[PATCH] libata EH document update

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

diff --git a/Documentation/DocBook/libata.tmpl b/Documentation/DocBook/libata.tmpl
index 375ae76..fcbb58f 100644 (file)
--- a/Documentation/DocBook/libata.tmpl
+++ b/Documentation/DocBook/libata.tmpl
@@ -413,6 +413,362 @@ and other resources, etc.
        </sect2>
 
      </sect1>
+     <sect1>
+        <title>Error handling</title>
+
+       <para>
+       This chapter describes how errors are handled under libata.
+       Readers are advised to read SCSI EH
+       (Documentation/scsi/scsi_eh.txt) and ATA exceptions doc first.
+       </para>
+
+       <sect2><title>Origins of commands</title>
+       <para>
+       In libata, a command is represented with struct ata_queued_cmd
+       or qc.  qc's are preallocated during port initialization and
+       repetitively used for command executions.  Currently only one
+       qc is allocated per port but yet-to-be-merged NCQ branch
+       allocates one for each tag and maps each qc to NCQ tag 1-to-1.
+       </para>
+       <para>
+       libata commands can originate from two sources - libata itself
+       and SCSI midlayer.  libata internal commands are used for
+       initialization and error handling.  All normal blk requests
+       and commands for SCSI emulation are passed as SCSI commands
+       through queuecommand callback of SCSI host template.
+       </para>
+       </sect2>
+
+       <sect2><title>How commands are issued</title>
+
+       <variablelist>
+
+       <varlistentry><term>Internal commands</term>
+       <listitem>
+       <para>
+       First, qc is allocated and initialized using
+       ata_qc_new_init().  Although ata_qc_new_init() doesn't
+       implement any wait or retry mechanism when qc is not
+       available, internal commands are currently issued only during
+       initialization and error recovery, so no other command is
+       active and allocation is guaranteed to succeed.
+       </para>
+       <para>
+       Once allocated qc's taskfile is initialized for the command to
+       be executed.  qc currently has two mechanisms to notify
+       completion.  One is via qc->complete_fn() callback and the
+       other is completion qc->waiting.  qc->complete_fn() callback
+       is the asynchronous path used by normal SCSI translated
+       commands and qc->waiting is the synchronous (issuer sleeps in
+       process context) path used by internal commands.
+       </para>
+       <para>
+       Once initialization is complete, host_set lock is acquired
+       and the qc is issued.
+       </para>
+       </listitem>
+       </varlistentry>
+
+       <varlistentry><term>SCSI commands</term>
+       <listitem>
+       <para>
+       All libata drivers use ata_scsi_queuecmd() as
+       hostt->queuecommand callback.  scmds can either be simulated
+       or translated.  No qc is involved in processing a simulated
+       scmd.  The result is computed right away and the scmd is
+       completed.
+       </para>
+       <para>
+       For a translated scmd, ata_qc_new_init() is invoked to
+       allocate a qc and the scmd is translated into the qc.  SCSI
+       midlayer's completion notification function pointer is stored
+       into qc->scsidone.
+       </para>
+       <para>
+       qc->complete_fn() callback is used for completion
+       notification.  ATA commands use ata_scsi_qc_complete() while
+       ATAPI commands use atapi_qc_complete().  Both functions end up
+       calling qc->scsidone to notify upper layer when the qc is
+       finished.  After translation is completed, the qc is issued
+       with ata_qc_issue().
+       </para>
+       <para>
+       Note that SCSI midlayer invokes hostt->queuecommand while
+       holding host_set lock, so all above occur while holding
+       host_set lock.
+       </para>
+       </listitem>
+       </varlistentry>
+
+       </variablelist>
+       </sect2>
+
+       <sect2><title>How commands are processed</title>
+       <para>
+       Depending on which protocol and which controller are used,
+       commands are processed differently.  For the purpose of
+       discussion, a controller which uses taskfile interface and all
+       standard callbacks is assumed.
+       </para>
+       <para>
+       Currently 6 ATA command protocols are used.  They can be
+       sorted into the following four categories according to how
+       they are processed.
+       </para>
+
+       <variablelist>
+          <varlistentry><term>ATA NO DATA or DMA</term>
+          <listitem>
+          <para>
+          ATA_PROT_NODATA and ATA_PROT_DMA fall into this category.
+          These types of commands don't require any software
+          intervention once issued.  Device will raise interrupt on
+          completion.
+          </para>
+          </listitem>
+          </varlistentry>
+
+          <varlistentry><term>ATA PIO</term>
+          <listitem>
+          <para>
+          ATA_PROT_PIO is in this category.  libata currently
+          implements PIO with polling.  ATA_NIEN bit is set to turn
+          off interrupt and pio_task on ata_wq performs polling and
+          IO.
+          </para>
+          </listitem>
+          </varlistentry>
+
+          <varlistentry><term>ATAPI NODATA or DMA</term>
+          <listitem>
+          <para>
+          ATA_PROT_ATAPI_NODATA and ATA_PROT_ATAPI_DMA are in this
+          category.  packet_task is used to poll BSY bit after
+          issuing PACKET command.  Once BSY is turned off by the
+          device, packet_task transfers CDB and hands off processing
+          to interrupt handler.
+          </para>
+          </listitem>
+          </varlistentry>
+
+          <varlistentry><term>ATAPI PIO</term>
+          <listitem>
+          <para>
+          ATA_PROT_ATAPI is in this category.  ATA_NIEN bit is set
+          and, as in ATAPI NODATA or DMA, packet_task submits cdb.
+          However, after submitting cdb, further processing (data
+          transfer) is handed off to pio_task.
+          </para>
+          </listitem>
+          </varlistentry>
+       </variablelist>
+        </sect2>
+
+       <sect2><title>How commands are completed</title>
+       <para>
+       Once issued, all qc's are either completed with
+       ata_qc_complete() or time out.  For commands which are handled
+       by interrupts, ata_host_intr() invokes ata_qc_complete(), and,
+       for PIO tasks, pio_task invokes ata_qc_complete().  In error
+       cases, packet_task may also complete commands.
+       </para>
+       <para>
+       ata_qc_complete() does the following.
+       </para>
+
+       <orderedlist>
+
+       <listitem>
+       <para>
+       DMA memory is unmapped.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       ATA_QCFLAG_ACTIVE is clared from qc->flags.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       qc->complete_fn() callback is invoked.  If the return value of
+       the callback is not zero.  Completion is short circuited and
+       ata_qc_complete() returns.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       __ata_qc_complete() is called, which does
+          <orderedlist>
+
+          <listitem>
+          <para>
+          qc->flags is cleared to zero.
+          </para>
+          </listitem>
+
+          <listitem>
+          <para>
+          ap->active_tag and qc->tag are poisoned.
+          </para>
+          </listitem>
+
+          <listitem>
+          <para>
+          qc->waiting is claread &amp; completed (in that order).
+          </para>
+          </listitem>
+
+          <listitem>
+          <para>
+          qc is deallocated by clearing appropriate bit in ap->qactive.
+          </para>
+          </listitem>
+
+          </orderedlist>
+       </para>
+       </listitem>
+
+       </orderedlist>
+
+       <para>
+       So, it basically notifies upper layer and deallocates qc.  One
+       exception is short-circuit path in #3 which is used by
+       atapi_qc_complete().
+       </para>
+       <para>
+       For all non-ATAPI commands, whether it fails or not, almost
+       the same code path is taken and very little error handling
+       takes place.  A qc is completed with success status if it
+       succeeded, with failed status otherwise.
+       </para>
+       <para>
+       However, failed ATAPI commands require more handling as
+       REQUEST SENSE is needed to acquire sense data.  If an ATAPI
+       command fails, ata_qc_complete() is invoked with error status,
+       which in turn invokes atapi_qc_complete() via
+       qc->complete_fn() callback.
+       </para>
+       <para>
+       This makes atapi_qc_complete() set scmd->result to
+       SAM_STAT_CHECK_CONDITION, complete the scmd and return 1.  As
+       the sense data is empty but scmd->result is CHECK CONDITION,
+       SCSI midlayer will invoke EH for the scmd, and returning 1
+       makes ata_qc_complete() to return without deallocating the qc.
+       This leads us to ata_scsi_error() with partially completed qc.
+       </para>
+
+       </sect2>
+
+       <sect2><title>ata_scsi_error()</title>
+       <para>
+       ata_scsi_error() is the current hostt->eh_strategy_handler()
+       for libata.  As discussed above, this will be entered in two
+       cases - timeout and ATAPI error completion.  This function
+       calls low level libata driver's eng_timeout() callback, the
+       standard callback for which is ata_eng_timeout().  It checks
+       if a qc is active and calls ata_qc_timeout() on the qc if so.
+       Actual error handling occurs in ata_qc_timeout().
+       </para>
+       <para>
+       If EH is invoked for timeout, ata_qc_timeout() stops BMDMA and
+       completes the qc.  Note that as we're currently in EH, we
+       cannot call scsi_done.  As described in SCSI EH doc, a
+       recovered scmd should be either retried with
+       scsi_queue_insert() or finished with scsi_finish_command().
+       Here, we override qc->scsidone with scsi_finish_command() and
+       calls ata_qc_complete().
+       </para>
+       <para>
+       If EH is invoked due to a failed ATAPI qc, the qc here is
+       completed but not deallocated.  The purpose of this
+       half-completion is to use the qc as place holder to make EH
+       code reach this place.  This is a bit hackish, but it works.
+       </para>
+       <para>
+       Once control reaches here, the qc is deallocated by invoking
+       __ata_qc_complete() explicitly.  Then, internal qc for REQUEST
+       SENSE is issued.  Once sense data is acquired, scmd is
+       finished by directly invoking scsi_finish_command() on the
+       scmd.  Note that as we already have completed and deallocated
+       the qc which was associated with the scmd, we don't need
+       to/cannot call ata_qc_complete() again.
+       </para>
+
+       </sect2>
+
+       <sect2><title>Problems with the current EH</title>
+
+       <itemizedlist>
+
+       <listitem>
+       <para>
+       Error representation is too crude.  Currently any and all
+       error conditions are represented with ATA STATUS and ERROR
+       registers.  Errors which aren't ATA device errors are treated
+       as ATA device errors by setting ATA_ERR bit.  Better error
+       descriptor which can properly represent ATA and other
+       errors/exceptions is needed.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       When handling timeouts, no action is taken to make device
+       forget about the timed out command and ready for new commands.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       EH handling via ata_scsi_error() is not properly protected
+       from usual command processing.  On EH entrance, the device is
+       not in quiescent state.  Timed out commands may succeed or
+       fail any time.  pio_task and atapi_task may still be running.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       Too weak error recovery.  Devices / controllers causing HSM
+       mismatch errors and other errors quite often require reset to
+       return to known state.  Also, advanced error handling is
+       necessary to support features like NCQ and hotplug.
+       </para>
+       </listitem>
+
+       <listitem>
+       <para>
+       ATA errors are directly handled in the interrupt handler and
+       PIO errors in pio_task.  This is problematic for advanced
+       error handling for the following reasons.
+       </para>
+       <para>
+       First, advanced error handling often requires context and
+       internal qc execution.
+       </para>
+       <para>
+       Second, even a simple failure (say, CRC error) needs
+       information gathering and could trigger complex error handling
+       (say, resetting &amp; reconfiguring).  Having multiple code
+       paths to gather information, enter EH and trigger actions
+       makes life painful.
+       </para>
+       <para>
+       Third, scattered EH code makes implementing low level drivers
+       difficult.  Low level drivers override libata callbacks.  If
+       EH is scattered over several places, each affected callbacks
+       should perform its part of error handling.  This can be error
+       prone and painful.
+       </para>
+       </listitem>
+
+       </itemizedlist>
+       </sect2>
+
+     </sect1>
   </chapter>
 
   <chapter id="libataExt">