <para>
            The function filters out modes larger than
            <parameter>max_width</parameter> and <parameter>max_height</parameter>
            if specified. It then calls the connector
            <methodname>mode_valid</methodname> helper operation for  each mode in
            the probed list to check whether the mode is valid for the connector.
          </para>
        </listitem>
      </itemizedlist>
    </sect2>
    <sect2>
      <title>CRTC Helper Operations</title>
      <itemizedlist>
        <listitem id="drm-helper-crtc-mode-fixup">
          <synopsis>bool (*mode_fixup)(struct drm_crtc *crtc,
                       const struct drm_display_mode *mode,
                       struct drm_display_mode *adjusted_mode);</synopsis>
          <para>
            Let CRTCs adjust the requested mode or reject it completely. This
            operation returns true if the mode is accepted (possibly after being
            adjusted) or false if it is rejected.
          </para>
          <para>
            The <methodname>mode_fixup</methodname> operation should reject the
            mode if it can't reasonably use it. The definition of "reasonable"
            is currently fuzzy in this context. One possible behaviour would be
            to set the adjusted mode to the panel timings when a fixed-mode
            panel is used with hardware capable of scaling. Another behaviour
            would be to accept any input mode and adjust it to the closest mode
            supported by the hardware (FIXME: This needs to be clarified).
          </para>
        </listitem>
        <listitem>
          <synopsis>int (*mode_set_base)(struct drm_crtc *crtc, int x, int y,
                     struct drm_framebuffer *old_fb)</synopsis>
          <para>
            Move the CRTC on the current frame buffer (stored in
            <literal>crtc-&gt;fb</literal>) to position (x,y). Any of the frame
            buffer, x position or y position may have been modified.
          </para>
          <para>
            This helper operation is optional. If not provided, the
            <function>drm_crtc_helper_set_config</function> function will fall
            back to the <methodname>mode_set</methodname> helper operation.
          </para>
          <note><para>
            FIXME: Why are x and y passed as arguments, as they can be accessed
            through <literal>crtc-&gt;x</literal> and
            <literal>crtc-&gt;y</literal>?
          </para></note>
        </listitem>
        <listitem>
          <synopsis>void (*prepare)(struct drm_crtc *crtc);</synopsis>
          <para>
            Prepare the CRTC for mode setting. This operation is called after
            validating the requested mode. Drivers use it to perform
            device-specific operations required before setting the new mode.
          </para>
        </listitem>
        <listitem>
          <synopsis>int (*mode_set)(struct drm_crtc *crtc, struct drm_display_mode *mode,
                struct drm_display_mode *adjusted_mode, int x, int y,
                struct drm_framebuffer *old_fb);</synopsis>
          <para>
            Set a new mode, position and frame buffer. Depending on the device
            requirements, the mode can be stored internally by the driver and
            applied in the <methodname>commit</methodname> operation, or
            programmed to the hardware immediately.
          </para>
          <para>
            The <methodname>mode_set</methodname> operation returns 0 on success
	    or a negative error code if an error occurs.
          </para>
        </listitem>
        <listitem>
          <synopsis>void (*commit)(struct drm_crtc *crtc);</synopsis>
          <para>
            Commit a mode. This operation is called after setting the new mode.
            Upon return the device must use the new mode and be fully
            operational.
          </para>
        </listitem>
      </itemizedlist>
    </sect2>
    <sect2>
      <title>Encoder Helper Operations</title>
      <itemizedlist>
        <listitem>
          <synopsis>bool (*mode_fixup)(struct drm_encoder *encoder,
                       const struct drm_display_mode *mode,
                       struct drm_display_mode *adjusted_mode);</synopsis>
          <note><para>
            FIXME: The mode argument be const, but the i915 driver modifies
            mode-&gt;clock in <function>intel_dp_mode_fixup</function>.
          </para></note>
          <para>
            Let encoders adjust the requested mode or reject it completely. This
            operation returns true if the mode is accepted (possibly after being
            adjusted) or false if it is rejected. See the
            <link linkend="drm-helper-crtc-mode-fixup">mode_fixup CRTC helper
            operation</link> for an explanation of the allowed adjustments.
          </para>
        </listitem>
        <listitem>
          <synopsis>void (*prepare)(struct drm_encoder *encoder);</synopsis>
          <para>
            Prepare the encoder for mode setting. This operation is called after
            validating the requested mode. Drivers use it to perform
            device-specific operations required before setting the new mode.
          </para>
        </listitem>
        <listitem>
          <synopsis>void (*mode_set)(struct drm_encoder *encoder,
                 struct drm_display_mode *mode,
                 struct drm_display_mode *adjusted_mode);</synopsis>
          <para>
            Set a new mode. Depending on the device requirements, the mode can
            be stored internally by the driver and applied in the
            <methodname>commit</methodname> operation, or programmed to the
            hardware immediately.
          </para>
        </listitem>
        <listitem>
          <synopsis>void (*commit)(struct drm_encoder *encoder);</synopsis>
          <para>
            Commit a mode. This operation is called after setting the new mode.
            Upon return the device must use the new mode and be fully
            operational.
          </para>
        </listitem>
      </itemizedlist>
    </sect2>
    <sect2>
      <title>Connector Helper Operations</title>
      <itemizedlist>
        <listitem>
          <synopsis>struct drm_encoder *(*best_encoder)(struct drm_connector *connector);</synopsis>
          <para>
            Return a pointer to the best encoder for the connecter. Device that
            map connectors to encoders 1:1 simply return the pointer to the
            associated encoder. This operation is mandatory.
          </para>
        </listitem>
        <listitem>
          <synopsis>int (*get_modes)(struct drm_connector *connector);</synopsis>
          <para>
            Fill the connector's <structfield>probed_modes</structfield> list
            by parsing EDID data with <function>drm_add_edid_modes</function> or
            calling <function>drm_mode_probed_add</function> directly for every
            supported mode and return the number of modes it has detected. This
            operation is mandatory.
          </para>
          <para>
            When adding modes manually the driver creates each mode with a call to
            <function>drm_mode_create</function> and must fill the following fields.
            <itemizedlist>
              <listitem>
                <synopsis>__u32 type;</synopsis>
                <para>
                  Mode type bitmask, a combination of
                  <variablelist>
                    <varlistentry>
                      <term>DRM_MODE_TYPE_BUILTIN</term>
                      <listitem><para>not used?</para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_TYPE_CLOCK_C</term>
                      <listitem><para>not used?</para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_TYPE_CRTC_C</term>
                      <listitem><para>not used?</para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>
        DRM_MODE_TYPE_PREFERRED - The preferred mode for the connector
                      </term>
                      <listitem>
                        <para>not used?</para>
                      </listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_TYPE_DEFAULT</term>
                      <listitem><para>not used?</para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_TYPE_USERDEF</term>
                      <listitem><para>not used?</para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_TYPE_DRIVER</term>
                      <listitem>
                        <para>
                          The mode has been created by the driver (as opposed to
                          to user-created modes).
                        </para>
                      </listitem>
                    </varlistentry>
                  </variablelist>
                  Drivers must set the DRM_MODE_TYPE_DRIVER bit for all modes they
                  create, and set the DRM_MODE_TYPE_PREFERRED bit for the preferred
                  mode.
                </para>
              </listitem>
              <listitem>
                <synopsis>__u32 clock;</synopsis>
                <para>Pixel clock frequency in kHz unit</para>
              </listitem>
              <listitem>
                <synopsis>__u16 hdisplay, hsync_start, hsync_end, htotal;
    __u16 vdisplay, vsync_start, vsync_end, vtotal;</synopsis>
                <para>Horizontal and vertical timing information</para>
                <screen><![CDATA[
             Active                 Front           Sync           Back
             Region                 Porch                          Porch
    <-----------------------><----------------><-------------><-------------->

      //////////////////////|
     ////////////////////// |
    //////////////////////  |..................               ................
                                               _______________

    <----- [hv]display ----->
    <------------- [hv]sync_start ------------>
    <--------------------- [hv]sync_end --------------------->
    <-------------------------------- [hv]total ----------------------------->
]]></screen>
              </listitem>
              <listitem>
                <synopsis>__u16 hskew;
    __u16 vscan;</synopsis>
                <para>Unknown</para>
              </listitem>
              <listitem>
                <synopsis>__u32 flags;</synopsis>
                <para>
                  Mode flags, a combination of
                  <variablelist>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_PHSYNC</term>
                      <listitem><para>
                        Horizontal sync is active high
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_NHSYNC</term>
                      <listitem><para>
                        Horizontal sync is active low
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_PVSYNC</term>
                      <listitem><para>
                        Vertical sync is active high
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_NVSYNC</term>
                      <listitem><para>
                        Vertical sync is active low
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_INTERLACE</term>
                      <listitem><para>
                        Mode is interlaced
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_DBLSCAN</term>
                      <listitem><para>
                        Mode uses doublescan
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_CSYNC</term>
                      <listitem><para>
                        Mode uses composite sync
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_PCSYNC</term>
                      <listitem><para>
                        Composite sync is active high
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_NCSYNC</term>
                      <listitem><para>
                        Composite sync is active low
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_HSKEW</term>
                      <listitem><para>
                        hskew provided (not used?)
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_BCAST</term>
                      <listitem><para>
                        not used?
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_PIXMUX</term>
                      <listitem><para>
                        not used?
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_DBLCLK</term>
                      <listitem><para>
                        not used?
                      </para></listitem>
                    </varlistentry>
                    <varlistentry>
                      <term>DRM_MODE_FLAG_CLKDIV2</term>
                      <listitem><para>
                        ?
                      </para></listitem>
                    </varlistentry>
                  </variablelist>
                </para>
                <para>
                  Note that modes marked with the INTERLACE or DBLSCAN flags will be
                  filtered out by
                  <function>drm_helper_probe_single_connector_modes</function> if
                  the connector's <structfield>interlace_allowed</structfield> or
                  <structfield>doublescan_allowed</structfield> field is set to 0.
                </para>
              </listitem>
              <listitem>
                <synopsis>char name[DRM_DISPLAY_MODE_LEN];</synopsis>
                <para>
                  Mode name. The driver must call
                  <function>drm_mode_set_name</function> to fill the mode name from
                  <structfield>hdisplay</structfield>,
                  <structfield>vdisplay</structfield> and interlace flag after
                  filling the corresponding fields.
                </para>
              </listitem>
            </itemizedlist>
          </para>
          <para>
            The <structfield>vrefresh</structfield> value is computed by
            <function>drm_helper_probe_single_connector_modes</function>.
          </para>
          <para>
            When parsing EDID data, <function>drm_add_edid_modes</function> fill the
            connector <structfield>display_info</structfield>
            <structfield>width_mm</structfield> and
            <structfield>height_mm</structfield> fields. When creating modes
            manually the <methodname>get_modes</methodname> helper operation must
            set the <structfield>display_info</structfield>
            <structfield>width_mm</structfield> and
            <structfield>height_mm</structfield> fields if they haven't been set
            already (for instance at initilization time when a fixed-size panel is
            attached to the connector). The mode <structfield>width_mm</structfield>
            and <structfield>height_mm</structfield> fields are only used internally
            during EDID parsing and should not be set when creating modes manually.
          </para>
        </listitem>
        <listitem>
          <synopsis>int (*mode_valid)(struct drm_connector *connector,
		  struct drm_display_mode *mode);</synopsis>
          <para>
            Verify whether a mode is valid for the connector. Return MODE_OK for
            supported modes and one of the enum drm_mode_status values (MODE_*)
            for unsupported modes. This operation is mandatory.
          </para>
          <para>
            As the mode rejection reason is currently not used beside for
            immediately removing the unsupported mode, an implementation can
            return MODE_BAD regardless of the exact reason why the mode is not
            valid.
          </para>
          <note><para>
            Note that the <methodname>mode_valid</methodname> helper operation is
            only called for modes detected by the device, and
            <emphasis>not</emphasis> for modes set by the user through the CRTC
            <methodname>set_config</methodname> operation.
          </para></note>
        </listitem>
      </itemizedlist>
    </sect2>
  </sect1>

  <!-- Internals: vertical blanking -->

  <sect1 id="drm-vertical-blank">
    <title>Vertical Blanking</title>
    <para>
      Vertical blanking plays a major role in graphics rendering. To achieve
      tear-free display, users must synchronize page flips and/or rendering to
      vertical blanking. The DRM API offers ioctls to perform page flips
      synchronized to vertical blanking and wait for vertical blanking.
    </para>
    <para>
      The DRM core handles most of the vertical blanking management logic, which
      involves filtering out spurious interrupts, keeping race-free blanking
      counters, coping with counter wrap-around and resets and keeping use
      counts. It relies on the driver to generate vertical blanking interrupts
      and optionally provide a hardware vertical blanking counter. Drivers must
      implement the following operations.
    </para>
    <itemizedlist>
      <listitem>
        <synopsis>int (*enable_vblank) (struct drm_device *dev, int crtc);
void (*disable_vblank) (struct drm_device *dev, int crtc);</synopsis>
        <para>
	  Enable or disable vertical blanking interrupts for the given CRTC.
	</para>
      </listitem>
      <listitem>
        <synopsis>u32 (*get_vblank_counter) (struct drm_device *dev, int crtc);</synopsis>
        <para>
	  Retrieve the value of the vertical blanking counter for the given
	  CRTC. If the hardware maintains a vertical blanking counter its value
	  should be returned. Otherwise drivers can use the
	  <function>drm_vblank_count</function> helper function to handle this
	  operation.
	</para>
      </listitem>
    </itemizedlist>
    <para>
      Drivers must initialize the vertical blanking handling core with a call to
      <function>drm_vblank_init</function> in their
      <methodname>load</methodname> operation. The function will set the struct
      <structname>drm_device</structname>
      <structfield>vblank_disable_allowed</structfield> field to 0. This will
      keep vertical blanking interrupts enabled permanently until the first mode
      set operation, where <structfield>vblank_disable_allowed</structfield> is
      set to 1. The reason behind this is not clear. Drivers can set the field
      to 1 after <function>calling drm_vblank_init</function> to make vertical
      blanking interrupts dynamically managed from the beginning.
    </para>
    <para>
      Vertical blanking interrupts can be enabled by the DRM core or by drivers
      themselves (for instance to handle page flipping operations). The DRM core
      maintains a vertical blanking use count to ensure that the interrupts are
      not disabled while a user still needs them. To increment the use count,
      drivers call <function>drm_vblank_get</function>. Upon return vertical
      blanking interrupts are guaranteed to be enabled.
    </para>
    <para>
      To decrement the use count drivers call
      <function>drm_vblank_put</function>. Only when the use count drops to zero
      will the DRM core disable the vertical blanking interrupts after a delay
      by scheduling a timer. The delay is accessible through the vblankoffdelay
      module parameter or the <varname>drm_vblank_offdelay</varname> global
      variable and expressed in milliseconds. Its default value is 5000 ms.
    </para>
    <para>
      When a vertical blanking interrupt occurs drivers only need to call the
      <function>drm_handle_vblank</function> function to account for the
      interrupt.
    </para>
    <para>
      Resources allocated by <function>drm_vblank_init</function> must be freed
      with a call to <function>drm_vblank_cleanup</function> in the driver
      <methodname>unload</methodname> operation handler.
    </para>
  </sect1>

  <!-- Internals: open/close, file operations and ioctls -->

  <sect1>
    <title>Open/Close, File Operations and IOCTLs</title>
    <sect2>
      <title>Open and Close</title>
      <synopsis>int (*firstopen) (struct drm_device *);
void (*lastclose) (struct drm_device *);
int (*open) (struct drm_device *, struct drm_file *);
void (*preclose) (struct drm_device *, struct drm_file *);
void (*postclose) (struct drm_device *, struct drm_file *);</synopsis>
      <abstract>Open and close handlers. None of those methods are mandatory.
      </abstract>
      <para>
        The <methodname>firstopen</methodname> method is called by the DRM core
	when an application opens a device that has no other opened file handle.
	Similarly the <methodname>lastclose</methodname> method is called when
	the last application holding a file handle opened on the device closes
	it. Both methods are mostly used for UMS (User Mode Setting) drivers to
	acquire and release device resources which should be done in the
	<methodname>load</methodname> and <methodname>unload</methodname>
	methods for KMS drivers.
      </para>
      <para>
        Note that the <methodname>lastclose</methodname> method is also called
	at module unload time or, for hot-pluggable devices, when the device is
	unplugged. The <methodname>firstopen</methodname> and
	<methodname>lastclose</methodname> calls can thus be unbalanced.
      </para>
      <para>
        The <methodname>open</methodname> method is called every time the device
	is opened by an application. Drivers can allocate per-file private data
	in this method and store them in the struct
	<structname>drm_file</structname> <structfield>driver_priv</structfield>
	field. Note that the <methodname>open</methodname> method is called
	before <methodname>firstopen</methodname>.
      </para>
      <para>
        The close operation is split into <methodname>preclose</methodname> and
	<methodname>postclose</methodname> methods. Drivers must stop and
	cleanup all per-file operations in the <methodname>preclose</methodname>
	method. For instance pending vertical blanking and page flip events must
	be cancelled. No per-file operation is allowed on the file handle after
	returning from the <methodname>preclose</methodname> method.
      </para>
      <para>
        Finally the <methodname>postclose</methodname> method is called as the
	last step of the close operation, right before calling the
	<methodname>lastclose</methodname> method if no other open file handle
	exists for the device. Drivers that have allocated per-file private data
	in the <methodname>open</methodname> method should free it here.
      </para>
      <para>
        The <methodname>lastclose</methodname> method should restore CRTC and
	plane properties to default value, so that a subsequent open of the
	device will not inherit state from the previous user.
      </para>
    </sect2>
    <sect2>
      <title>File Operations</title>
      <synopsis>const struct file_operations *fops</synopsis>
      <abstract>File operations for the DRM device node.</abstract>
      <para>
        Drivers must define the file operations structure that forms the DRM
	userspace API entry point, even though most of those operations are
	implemented in the DRM core. The <methodname>open</methodname>,
	<methodname>release</methodname> and <methodname>ioctl</methodname>
	operations are handled by
	<programlisting>
	.owner = THIS_MODULE,
	.open = drm_open,
	.release = drm_release,
	.unlocked_ioctl = drm_ioctl,
  #ifdef CONFIG_COMPAT
	.compat_ioctl = drm_compat_ioctl,
  #endif
        </programlisting>
      </para>
      <para>
        Drivers that implement private ioctls that requires 32/64bit
	compatibility support must provide their own
	<methodname>compat_ioctl</methodname> handler that processes private
	ioctls and calls <function>drm_compat_ioctl</function> for core ioctls.
      </para>
      <para>
        The <methodname>read</methodname> and <methodname>poll</methodname>
	operations provide support for reading DRM events and polling them. They
	are implemented by
	<programlisting>
	.poll = drm_poll,
	.read = drm_read,
	.fasync = drm_fasync,
	.llseek = no_llseek,
	</programlisting>
      </para>
      <para>
        The memory mapping implementation varies depending on how the driver
	manages memory. Pre-GEM drivers will use <function>drm_mmap</function>,
	while GEM-aware drivers will use <function>drm_gem_mmap</function>. See
	<xref linkend="drm-gem"/>.
	<programlisting>
	.mmap = drm_gem_mmap,
	</programlisting>
      </para>
      <para>
        No other file operation is supported by the DRM API.
      </para>
    </sect2>
    <sect2>
      <title>IOCTLs</title>
      <synopsis>struct drm_ioctl_desc *ioctls;
int num_ioctls;</synopsis>
      <abstract>Driver-specific ioctls descriptors table.</abstract>
      <para>
        Driver-specific ioctls numbers start at DRM_COMMAND_BASE. The ioctls
	descriptors table is indexed by the ioctl number offset from the base
	value. Drivers can use the DRM_IOCTL_DEF_DRV() macro to initialize the
	table entries.
      </para>
      <para>
        <programlisting>DRM_IOCTL_DEF_DRV(ioctl, func, flags)</programlisting>
	<para>
	  <parameter>ioctl</parameter> is the ioctl name. Drivers must define
	  the DRM_##ioctl and DRM_IOCTL_##ioctl macros to the ioctl number
	  offset from DRM_COMMAND_BASE and the ioctl number respectively. The
	  first macro is private to the device while the second must be exposed
	  to userspace in a public header.
	</para>
	<para>
	  <parameter>func</parameter> is a pointer to the ioctl handler function
	  compatible with the <type>drm_ioctl_t</type> type.
	  <programlisting>typedef int drm_ioctl_t(struct drm_device *dev, void *data,
		struct drm_file *file_priv);</programlisting>
	</para>
	<para>
	  <parameter>flags</parameter> is a bitmask combination of the following
	  values. It restricts how the ioctl is allowed to be called.
	  <itemizedlist>
	    <listitem><para>
	      DRM_AUTH - Only authenticated callers allowed
	    </para></listitem>
	    <listitem><para>
	      DRM_MASTER - The ioctl can only be called on the master file
	      handle
	    </para></listitem>
            <listitem><para>
	      DRM_ROOT_ONLY - Only callers with the SYSADMIN capability allowed
	    </para></listitem>
            <listitem><para>
	      DRM_CONTROL_ALLOW - The ioctl can only be called on a control
	      device
	    </para></listitem>
            <listitem><para>
	      DRM_UNLOCKED - The ioctl handler will be called without locking
	      the DRM global mutex
	    </para></listitem>
	  </itemizedlist>
	</para>
      </para>
    </sect2>
  </sect1>

  <sect1>
    <title>Command submission &amp; fencing</title>
    <para>
      This should cover a few device-specific command submission
      implementations.
    </para>
  </sect1>

  <!-- Internals: suspend/resume -->

  <sect1>
    <title>Suspend/Resume</title>
    <para>
      The DRM core provides some suspend/resume code, but drivers wanting full
      suspend/resume support should provide save() and restore() functions.
      These are called at suspend, hibernate, or resume time, and should perform
      any state save or restore required by your device across suspend or
      hibernate states.
    </para>
    <synopsis>int (*suspend) (struct drm_device *, pm_message_t state);
int (*resume) (struct drm_device *);</synopsis>
    <para>
      Those are legacy suspend and resume methods. New driver should use the
      power management interface provided by their bus type (usually through
      the struct <structname>device_driver</structname> dev_pm_ops) and set
      these methods to NULL.
    </para>
  </sect1>

  <sect1>
    <title>DMA services</title>
    <para>
      This should cover how DMA mapping etc. is supported by the core.
      These functions are deprecated and should not be used.
    </para>
  </sect1>
  </chapter>

<!-- TODO

- Add a glossary
- Document the struct_mutex catch-all lock
- Document connector properties

- Why is the load method optional?
- What are drivers supposed to set the initial display state to, and how?
  Connector's DPMS states are not initialized and are thus equal to
  DRM_MODE_DPMS_ON. The fbcon compatibility layer calls
  drm_helper_disable_unused_functions(), which disables unused encoders and
  CRTCs, but doesn't touch the connectors' DPMS state, and
  drm_helper_connector_dpms() in reaction to fbdev blanking events. Do drivers
  that don't implement (or just don't use) fbcon compatibility need to call
  those functions themselves?
- KMS drivers must call drm_vblank_pre_modeset() and drm_vblank_post_modeset()
  around mode setting. Should this be done in the DRM core?
- vblank_disable_allowed is set to 1 in the first drm_vblank_post_modeset()
  call and never set back to 0. It seems to be safe to permanently set it to 1
  in drm_vblank_init() for KMS driver, and it might be safe for UMS drivers as
  well. This should be investigated.
- crtc and connector .save and .restore operations are only used internally in
  drivers, should they be removed from the core?
- encoder mid-layer .save and .restore operations are only used internally in
  drivers, should they be removed from the core?
- encoder mid-layer .detect operation is only used internally in drivers,
  should it be removed from the core?
-->

  <!-- External interfaces -->

  <chapter id="drmExternals">
    <title>Userland interfaces</title>
    <para>
      The DRM core exports several interfaces to applications,
      generally intended to be used through corresponding libdrm
      wrapper functions.  In addition, drivers export device-specific
      interfaces for use by userspace drivers &amp; device-aware
      applications through ioctls and sysfs files.
    </para>
    <para>
      External interfaces include: memory mapping, context management,
      DMA operations, AGP management, vblank control, fence
      management, memory management, and output management.
    </para>
    <para>
      Cover generic ioctls and sysfs layout here.  We only need high-level
      info, since man pages should cover the rest.
    </para>

  <!-- External: vblank handling -->

    <sect1>
      <title>VBlank event handling</title>
      <para>
        The DRM core exposes two vertical blank related ioctls:
        <variablelist>
          <varlistentry>
            <term>DRM_IOCTL_WAIT_VBLANK</term>
            <listitem>
              <para>
                This takes a struct drm_wait_vblank structure as its argument,
                and it is used to block or request a signal when a specified
                vblank event occurs.
              </para>
            </listitem>
          </varlistentry>
          <varlistentry>
            <term>DRM_IOCTL_MODESET_CTL</term>
            <listitem>
              <para>
                This should be called by application level drivers before and
                after mode setting, since on many devices the vertical blank
                counter is reset at that time.  Internally, the DRM snapshots
                the last vblank count when the ioctl is called with the
                _DRM_PRE_MODESET command, so that the counter won't go backwards
                (which is dealt with when _DRM_POST_MODESET is used).
              </para>
            </listitem>
          </varlistentry>
        </variablelist>
<!--!Edrivers/char/drm/drm_irq.c-->
      </para>
    </sect1>

  </chapter>

  <!-- API reference -->

  <appendix id="drmDriverApi">
    <title>DRM Driver API</title>
    <para>
      Include auto-generated API reference here (need to reference it
      from paragraphs above too).
    </para>
  </appendix>

</book>