Merge branch 'upstream'
diff --git a/Documentation/DocBook/libata.tmpl b/Documentation/DocBook/libata.tmpl
index 375ae76..d260d92 100644
--- a/Documentation/DocBook/libata.tmpl
+++ b/Documentation/DocBook/libata.tmpl
@@ -415,6 +415,362 @@
      </sect1>
   </chapter>
 
+  <chapter id="libataEH">
+        <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>
+
+	<sect1><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>
+	</sect1>
+
+	<sect1><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>
+	</sect1>
+
+	<sect1><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>
+        </sect1>
+
+	<sect1><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>
+
+	</sect1>
+
+	<sect1><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>
+
+	</sect1>
+
+	<sect1><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>
+	</sect1>
+  </chapter>
+
   <chapter id="libataExt">
      <title>libata Library</title>
 !Edrivers/scsi/libata-core.c
@@ -431,6 +787,722 @@
 !Idrivers/scsi/libata-scsi.c
   </chapter>
 
+  <chapter id="ataExceptions">
+     <title>ATA errors &amp; exceptions</title>
+
+  <para>
+  This chapter tries to identify what error/exception conditions exist
+  for ATA/ATAPI devices and describe how they should be handled in
+  implementation-neutral way.
+  </para>
+
+  <para>
+  The term 'error' is used to describe conditions where either an
+  explicit error condition is reported from device or a command has
+  timed out.
+  </para>
+
+  <para>
+  The term 'exception' is either used to describe exceptional
+  conditions which are not errors (say, power or hotplug events), or
+  to describe both errors and non-error exceptional conditions.  Where
+  explicit distinction between error and exception is necessary, the
+  term 'non-error exception' is used.
+  </para>
+
+  <sect1 id="excat">
+     <title>Exception categories</title>
+     <para>
+     Exceptions are described primarily with respect to legacy
+     taskfile + bus master IDE interface.  If a controller provides
+     other better mechanism for error reporting, mapping those into
+     categories described below shouldn't be difficult.
+     </para>
+
+     <para>
+     In the following sections, two recovery actions - reset and
+     reconfiguring transport - are mentioned.  These are described
+     further in <xref linkend="exrec"/>.
+     </para>
+
+     <sect2 id="excatHSMviolation">
+        <title>HSM violation</title>
+        <para>
+        This error is indicated when STATUS value doesn't match HSM
+        requirement during issuing or excution any ATA/ATAPI command.
+        </para>
+
+	<itemizedlist>
+	<title>Examples</title>
+
+        <listitem>
+	<para>
+	ATA_STATUS doesn't contain !BSY &amp;&amp; DRDY &amp;&amp; !DRQ while trying
+	to issue a command.
+        </para>
+	</listitem>
+
+        <listitem>
+	<para>
+	!BSY &amp;&amp; !DRQ during PIO data transfer.
+        </para>
+	</listitem>
+
+        <listitem>
+	<para>
+	DRQ on command completion.
+        </para>
+	</listitem>
+
+        <listitem>
+	<para>
+	!BSY &amp;&amp; ERR after CDB tranfer starts but before the
+        last byte of CDB is transferred.  ATA/ATAPI standard states
+        that &quot;The device shall not terminate the PACKET command
+        with an error before the last byte of the command packet has
+        been written&quot; in the error outputs description of PACKET
+        command and the state diagram doesn't include such
+        transitions.
+	</para>
+	</listitem>
+
+	</itemizedlist>
+
+	<para>
+	In these cases, HSM is violated and not much information
+	regarding the error can be acquired from STATUS or ERROR
+	register.  IOW, this error can be anything - driver bug,
+	faulty device, controller and/or cable.
+	</para>
+
+	<para>
+	As HSM is violated, reset is necessary to restore known state.
+	Reconfiguring transport for lower speed might be helpful too
+	as transmission errors sometimes cause this kind of errors.
+	</para>
+     </sect2>
+     
+     <sect2 id="excatDevErr">
+        <title>ATA/ATAPI device error (non-NCQ / non-CHECK CONDITION)</title>
+
+	<para>
+	These are errors detected and reported by ATA/ATAPI devices
+	indicating device problems.  For this type of errors, STATUS
+	and ERROR register values are valid and describe error
+	condition.  Note that some of ATA bus errors are detected by
+	ATA/ATAPI devices and reported using the same mechanism as
+	device errors.  Those cases are described later in this
+	section.
+	</para>
+
+	<para>
+	For ATA commands, this type of errors are indicated by !BSY
+	&amp;&amp; ERR during command execution and on completion.
+	</para>
+
+	<para>For ATAPI commands,</para>
+
+	<itemizedlist>
+
+	<listitem>
+	<para>
+	!BSY &amp;&amp; ERR &amp;&amp; ABRT right after issuing PACKET
+	indicates that PACKET command is not supported and falls in
+	this category.
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	!BSY &amp;&amp; ERR(==CHK) &amp;&amp; !ABRT after the last
+	byte of CDB is transferred indicates CHECK CONDITION and
+	doesn't fall in this category.
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	!BSY &amp;&amp; ERR(==CHK) &amp;&amp; ABRT after the last byte
+        of CDB is transferred *probably* indicates CHECK CONDITION and
+        doesn't fall in this category.
+	</para>
+	</listitem>
+
+	</itemizedlist>
+
+	<para>
+	Of errors detected as above, the followings are not ATA/ATAPI
+	device errors but ATA bus errors and should be handled
+	according to <xref linkend="excatATAbusErr"/>.
+	</para>
+
+	<variablelist>
+
+	   <varlistentry>
+	   <term>CRC error during data transfer</term>
+	   <listitem>
+	   <para>
+	   This is indicated by ICRC bit in the ERROR register and
+	   means that corruption occurred during data transfer.  Upto
+	   ATA/ATAPI-7, the standard specifies that this bit is only
+	   applicable to UDMA transfers but ATA/ATAPI-8 draft revision
+	   1f says that the bit may be applicable to multiword DMA and
+	   PIO.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry>
+	   <term>ABRT error during data transfer or on completion</term>
+	   <listitem>
+	   <para>
+	   Upto ATA/ATAPI-7, the standard specifies that ABRT could be
+	   set on ICRC errors and on cases where a device is not able
+	   to complete a command.  Combined with the fact that MWDMA
+	   and PIO transfer errors aren't allowed to use ICRC bit upto
+	   ATA/ATAPI-7, it seems to imply that ABRT bit alone could
+	   indicate tranfer errors.
+	   </para>
+	   <para>
+	   However, ATA/ATAPI-8 draft revision 1f removes the part
+	   that ICRC errors can turn on ABRT.  So, this is kind of
+	   gray area.  Some heuristics are needed here.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	</variablelist>
+
+	<para>
+	ATA/ATAPI device errors can be further categorized as follows.
+	</para>
+
+	<variablelist>
+
+	   <varlistentry>
+	   <term>Media errors</term>
+	   <listitem>
+	   <para>
+	   This is indicated by UNC bit in the ERROR register.  ATA
+	   devices reports UNC error only after certain number of
+	   retries cannot recover the data, so there's nothing much
+	   else to do other than notifying upper layer.
+	   </para>
+	   <para>
+	   READ and WRITE commands report CHS or LBA of the first
+	   failed sector but ATA/ATAPI standard specifies that the
+	   amount of transferred data on error completion is
+	   indeterminate, so we cannot assume that sectors preceding
+	   the failed sector have been transferred and thus cannot
+	   complete those sectors successfully as SCSI does.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry>
+	   <term>Media changed / media change requested error</term>
+	   <listitem>
+	   <para>
+	   &lt;&lt;TODO: fill here&gt;&gt;
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry><term>Address error</term>
+	   <listitem>
+	   <para>
+	   This is indicated by IDNF bit in the ERROR register.
+	   Report to upper layer.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry><term>Other errors</term>
+	   <listitem>
+	   <para>
+	   This can be invalid command or parameter indicated by ABRT
+	   ERROR bit or some other error condition.  Note that ABRT
+	   bit can indicate a lot of things including ICRC and Address
+	   errors.  Heuristics needed.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	</variablelist>
+
+	<para>
+	Depending on commands, not all STATUS/ERROR bits are
+	applicable.  These non-applicable bits are marked with
+	&quot;na&quot; in the output descriptions but upto ATA/ATAPI-7
+	no definition of &quot;na&quot; can be found.  However,
+	ATA/ATAPI-8 draft revision 1f describes &quot;N/A&quot; as
+	follows.
+	</para>
+
+	<blockquote>
+	<variablelist>
+	   <varlistentry><term>3.2.3.3a N/A</term>
+	   <listitem>
+	   <para>
+	   A keyword the indicates a field has no defined value in
+	   this standard and should not be checked by the host or
+	   device. N/A fields should be cleared to zero.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+	</variablelist>
+	</blockquote>
+
+	<para>
+	So, it seems reasonable to assume that &quot;na&quot; bits are
+	cleared to zero by devices and thus need no explicit masking.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatATAPIcc">
+        <title>ATAPI device CHECK CONDITION</title>
+
+	<para>
+	ATAPI device CHECK CONDITION error is indicated by set CHK bit
+	(ERR bit) in the STATUS register after the last byte of CDB is
+	transferred for a PACKET command.  For this kind of errors,
+	sense data should be acquired to gather information regarding
+	the errors.  REQUEST SENSE packet command should be used to
+	acquire sense data.
+	</para>
+
+	<para>
+	Once sense data is acquired, this type of errors can be
+	handled similary to other SCSI errors.  Note that sense data
+	may indicate ATA bus error (e.g. Sense Key 04h HARDWARE ERROR
+	&amp;&amp; ASC/ASCQ 47h/00h SCSI PARITY ERROR).  In such
+	cases, the error should be considered as an ATA bus error and
+	handled according to <xref linkend="excatATAbusErr"/>.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatNCQerr">
+        <title>ATA device error (NCQ)</title>
+
+	<para>
+	NCQ command error is indicated by cleared BSY and set ERR bit
+	during NCQ command phase (one or more NCQ commands
+	outstanding).  Although STATUS and ERROR registers will
+	contain valid values describing the error, READ LOG EXT is
+	required to clear the error condition, determine which command
+	has failed and acquire more information.
+	</para>
+
+	<para>
+	READ LOG EXT Log Page 10h reports which tag has failed and
+	taskfile register values describing the error.  With this
+	information the failed command can be handled as a normal ATA
+	command error as in <xref linkend="excatDevErr"/> and all
+	other in-flight commands must be retried.  Note that this
+	retry should not be counted - it's likely that commands
+	retried this way would have completed normally if it were not
+	for the failed command.
+	</para>
+
+	<para>
+	Note that ATA bus errors can be reported as ATA device NCQ
+	errors.  This should be handled as described in <xref
+	linkend="excatATAbusErr"/>.
+	</para>
+
+	<para>
+	If READ LOG EXT Log Page 10h fails or reports NQ, we're
+	thoroughly screwed.  This condition should be treated
+	according to <xref linkend="excatHSMviolation"/>.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatATAbusErr">
+        <title>ATA bus error</title>
+
+	<para>
+	ATA bus error means that data corruption occurred during
+	transmission over ATA bus (SATA or PATA).  This type of errors
+	can be indicated by
+	</para>
+
+	<itemizedlist>
+
+	<listitem>
+	<para>
+	ICRC or ABRT error as described in <xref linkend="excatDevErr"/>.
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	Controller-specific error completion with error information
+	indicating transmission error.
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	On some controllers, command timeout.  In this case, there may
+	be a mechanism to determine that the timeout is due to
+	transmission error.
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	Unknown/random errors, timeouts and all sorts of weirdities.
+	</para>
+	</listitem>
+
+	</itemizedlist>
+
+	<para>
+	As described above, transmission errors can cause wide variety
+	of symptoms ranging from device ICRC error to random device
+	lockup, and, for many cases, there is no way to tell if an
+	error condition is due to transmission error or not;
+	therefore, it's necessary to employ some kind of heuristic
+	when dealing with errors and timeouts.  For example,
+	encountering repetitive ABRT errors for known supported
+	command is likely to indicate ATA bus error.
+	</para>
+
+	<para>
+	Once it's determined that ATA bus errors have possibly
+	occurred, lowering ATA bus transmission speed is one of
+	actions which may alleviate the problem.  See <xref
+	linkend="exrecReconf"/> for more information.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatPCIbusErr">
+        <title>PCI bus error</title>
+
+	<para>
+	Data corruption or other failures during transmission over PCI
+	(or other system bus).  For standard BMDMA, this is indicated
+	by Error bit in the BMDMA Status register.  This type of
+	errors must be logged as it indicates something is very wrong
+	with the system.  Resetting host controller is recommended.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatLateCompletion">
+        <title>Late completion</title>
+
+	<para>
+	This occurs when timeout occurs and the timeout handler finds
+	out that the timed out command has completed successfully or
+	with error.  This is usually caused by lost interrupts.  This
+	type of errors must be logged.  Resetting host controller is
+	recommended.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatUnknown">
+        <title>Unknown error (timeout)</title>
+
+	<para>
+	This is when timeout occurs and the command is still
+	processing or the host and device are in unknown state.  When
+	this occurs, HSM could be in any valid or invalid state.  To
+	bring the device to known state and make it forget about the
+	timed out command, resetting is necessary.  The timed out
+	command may be retried.
+	</para>
+
+	<para>
+	Timeouts can also be caused by transmission errors.  Refer to
+	<xref linkend="excatATAbusErr"/> for more details.
+	</para>
+
+     </sect2>
+
+     <sect2 id="excatHoplugPM">
+        <title>Hotplug and power management exceptions</title>
+
+	<para>
+	&lt;&lt;TODO: fill here&gt;&gt;
+	</para>
+
+     </sect2>
+
+  </sect1>
+
+  <sect1 id="exrec">
+     <title>EH recovery actions</title>
+
+     <para>
+     This section discusses several important recovery actions.
+     </para>
+
+     <sect2 id="exrecClr">
+        <title>Clearing error condition</title>
+
+	<para>
+	Many controllers require its error registers to be cleared by
+	error handler.  Different controllers may have different
+	requirements.
+	</para>
+
+	<para>
+	For SATA, it's strongly recommended to clear at least SError
+	register during error handling.
+	</para>
+     </sect2>
+
+     <sect2 id="exrecRst">
+        <title>Reset</title>
+
+	<para>
+	During EH, resetting is necessary in the following cases.
+	</para>
+
+	<itemizedlist>
+
+	<listitem>
+	<para>
+	HSM is in unknown or invalid state
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	HBA is in unknown or invalid state
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	EH needs to make HBA/device forget about in-flight commands
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	HBA/device behaves weirdly
+	</para>
+	</listitem>
+
+	</itemizedlist>
+
+	<para>
+	Resetting during EH might be a good idea regardless of error
+	condition to improve EH robustness.  Whether to reset both or
+	either one of HBA and device depends on situation but the
+	following scheme is recommended.
+	</para>
+
+	<itemizedlist>
+
+	<listitem>
+	<para>
+	When it's known that HBA is in ready state but ATA/ATAPI
+	device in in unknown state, reset only device.
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	If HBA is in unknown state, reset both HBA and device.
+	</para>
+	</listitem>
+
+	</itemizedlist>
+
+	<para>
+	HBA resetting is implementation specific.  For a controller
+	complying to taskfile/BMDMA PCI IDE, stopping active DMA
+	transaction may be sufficient iff BMDMA state is the only HBA
+	context.  But even mostly taskfile/BMDMA PCI IDE complying
+	controllers may have implementation specific requirements and
+	mechanism to reset themselves.  This must be addressed by
+	specific drivers.
+	</para>
+
+	<para>
+	OTOH, ATA/ATAPI standard describes in detail ways to reset
+	ATA/ATAPI devices.
+	</para>
+
+	<variablelist>
+
+	   <varlistentry><term>PATA hardware reset</term>
+	   <listitem>
+	   <para>
+	   This is hardware initiated device reset signalled with
+	   asserted PATA RESET- signal.  There is no standard way to
+	   initiate hardware reset from software although some
+	   hardware provides registers that allow driver to directly
+	   tweak the RESET- signal.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry><term>Software reset</term>
+	   <listitem>
+	   <para>
+	   This is achieved by turning CONTROL SRST bit on for at
+	   least 5us.  Both PATA and SATA support it but, in case of
+	   SATA, this may require controller-specific support as the
+	   second Register FIS to clear SRST should be transmitted
+	   while BSY bit is still set.  Note that on PATA, this resets
+	   both master and slave devices on a channel.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry><term>EXECUTE DEVICE DIAGNOSTIC command</term>
+	   <listitem>
+	   <para>
+	   Although ATA/ATAPI standard doesn't describe exactly, EDD
+	   implies some level of resetting, possibly similar level
+	   with software reset.  Host-side EDD protocol can be handled
+	   with normal command processing and most SATA controllers
+	   should be able to handle EDD's just like other commands.
+	   As in software reset, EDD affects both devices on a PATA
+	   bus.
+	   </para>
+	   <para>
+	   Although EDD does reset devices, this doesn't suit error
+	   handling as EDD cannot be issued while BSY is set and it's
+	   unclear how it will act when device is in unknown/weird
+	   state.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry><term>ATAPI DEVICE RESET command</term>
+	   <listitem>
+	   <para>
+	   This is very similar to software reset except that reset
+	   can be restricted to the selected device without affecting
+	   the other device sharing the cable.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	   <varlistentry><term>SATA phy reset</term>
+	   <listitem>
+	   <para>
+	   This is the preferred way of resetting a SATA device.  In
+	   effect, it's identical to PATA hardware reset.  Note that
+	   this can be done with the standard SCR Control register.
+	   As such, it's usually easier to implement than software
+	   reset.
+	   </para>
+	   </listitem>
+	   </varlistentry>
+
+	</variablelist>
+
+	<para>
+	One more thing to consider when resetting devices is that
+	resetting clears certain configuration parameters and they
+	need to be set to their previous or newly adjusted values
+	after reset.
+	</para>
+
+	<para>
+	Parameters affected are.
+	</para>
+
+	<itemizedlist>
+
+	<listitem>
+	<para>
+	CHS set up with INITIALIZE DEVICE PARAMETERS (seldomly used)
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	Parameters set with SET FEATURES including transfer mode setting
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	Block count set with SET MULTIPLE MODE
+	</para>
+	</listitem>
+
+	<listitem>
+	<para>
+	Other parameters (SET MAX, MEDIA LOCK...)
+	</para>
+	</listitem>
+
+	</itemizedlist>
+
+	<para>
+	ATA/ATAPI standard specifies that some parameters must be
+	maintained across hardware or software reset, but doesn't
+	strictly specify all of them.  Always reconfiguring needed
+	parameters after reset is required for robustness.  Note that
+	this also applies when resuming from deep sleep (power-off).
+	</para>
+
+	<para>
+	Also, ATA/ATAPI standard requires that IDENTIFY DEVICE /
+	IDENTIFY PACKET DEVICE is issued after any configuration
+	parameter is updated or a hardware reset and the result used
+	for further operation.  OS driver is required to implement
+	revalidation mechanism to support this.
+	</para>
+
+     </sect2>
+
+     <sect2 id="exrecReconf">
+        <title>Reconfigure transport</title>
+
+	<para>
+	For both PATA and SATA, a lot of corners are cut for cheap
+	connectors, cables or controllers and it's quite common to see
+	high transmission error rate.  This can be mitigated by
+	lowering transmission speed.
+	</para>
+
+	<para>
+	The following is a possible scheme Jeff Garzik suggested.
+	</para>
+
+	<blockquote>
+	<para>
+	If more than $N (3?) transmission errors happen in 15 minutes,
+	</para>	
+	<itemizedlist>
+	<listitem>
+	<para>
+	if SATA, decrease SATA PHY speed.  if speed cannot be decreased,
+	</para>
+	</listitem>
+	<listitem>
+	<para>
+	decrease UDMA xfer speed.  if at UDMA0, switch to PIO4,
+	</para>
+	</listitem>
+	<listitem>
+	<para>
+	decrease PIO xfer speed.  if at PIO3, complain, but continue
+	</para>
+	</listitem>
+	</itemizedlist>
+	</blockquote>
+
+     </sect2>
+
+  </sect1>
+
+  </chapter>
+
   <chapter id="PiixInt">
      <title>ata_piix Internals</title>
 !Idrivers/scsi/ata_piix.c
diff --git a/arch/arm/kernel/sys_arm.c b/arch/arm/kernel/sys_arm.c
index 42629ff..ea569ba 100644
--- a/arch/arm/kernel/sys_arm.c
+++ b/arch/arm/kernel/sys_arm.c
@@ -305,7 +305,7 @@
 		  "Ir" (THREAD_START_SP - sizeof(regs)),
 		  "r" (&regs),
 		  "Ir" (sizeof(regs))
-		: "r0", "r1", "r2", "r3", "ip", "memory");
+		: "r0", "r1", "r2", "r3", "ip", "lr", "memory");
 
  out:
 	return ret;
diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
index e7d22db..f6de76e 100644
--- a/arch/arm/kernel/traps.c
+++ b/arch/arm/kernel/traps.c
@@ -504,7 +504,7 @@
 
 		bad_access:
 		spin_unlock(&mm->page_table_lock);
-		/* simulate a read access fault */
+		/* simulate a write access fault */
 		do_DataAbort(addr, 15 + (1 << 11), regs);
 		return -1;
 	}
diff --git a/arch/arm/mach-imx/generic.c b/arch/arm/mach-imx/generic.c
index 41e5849..f8a742b 100644
--- a/arch/arm/mach-imx/generic.c
+++ b/arch/arm/mach-imx/generic.c
@@ -28,14 +28,15 @@
 #include <linux/module.h>
 #include <asm/arch/imxfb.h>
 #include <asm/hardware.h>
+#include <asm/arch/imx-regs.h>
 
 #include <asm/mach/map.h>
 
 void imx_gpio_mode(int gpio_mode)
 {
 	unsigned int pin = gpio_mode & GPIO_PIN_MASK;
-	unsigned int port = (gpio_mode & GPIO_PORT_MASK) >> 5;
-	unsigned int ocr = (gpio_mode & GPIO_OCR_MASK) >> 10;
+	unsigned int port = (gpio_mode & GPIO_PORT_MASK) >> GPIO_PORT_SHIFT;
+	unsigned int ocr = (gpio_mode & GPIO_OCR_MASK) >> GPIO_OCR_SHIFT;
 	unsigned int tmp;
 
 	/* Pullup enable */
@@ -57,7 +58,7 @@
 		GPR(port) &= ~(1<<pin);
 
 	/* use as gpio? */
-	if( ocr == 3 )
+	if(gpio_mode &  GPIO_GIUS)
 		GIUS(port) |= (1<<pin);
 	else
 		GIUS(port) &= ~(1<<pin);
@@ -72,20 +73,20 @@
 		tmp |= (ocr << (pin*2));
 		OCR1(port) = tmp;
 
-		if( gpio_mode &	GPIO_AOUT )
-			ICONFA1(port) &= ~( 3<<(pin*2));
-		if( gpio_mode &	GPIO_BOUT )
-			ICONFB1(port) &= ~( 3<<(pin*2));
+		ICONFA1(port) &= ~( 3<<(pin*2));
+		ICONFA1(port) |= ((gpio_mode >> GPIO_AOUT_SHIFT) & 3) << (pin * 2);
+		ICONFB1(port) &= ~( 3<<(pin*2));
+		ICONFB1(port) |= ((gpio_mode >> GPIO_BOUT_SHIFT) & 3) << (pin * 2);
 	} else {
 		tmp = OCR2(port);
 		tmp &= ~( 3<<((pin-16)*2));
 		tmp |= (ocr << ((pin-16)*2));
 		OCR2(port) = tmp;
 
-		if( gpio_mode &	GPIO_AOUT )
-			ICONFA2(port) &= ~( 3<<((pin-16)*2));
-		if( gpio_mode &	GPIO_BOUT )
-			ICONFB2(port) &= ~( 3<<((pin-16)*2));
+		ICONFA2(port) &= ~( 3<<((pin-16)*2));
+		ICONFA2(port) |= ((gpio_mode >> GPIO_AOUT_SHIFT) & 3) << ((pin-16) * 2);
+		ICONFB2(port) &= ~( 3<<((pin-16)*2));
+		ICONFB2(port) |= ((gpio_mode >> GPIO_BOUT_SHIFT) & 3) << ((pin-16) * 2);
 	}
 }
 
diff --git a/arch/arm/mach-imx/mx1ads.c b/arch/arm/mach-imx/mx1ads.c
index 5d25434..a7511dd 100644
--- a/arch/arm/mach-imx/mx1ads.c
+++ b/arch/arm/mach-imx/mx1ads.c
@@ -55,7 +55,7 @@
 mx1ads_init(void)
 {
 #ifdef CONFIG_LEDS
-	imx_gpio_mode(GPIO_PORTA | GPIO_OUT | GPIO_GPIO | 2);
+	imx_gpio_mode(GPIO_PORTA | GPIO_OUT | 2);
 #endif
 	platform_add_devices(devices, ARRAY_SIZE(devices));
 }
diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig
index db5e47d..c54e04c 100644
--- a/arch/arm/mm/Kconfig
+++ b/arch/arm/mm/Kconfig
@@ -370,21 +370,21 @@
 
 config CPU_ICACHE_DISABLE
 	bool "Disable I-Cache"
-	depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020
+	depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 || CPU_V6
 	help
 	  Say Y here to disable the processor instruction cache. Unless
 	  you have a reason not to or are unsure, say N.
 
 config CPU_DCACHE_DISABLE
 	bool "Disable D-Cache"
-	depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020
+	depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 || CPU_V6
 	help
 	  Say Y here to disable the processor data cache. Unless
 	  you have a reason not to or are unsure, say N.
 
 config CPU_DCACHE_WRITETHROUGH
 	bool "Force write through D-cache"
-	depends on (CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020) && !CPU_DCACHE_DISABLE
+	depends on (CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 || CPU_V6) && !CPU_DCACHE_DISABLE
 	default y if CPU_ARM925T
 	help
 	  Say Y here to use the data cache in writethrough mode. Unless you
@@ -399,7 +399,7 @@
 
 config CPU_BPREDICT_DISABLE
 	bool "Disable branch prediction"
-	depends on CPU_ARM1020
+	depends on CPU_ARM1020 || CPU_V6
 	help
 	  Say Y here to disable branch prediction.  If unsure, say N.
 
diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c
index 6dc726a..d0a5106 100644
--- a/arch/ia64/kernel/mca.c
+++ b/arch/ia64/kernel/mca.c
@@ -1016,6 +1016,11 @@
 
 			cmc_polling_enabled = 1;
 			spin_unlock(&cmc_history_lock);
+			/* If we're being hit with CMC interrupts, we won't
+			 * ever execute the schedule_work() below.  Need to
+			 * disable CMC interrupts on this processor now.
+			 */
+			ia64_mca_cmc_vector_disable(NULL);
 			schedule_work(&cmc_disable_work);
 
 			/*
diff --git a/arch/ppc/platforms/pmac_time.c b/arch/ppc/platforms/pmac_time.c
index 778ce4f..efb819f 100644
--- a/arch/ppc/platforms/pmac_time.c
+++ b/arch/ppc/platforms/pmac_time.c
@@ -195,7 +195,7 @@
 		;
 	dend = get_dec();
 
-	tb_ticks_per_jiffy = (dstart - dend) / (6 * (HZ/100));
+	tb_ticks_per_jiffy = (dstart - dend) / ((6 * HZ)/100);
 	tb_to_us = mulhwu_scale_factor(dstart - dend, 60000);
 
 	printk(KERN_INFO "via_calibrate_decr: ticks per jiffy = %u (%u ticks)\n",
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index aba0539..6537445 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -25,62 +25,6 @@
 
 menu "General machine setup"
 
-config VT
-	bool
-	select INPUT
-	default y
-	---help---
-	  If you say Y here, you will get support for terminal devices with
-	  display and keyboard devices. These are called "virtual" because you
-	  can run several virtual terminals (also called virtual consoles) on
-	  one physical terminal. This is rather useful, for example one
-	  virtual terminal can collect system messages and warnings, another
-	  one can be used for a text-mode user session, and a third could run
-	  an X session, all in parallel. Switching between virtual terminals
-	  is done with certain key combinations, usually Alt-<function key>.
-
-	  The setterm command ("man setterm") can be used to change the
-	  properties (such as colors or beeping) of a virtual terminal. The
-	  man page console_codes(4) ("man console_codes") contains the special
-	  character sequences that can be used to change those properties
-	  directly. The fonts used on virtual terminals can be changed with
-	  the setfont ("man setfont") command and the key bindings are defined
-	  with the loadkeys ("man loadkeys") command.
-
-	  You need at least one virtual terminal device in order to make use
-	  of your keyboard and monitor. Therefore, only people configuring an
-	  embedded system would want to say N here in order to save some
-	  memory; the only way to log into such a system is then via a serial
-	  or network connection.
-
-	  If unsure, say Y, or else you won't be able to do much with your new
-	  shiny Linux system :-)
-
-config VT_CONSOLE
-	bool
-	default y
-	---help---
-	  The system console is the device which receives all kernel messages
-	  and warnings and which allows logins in single user mode. If you
-	  answer Y here, a virtual terminal (the device used to interact with
-	  a physical terminal) can be used as system console. This is the most
-	  common mode of operations, so you should say Y here unless you want
-	  the kernel messages be output only to a serial port (in which case
-	  you should say Y to "Console on serial port", below).
-
-	  If you do say Y here, by default the currently visible virtual
-	  terminal (/dev/tty0) will be used as system console. You can change
-	  that with a kernel command line option such as "console=tty3" which
-	  would use the third virtual terminal as system console. (Try "man
-	  bootparam" or see the documentation of your boot loader (lilo or
-	  loadlin) about how to pass options to the kernel at boot time.)
-
-	  If unsure, say Y.
-
-config HW_CONSOLE
-	bool
-	default y
-
 config SMP
 	bool "Symmetric multi-processing support (does not work on sun4/sun4c)"
 	depends on BROKEN
diff --git a/arch/sparc/kernel/time.c b/arch/sparc/kernel/time.c
index bc015e9..279a626 100644
--- a/arch/sparc/kernel/time.c
+++ b/arch/sparc/kernel/time.c
@@ -457,7 +457,7 @@
 	sbus_time_init();
 }
 
-extern __inline__ unsigned long do_gettimeoffset(void)
+static inline unsigned long do_gettimeoffset(void)
 {
 	return (*master_l10_counter >> 10) & 0x1fffff;
 }
diff --git a/arch/sparc/mm/srmmu.c b/arch/sparc/mm/srmmu.c
index c89a803..c664b96 100644
--- a/arch/sparc/mm/srmmu.c
+++ b/arch/sparc/mm/srmmu.c
@@ -260,7 +260,7 @@
 { return __pte((pte_val(pte) & SRMMU_CHG_MASK) | pgprot_val(newprot)); }
 
 /* to find an entry in a top-level page table... */
-extern inline pgd_t *srmmu_pgd_offset(struct mm_struct * mm, unsigned long address)
+static inline pgd_t *srmmu_pgd_offset(struct mm_struct * mm, unsigned long address)
 { return mm->pgd + (address >> SRMMU_PGDIR_SHIFT); }
 
 /* Find an entry in the second-level page table.. */
diff --git a/arch/sparc64/kernel/entry.S b/arch/sparc64/kernel/entry.S
index 2879b10..f685035dbd 100644
--- a/arch/sparc64/kernel/entry.S
+++ b/arch/sparc64/kernel/entry.S
@@ -97,8 +97,8 @@
 	faddd		%f0, %f2, %f4
 	fmuld		%f0, %f2, %f6
 	ldxa		[%g3] ASI_DMMU, %g5
-cplus_fptrap_insn_1:
-	sethi		%hi(0), %g2
+	sethi		%hi(sparc64_kern_sec_context), %g2
+	ldx		[%g2 + %lo(sparc64_kern_sec_context)], %g2
 	stxa		%g2, [%g3] ASI_DMMU
 	membar		#Sync
 	add		%g6, TI_FPREGS + 0xc0, %g2
@@ -126,8 +126,8 @@
 	fzero		%f34
 	ldxa		[%g3] ASI_DMMU, %g5
 	add		%g6, TI_FPREGS, %g1
-cplus_fptrap_insn_2:
-	sethi		%hi(0), %g2
+	sethi		%hi(sparc64_kern_sec_context), %g2
+	ldx		[%g2 + %lo(sparc64_kern_sec_context)], %g2
 	stxa		%g2, [%g3] ASI_DMMU
 	membar		#Sync
 	add		%g6, TI_FPREGS + 0x40, %g2
@@ -153,8 +153,8 @@
 3:	mov		SECONDARY_CONTEXT, %g3
 	add		%g6, TI_FPREGS, %g1
 	ldxa		[%g3] ASI_DMMU, %g5
-cplus_fptrap_insn_3:
-	sethi		%hi(0), %g2
+	sethi		%hi(sparc64_kern_sec_context), %g2
+	ldx		[%g2 + %lo(sparc64_kern_sec_context)], %g2
 	stxa		%g2, [%g3] ASI_DMMU
 	membar		#Sync
 	mov		0x40, %g2
@@ -319,8 +319,8 @@
 	stx		%g3, [%g6 + TI_GSR]
 	mov		SECONDARY_CONTEXT, %g3
 	ldxa		[%g3] ASI_DMMU, %g5
-cplus_fptrap_insn_4:
-	sethi		%hi(0), %g2
+	sethi		%hi(sparc64_kern_sec_context), %g2
+	ldx		[%g2 + %lo(sparc64_kern_sec_context)], %g2
 	stxa		%g2, [%g3] ASI_DMMU
 	membar		#Sync
 	add		%g6, TI_FPREGS, %g2
@@ -341,33 +341,6 @@
 	ba,pt		%xcc, etrap
 	 wr		%g0, 0, %fprs
 
-cplus_fptrap_1:
-	sethi		%hi(CTX_CHEETAH_PLUS_CTX0), %g2
-
-	.globl		cheetah_plus_patch_fpdis
-cheetah_plus_patch_fpdis:
-	/* We configure the dTLB512_0 for 4MB pages and the
-	 * dTLB512_1 for 8K pages when in context zero.
-	 */
-	sethi			%hi(cplus_fptrap_1), %o0
-	lduw			[%o0 + %lo(cplus_fptrap_1)], %o1
-
-	set			cplus_fptrap_insn_1, %o2
-	stw			%o1, [%o2]
-	flush			%o2
-	set			cplus_fptrap_insn_2, %o2
-	stw			%o1, [%o2]
-	flush			%o2
-	set			cplus_fptrap_insn_3, %o2
-	stw			%o1, [%o2]
-	flush			%o2
-	set			cplus_fptrap_insn_4, %o2
-	stw			%o1, [%o2]
-	flush			%o2
-
-	retl
-	 nop
-
 	/* The registers for cross calls will be:
 	 *
 	 * DATA 0: [low 32-bits]  Address of function to call, jmp to this
diff --git a/arch/sparc64/kernel/etrap.S b/arch/sparc64/kernel/etrap.S
index 50d2af1..0d8eba2 100644
--- a/arch/sparc64/kernel/etrap.S
+++ b/arch/sparc64/kernel/etrap.S
@@ -68,12 +68,8 @@
 
 		wrpr	%g3, 0, %otherwin
 		wrpr	%g2, 0, %wstate
-cplus_etrap_insn_1:
-		sethi	%hi(0), %g3
-		sllx	%g3, 32, %g3
-cplus_etrap_insn_2:
-		sethi	%hi(0), %g2
-		or	%g3, %g2, %g3
+		sethi	%hi(sparc64_kern_pri_context), %g2
+		ldx	[%g2 + %lo(sparc64_kern_pri_context)], %g3
 		stxa	%g3, [%l4] ASI_DMMU
 		flush	%l6
 		wr	%g0, ASI_AIUS, %asi
@@ -215,12 +211,8 @@
 		mov	PRIMARY_CONTEXT, %l4
 		wrpr	%g3, 0, %otherwin
 		wrpr	%g2, 0, %wstate
-cplus_etrap_insn_3:
-		sethi	%hi(0), %g3
-		sllx	%g3, 32, %g3
-cplus_etrap_insn_4:
-		sethi	%hi(0), %g2
-		or	%g3, %g2, %g3
+		sethi	%hi(sparc64_kern_pri_context), %g2
+		ldx	[%g2 + %lo(sparc64_kern_pri_context)], %g3
 		stxa	%g3, [%l4] ASI_DMMU
 		flush	%l6
 
@@ -264,38 +256,3 @@
 
 #undef TASK_REGOFF
 #undef ETRAP_PSTATE1
-
-cplus_einsn_1:
-		sethi			%uhi(CTX_CHEETAH_PLUS_NUC), %g3
-cplus_einsn_2:
-		sethi			%hi(CTX_CHEETAH_PLUS_CTX0), %g2
-
-		.globl			cheetah_plus_patch_etrap
-cheetah_plus_patch_etrap:
-		/* We configure the dTLB512_0 for 4MB pages and the
-		 * dTLB512_1 for 8K pages when in context zero.
-		 */
-		sethi			%hi(cplus_einsn_1), %o0
-		sethi			%hi(cplus_etrap_insn_1), %o2
-		lduw			[%o0 + %lo(cplus_einsn_1)], %o1
-		or			%o2, %lo(cplus_etrap_insn_1), %o2
-		stw			%o1, [%o2]
-		flush			%o2
-		sethi			%hi(cplus_etrap_insn_3), %o2
-		or			%o2, %lo(cplus_etrap_insn_3), %o2
-		stw			%o1, [%o2]
-		flush			%o2
-
-		sethi			%hi(cplus_einsn_2), %o0
-		sethi			%hi(cplus_etrap_insn_2), %o2
-		lduw			[%o0 + %lo(cplus_einsn_2)], %o1
-		or			%o2, %lo(cplus_etrap_insn_2), %o2
-		stw			%o1, [%o2]
-		flush			%o2
-		sethi			%hi(cplus_etrap_insn_4), %o2
-		or			%o2, %lo(cplus_etrap_insn_4), %o2
-		stw			%o1, [%o2]
-		flush			%o2
-
-		retl
-		 nop
diff --git a/arch/sparc64/kernel/head.S b/arch/sparc64/kernel/head.S
index 89406f9..2434049 100644
--- a/arch/sparc64/kernel/head.S
+++ b/arch/sparc64/kernel/head.S
@@ -325,23 +325,7 @@
 1:	sethi	%hi(tlb_type), %g1
 	stw	%g2, [%g1 + %lo(tlb_type)]
 
-	BRANCH_IF_CHEETAH_PLUS_OR_FOLLOWON(g1,g7,1f)
-	ba,pt	%xcc, 2f
-	 nop
-
-1:	/* Patch context register writes to support nucleus page
-	 * size correctly.
-	 */
-	call	cheetah_plus_patch_etrap
-	 nop
-	call	cheetah_plus_patch_rtrap
-	 nop
-	call	cheetah_plus_patch_fpdis
-	 nop
-	call	cheetah_plus_patch_winfixup
-	 nop
-
-2:	/* Patch copy/page operations to cheetah optimized versions. */
+	/* Patch copy/page operations to cheetah optimized versions. */
 	call	cheetah_patch_copyops
 	 nop
 	call	cheetah_patch_copy_page
@@ -484,20 +468,13 @@
 	call	prom_set_trap_table
 	 sethi	%hi(sparc64_ttable_tl0), %o0
 
-	BRANCH_IF_CHEETAH_PLUS_OR_FOLLOWON(g2,g3,1f)
-	ba,pt	%xcc, 2f
-	 nop
-
-1:	/* Start using proper page size encodings in ctx register.  */
-	sethi	%uhi(CTX_CHEETAH_PLUS_NUC), %g3
+	/* Start using proper page size encodings in ctx register.  */
+	sethi	%hi(sparc64_kern_pri_context), %g3
+	ldx	[%g3 + %lo(sparc64_kern_pri_context)], %g2
 	mov	PRIMARY_CONTEXT, %g1
-	sllx	%g3, 32, %g3
-	sethi	%hi(CTX_CHEETAH_PLUS_CTX0), %g2
-	or	%g3, %g2, %g3
-	stxa	%g3, [%g1] ASI_DMMU
+	stxa	%g2, [%g1] ASI_DMMU
 	membar	#Sync
 
-2:
 	rdpr	%pstate, %o1
 	or	%o1, PSTATE_IE, %o1
 	wrpr	%o1, 0, %pstate
diff --git a/arch/sparc64/kernel/rtrap.S b/arch/sparc64/kernel/rtrap.S
index fafd227..ecfb42a 100644
--- a/arch/sparc64/kernel/rtrap.S
+++ b/arch/sparc64/kernel/rtrap.S
@@ -256,9 +256,8 @@
 		brnz,pn			%l3, kern_rtt
 		 mov			PRIMARY_CONTEXT, %l7
 		ldxa			[%l7 + %l7] ASI_DMMU, %l0
-cplus_rtrap_insn_1:
-		sethi			%hi(0), %l1
-		sllx			%l1, 32, %l1
+		sethi			%hi(sparc64_kern_pri_nuc_bits), %l1
+		ldx			[%l1 + %lo(sparc64_kern_pri_nuc_bits)], %l1
 		or			%l0, %l1, %l0
 		stxa			%l0, [%l7] ASI_DMMU
 		flush			%g6
@@ -345,21 +344,3 @@
 		wr			%g0, FPRS_DU, %fprs
 		ba,pt			%xcc, rt_continue
 		 stb			%l5, [%g6 + TI_FPDEPTH]
-
-cplus_rinsn_1:
-		sethi			%uhi(CTX_CHEETAH_PLUS_NUC), %l1
-
-		.globl			cheetah_plus_patch_rtrap
-cheetah_plus_patch_rtrap:
-		/* We configure the dTLB512_0 for 4MB pages and the
-		 * dTLB512_1 for 8K pages when in context zero.
-		 */
-		sethi			%hi(cplus_rinsn_1), %o0
-		sethi			%hi(cplus_rtrap_insn_1), %o2
-		lduw			[%o0 + %lo(cplus_rinsn_1)], %o1
-		or			%o2, %lo(cplus_rtrap_insn_1), %o2
-		stw			%o1, [%o2]
-		flush			%o2
-
-		retl
-		 nop
diff --git a/arch/sparc64/kernel/setup.c b/arch/sparc64/kernel/setup.c
index 4c9c8f2..c1f3423 100644
--- a/arch/sparc64/kernel/setup.c
+++ b/arch/sparc64/kernel/setup.c
@@ -187,17 +187,13 @@
 		}
 
 		if ((va >= KERNBASE) && (va < (KERNBASE + (4 * 1024 * 1024)))) {
-			unsigned long kernel_pctx = 0;
-
-			if (tlb_type == cheetah_plus)
-				kernel_pctx |= (CTX_CHEETAH_PLUS_NUC |
-						CTX_CHEETAH_PLUS_CTX0);
+			extern unsigned long sparc64_kern_pri_context;
 
 			/* Spitfire Errata #32 workaround */
 			__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
 					     "flush	%%g6"
 					     : /* No outputs */
-					     : "r" (kernel_pctx),
+					     : "r" (sparc64_kern_pri_context),
 					       "r" (PRIMARY_CONTEXT),
 					       "i" (ASI_DMMU));
 
diff --git a/arch/sparc64/kernel/trampoline.S b/arch/sparc64/kernel/trampoline.S
index 89f2fcf..9478551 100644
--- a/arch/sparc64/kernel/trampoline.S
+++ b/arch/sparc64/kernel/trampoline.S
@@ -336,20 +336,13 @@
 	call		init_irqwork_curcpu
 	 nop
 
-	BRANCH_IF_CHEETAH_PLUS_OR_FOLLOWON(g2,g3,1f)
-	ba,pt	%xcc, 2f
-	 nop
-
-1:	/* Start using proper page size encodings in ctx register.  */
-	sethi	%uhi(CTX_CHEETAH_PLUS_NUC), %g3
+	/* Start using proper page size encodings in ctx register.  */
+	sethi	%hi(sparc64_kern_pri_context), %g3
+	ldx	[%g3 + %lo(sparc64_kern_pri_context)], %g2
 	mov	PRIMARY_CONTEXT, %g1
-	sllx	%g3, 32, %g3
-	sethi	%hi(CTX_CHEETAH_PLUS_CTX0), %g2
-	or	%g3, %g2, %g3
-	stxa	%g3, [%g1] ASI_DMMU
+	stxa	%g2, [%g1] ASI_DMMU
 	membar	#Sync
 
-2:
 	rdpr		%pstate, %o1
 	or		%o1, PSTATE_IE, %o1
 	wrpr		%o1, 0, %pstate
diff --git a/arch/sparc64/kernel/winfixup.S b/arch/sparc64/kernel/winfixup.S
index 99c809a..3916092 100644
--- a/arch/sparc64/kernel/winfixup.S
+++ b/arch/sparc64/kernel/winfixup.S
@@ -16,23 +16,14 @@
 	.text
 
 set_pcontext:
-cplus_winfixup_insn_1:
-	sethi	%hi(0), %l1
+	sethi	%hi(sparc64_kern_pri_context), %l1
+	ldx	[%l1 + %lo(sparc64_kern_pri_context)], %l1
 	mov	PRIMARY_CONTEXT, %g1
-	sllx	%l1, 32, %l1
-cplus_winfixup_insn_2:
-	sethi	%hi(0), %g2
-	or	%l1, %g2, %l1
 	stxa	%l1, [%g1] ASI_DMMU
 	flush	%g6
 	retl
 	 nop
 
-cplus_wfinsn_1:
-	sethi	%uhi(CTX_CHEETAH_PLUS_NUC), %l1
-cplus_wfinsn_2:
-	sethi	%hi(CTX_CHEETAH_PLUS_CTX0), %g2
-
 	.align	32
 
 	/* Here are the rules, pay attention.
@@ -395,23 +386,3 @@
 	 add		%sp, PTREGS_OFF, %o0
 	ba,pt		%xcc, rtrap
 	 clr		%l6
-	
-
-	.globl		cheetah_plus_patch_winfixup
-cheetah_plus_patch_winfixup:
-	sethi			%hi(cplus_wfinsn_1), %o0
-	sethi			%hi(cplus_winfixup_insn_1), %o2
-	lduw			[%o0 + %lo(cplus_wfinsn_1)], %o1
-	or			%o2, %lo(cplus_winfixup_insn_1), %o2
-	stw			%o1, [%o2]
-	flush			%o2
-
-	sethi			%hi(cplus_wfinsn_2), %o0
-	sethi			%hi(cplus_winfixup_insn_2), %o2
-	lduw			[%o0 + %lo(cplus_wfinsn_2)], %o1
-	or			%o2, %lo(cplus_winfixup_insn_2), %o2
-	stw			%o1, [%o2]
-	flush			%o2
-
-	retl
-	 nop
diff --git a/arch/sparc64/mm/init.c b/arch/sparc64/mm/init.c
index 5db5052..0d2e967 100644
--- a/arch/sparc64/mm/init.c
+++ b/arch/sparc64/mm/init.c
@@ -133,6 +133,12 @@
 
 struct page *mem_map_zero __read_mostly;
 
+unsigned int sparc64_highest_unlocked_tlb_ent __read_mostly;
+
+unsigned long sparc64_kern_pri_context __read_mostly;
+unsigned long sparc64_kern_pri_nuc_bits __read_mostly;
+unsigned long sparc64_kern_sec_context __read_mostly;
+
 int bigkernel = 0;
 
 /* XXX Tune this... */
@@ -362,6 +368,7 @@
 	unsigned long data;
 };
 static struct linux_prom_translation prom_trans[512] __initdata;
+static unsigned int prom_trans_ents __initdata;
 
 extern unsigned long prom_boot_page;
 extern void prom_remap(unsigned long physpage, unsigned long virtpage, int mmu_ihandle);
@@ -375,57 +382,7 @@
 unsigned long prom_pmd_phys __read_mostly;
 unsigned int swapper_pgd_zero __read_mostly;
 
-/* Allocate power-of-2 aligned chunks from the end of the
- * kernel image.  Return physical address.
- */
-static inline unsigned long early_alloc_phys(unsigned long size)
-{
-	unsigned long base;
-
-	BUILD_BUG_ON(size & (size - 1));
-
-	kern_size = (kern_size + (size - 1)) & ~(size - 1);
-	base = kern_base + kern_size;
-	kern_size += size;
-
-	return base;
-}
-
-static inline unsigned long load_phys32(unsigned long pa)
-{
-	unsigned long val;
-
-	__asm__ __volatile__("lduwa	[%1] %2, %0"
-			     : "=&r" (val)
-			     : "r" (pa), "i" (ASI_PHYS_USE_EC));
-
-	return val;
-}
-
-static inline unsigned long load_phys64(unsigned long pa)
-{
-	unsigned long val;
-
-	__asm__ __volatile__("ldxa	[%1] %2, %0"
-			     : "=&r" (val)
-			     : "r" (pa), "i" (ASI_PHYS_USE_EC));
-
-	return val;
-}
-
-static inline void store_phys32(unsigned long pa, unsigned long val)
-{
-	__asm__ __volatile__("stwa	%0, [%1] %2"
-			     : /* no outputs */
-			     : "r" (val), "r" (pa), "i" (ASI_PHYS_USE_EC));
-}
-
-static inline void store_phys64(unsigned long pa, unsigned long val)
-{
-	__asm__ __volatile__("stxa	%0, [%1] %2"
-			     : /* no outputs */
-			     : "r" (val), "r" (pa), "i" (ASI_PHYS_USE_EC));
-}
+static pmd_t *prompmd __read_mostly;
 
 #define BASE_PAGE_SIZE 8192
 
@@ -435,34 +392,28 @@
  */
 unsigned long prom_virt_to_phys(unsigned long promva, int *error)
 {
-	unsigned long pmd_phys = (prom_pmd_phys +
-				  ((promva >> 23) & 0x7ff) * sizeof(pmd_t));
-	unsigned long pte_phys;
-	pmd_t pmd_ent;
-	pte_t pte_ent;
+	pmd_t *pmdp = prompmd + ((promva >> 23) & 0x7ff);
+	pte_t *ptep;
 	unsigned long base;
 
-	pmd_val(pmd_ent) = load_phys32(pmd_phys);
-	if (pmd_none(pmd_ent)) {
+	if (pmd_none(*pmdp)) {
 		if (error)
 			*error = 1;
 		return 0;
 	}
-
-	pte_phys = (unsigned long)pmd_val(pmd_ent) << 11UL;
-	pte_phys += ((promva >> 13) & 0x3ff) * sizeof(pte_t);
-	pte_val(pte_ent) = load_phys64(pte_phys);
-	if (!pte_present(pte_ent)) {
+	ptep = (pte_t *)__pmd_page(*pmdp) + ((promva >> 13) & 0x3ff);
+	if (!pte_present(*ptep)) {
 		if (error)
 			*error = 1;
 		return 0;
 	}
 	if (error) {
 		*error = 0;
-		return pte_val(pte_ent);
+		return pte_val(*ptep);
 	}
-	base = pte_val(pte_ent) & _PAGE_PADDR;
-	return (base + (promva & (BASE_PAGE_SIZE - 1)));
+	base = pte_val(*ptep) & _PAGE_PADDR;
+
+	return base + (promva & (BASE_PAGE_SIZE - 1));
 }
 
 /* The obp translations are saved based on 8k pagesize, since obp can
@@ -475,25 +426,20 @@
 	unsigned long vaddr;
 
 	for (vaddr = start; vaddr < end; vaddr += BASE_PAGE_SIZE) {
-		unsigned long val, pte_phys, pmd_phys;
-		pmd_t pmd_ent;
-		int i;
+		unsigned long val;
+		pmd_t *pmd;
+		pte_t *pte;
 
-		pmd_phys = (prom_pmd_phys +
-			    (((vaddr >> 23) & 0x7ff) * sizeof(pmd_t)));
-		pmd_val(pmd_ent) = load_phys32(pmd_phys);
-		if (pmd_none(pmd_ent)) {
-			pte_phys = early_alloc_phys(BASE_PAGE_SIZE);
-
-			for (i = 0; i < BASE_PAGE_SIZE / sizeof(pte_t); i++)
-				store_phys64(pte_phys+i*sizeof(pte_t),0);
-
-			pmd_val(pmd_ent) = pte_phys >> 11UL;
-			store_phys32(pmd_phys, pmd_val(pmd_ent));
+		pmd = prompmd + ((vaddr >> 23) & 0x7ff);
+		if (pmd_none(*pmd)) {
+			pte = __alloc_bootmem(BASE_PAGE_SIZE, BASE_PAGE_SIZE,
+					      PAGE_SIZE);
+			if (!pte)
+				prom_halt();
+			memset(pte, 0, BASE_PAGE_SIZE);
+			pmd_set(pmd, pte);
 		}
-
-		pte_phys = (unsigned long)pmd_val(pmd_ent) << 11UL;
-		pte_phys += (((vaddr >> 13) & 0x3ff) * sizeof(pte_t));
+		pte = (pte_t *) __pmd_page(*pmd) + ((vaddr >> 13) & 0x3ff);
 
 		val = data;
 
@@ -501,7 +447,8 @@
 		if (tlb_type == spitfire)
 			val &= ~0x0003fe0000000000UL;
 
-		store_phys64(pte_phys, val | _PAGE_MODIFIED);
+		set_pte_at(&init_mm, vaddr, pte,
+			   __pte(val | _PAGE_MODIFIED));
 
 		data += BASE_PAGE_SIZE;
 	}
@@ -514,13 +461,17 @@
 }
 
 #define OBP_PMD_SIZE 2048
-static void __init build_obp_pgtable(int prom_trans_ents)
+static void __init build_obp_pgtable(void)
 {
 	unsigned long i;
 
-	prom_pmd_phys = early_alloc_phys(OBP_PMD_SIZE);
-	for (i = 0; i < OBP_PMD_SIZE; i += 4)
-		store_phys32(prom_pmd_phys + i, 0);
+	prompmd = __alloc_bootmem(OBP_PMD_SIZE, OBP_PMD_SIZE, PAGE_SIZE);
+	if (!prompmd)
+		prom_halt();
+
+	memset(prompmd, 0, OBP_PMD_SIZE);
+
+	prom_pmd_phys = __pa(prompmd);
 
 	for (i = 0; i < prom_trans_ents; i++) {
 		unsigned long start, end;
@@ -540,7 +491,7 @@
 /* Read OBP translations property into 'prom_trans[]'.
  * Return the number of entries.
  */
-static int __init read_obp_translations(void)
+static void __init read_obp_translations(void)
 {
 	int n, node;
 
@@ -561,8 +512,10 @@
 		prom_printf("prom_mappings: Couldn't get property.\n");
 		prom_halt();
 	}
+
 	n = n / sizeof(struct linux_prom_translation);
-	return n;
+
+	prom_trans_ents = n;
 }
 
 static void __init remap_kernel(void)
@@ -582,28 +535,38 @@
 	prom_dtlb_load(tlb_ent, tte_data, tte_vaddr);
 	prom_itlb_load(tlb_ent, tte_data, tte_vaddr);
 	if (bigkernel) {
-		prom_dtlb_load(tlb_ent - 1,
+		tlb_ent -= 1;
+		prom_dtlb_load(tlb_ent,
 			       tte_data + 0x400000, 
 			       tte_vaddr + 0x400000);
-		prom_itlb_load(tlb_ent - 1,
+		prom_itlb_load(tlb_ent,
 			       tte_data + 0x400000, 
 			       tte_vaddr + 0x400000);
 	}
+	sparc64_highest_unlocked_tlb_ent = tlb_ent - 1;
+	if (tlb_type == cheetah_plus) {
+		sparc64_kern_pri_context = (CTX_CHEETAH_PLUS_CTX0 |
+					    CTX_CHEETAH_PLUS_NUC);
+		sparc64_kern_pri_nuc_bits = CTX_CHEETAH_PLUS_NUC;
+		sparc64_kern_sec_context = CTX_CHEETAH_PLUS_CTX0;
+	}
 }
 
-static void __init inherit_prom_mappings(void)
-{
-	int n;
 
-	n = read_obp_translations();
-	build_obp_pgtable(n);
+static void __init inherit_prom_mappings_pre(void)
+{
+	read_obp_translations();
 
 	/* Now fixup OBP's idea about where we really are mapped. */
 	prom_printf("Remapping the kernel... ");
 	remap_kernel();
 
 	prom_printf("done.\n");
+}
 
+static void __init inherit_prom_mappings_post(void)
+{
+	build_obp_pgtable();
 	register_prom_callbacks();
 }
 
@@ -788,8 +751,8 @@
 		}
 	}
 	if (tlb_type == spitfire) {
-		int high = SPITFIRE_HIGHEST_LOCKED_TLBENT - bigkernel;
-		for (i = 0; i < high; i++) {
+		int high = sparc64_highest_unlocked_tlb_ent;
+		for (i = 0; i <= high; i++) {
 			unsigned long data;
 
 			/* Spitfire Errata #32 workaround */
@@ -877,9 +840,9 @@
 			}
 		}
 	} else if (tlb_type == cheetah || tlb_type == cheetah_plus) {
-		int high = CHEETAH_HIGHEST_LOCKED_TLBENT - bigkernel;
+		int high = sparc64_highest_unlocked_tlb_ent;
 
-		for (i = 0; i < high; i++) {
+		for (i = 0; i <= high; i++) {
 			unsigned long data;
 
 			data = cheetah_get_ldtlb_data(i);
@@ -1556,8 +1519,7 @@
 	
 	swapper_pgd_zero = pgd_val(swapper_pg_dir[0]);
 	
-	/* Inherit non-locked OBP mappings. */
-	inherit_prom_mappings();
+	inherit_prom_mappings_pre();
 	
 	/* Ok, we can use our TLB miss and window trap handlers safely.
 	 * We need to do a quick peek here to see if we are on StarFire
@@ -1568,15 +1530,23 @@
 		extern void setup_tba(int);
 		setup_tba(this_is_starfire);
 	}
-
-	inherit_locked_prom_mappings(1);
-
 	__flush_tlb_all();
 
+	/* Everything from this point forward, until we are done with
+	 * inherit_prom_mappings_post(), must complete successfully
+	 * without calling into the firmware.  The firwmare page tables
+	 * have not been built, but we are running on the Linux kernel's
+	 * trap table.
+	 */
+
 	/* Setup bootmem... */
 	pages_avail = 0;
 	last_valid_pfn = end_pfn = bootmem_init(&pages_avail);
 
+	inherit_prom_mappings_post();
+
+	inherit_locked_prom_mappings(1);
+
 #ifdef CONFIG_DEBUG_PAGEALLOC
 	kernel_physical_mapping_init();
 #endif
diff --git a/arch/um/include/registers.h b/arch/um/include/registers.h
index 0a35e6d..4892e5f 100644
--- a/arch/um/include/registers.h
+++ b/arch/um/include/registers.h
@@ -15,16 +15,6 @@
 extern void restore_registers(int pid, union uml_pt_regs *regs);
 extern void init_registers(int pid);
 extern void get_safe_registers(unsigned long * regs);
+extern void get_thread_regs(union uml_pt_regs *uml_regs, void *buffer);
 
 #endif
-
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/sysdep-x86_64/ptrace.h b/arch/um/include/sysdep-x86_64/ptrace.h
index 331aa2d..8f06567 100644
--- a/arch/um/include/sysdep-x86_64/ptrace.h
+++ b/arch/um/include/sysdep-x86_64/ptrace.h
@@ -218,10 +218,6 @@
                 case RBP: UPT_RBP(regs) = __upt_val; break; \
                 case ORIG_RAX: UPT_ORIG_RAX(regs) = __upt_val; break; \
                 case CS: UPT_CS(regs) = __upt_val; break; \
-                case DS: UPT_DS(regs) = __upt_val; break; \
-                case ES: UPT_ES(regs) = __upt_val; break; \
-                case FS: UPT_FS(regs) = __upt_val; break; \
-                case GS: UPT_GS(regs) = __upt_val; break; \
                 case EFLAGS: UPT_EFLAGS(regs) = __upt_val; break; \
                 default :  \
                         panic("Bad register in UPT_SET : %d\n", reg);  \
diff --git a/arch/um/kernel/sysrq.c b/arch/um/kernel/sysrq.c
index f808500..b331e97 100644
--- a/arch/um/kernel/sysrq.c
+++ b/arch/um/kernel/sysrq.c
@@ -62,13 +62,7 @@
 
 	if (esp == NULL) {
 		if (task != current && task != NULL) {
-			/* XXX: Isn't this bogus? I.e. isn't this the
-			 * *userspace* stack of this task? If not so, use this
-			 * even when task == current (as in i386).
-			 */
 			esp = (unsigned long *) KSTK_ESP(task);
-			/* Which one? No actual difference - just coding style.*/
-			//esp = (unsigned long *) PT_REGS_IP(&task->thread.regs);
 		} else {
 			esp = (unsigned long *) &esp;
 		}
@@ -84,5 +78,5 @@
 	}
 
 	printk("Call Trace: \n");
-	show_trace(current, esp);
+	show_trace(task, esp);
 }
diff --git a/arch/um/os-Linux/sys-i386/registers.c b/arch/um/os-Linux/sys-i386/registers.c
index 3125d32..aee4812 100644
--- a/arch/um/os-Linux/sys-i386/registers.c
+++ b/arch/um/os-Linux/sys-i386/registers.c
@@ -5,6 +5,7 @@
 
 #include <errno.h>
 #include <string.h>
+#include <setjmp.h>
 #include "sysdep/ptrace_user.h"
 #include "sysdep/ptrace.h"
 #include "uml-config.h"
@@ -126,13 +127,11 @@
 	memcpy(regs, exec_regs, HOST_FRAME_SIZE * sizeof(unsigned long));
 }
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
+void get_thread_regs(union uml_pt_regs *uml_regs, void *buffer)
+{
+	struct __jmp_buf_tag *jmpbuf = buffer;
+
+	UPT_SET(uml_regs, EIP, jmpbuf->__jmpbuf[JB_PC]);
+	UPT_SET(uml_regs, UESP, jmpbuf->__jmpbuf[JB_SP]);
+	UPT_SET(uml_regs, EBP, jmpbuf->__jmpbuf[JB_BP]);
+}
diff --git a/arch/um/os-Linux/sys-x86_64/registers.c b/arch/um/os-Linux/sys-x86_64/registers.c
index 44438d1..4b638dfb 100644
--- a/arch/um/os-Linux/sys-x86_64/registers.c
+++ b/arch/um/os-Linux/sys-x86_64/registers.c
@@ -5,6 +5,7 @@
 
 #include <errno.h>
 #include <string.h>
+#include <setjmp.h>
 #include "ptrace_user.h"
 #include "uml-config.h"
 #include "skas_ptregs.h"
@@ -74,13 +75,11 @@
 	memcpy(regs, exec_regs, HOST_FRAME_SIZE * sizeof(unsigned long));
 }
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
+void get_thread_regs(union uml_pt_regs *uml_regs, void *buffer)
+{
+	struct __jmp_buf_tag *jmpbuf = buffer;
+
+	UPT_SET(uml_regs, RIP, jmpbuf->__jmpbuf[JB_PC]);
+	UPT_SET(uml_regs, RSP, jmpbuf->__jmpbuf[JB_RSP]);
+	UPT_SET(uml_regs, RBP, jmpbuf->__jmpbuf[JB_RBP]);
+}
diff --git a/arch/um/sys-i386/sysrq.c b/arch/um/sys-i386/sysrq.c
index e3706d1..d5244f0 100644
--- a/arch/um/sys-i386/sysrq.c
+++ b/arch/um/sys-i386/sysrq.c
@@ -88,9 +88,7 @@
 		task = current;
 
 	if (task != current) {
-		//ebp = (unsigned long) KSTK_EBP(task);
-		/* Which one? No actual difference - just coding style.*/
-		ebp = (unsigned long) PT_REGS_EBP(&task->thread.regs);
+		ebp = (unsigned long) KSTK_EBP(task);
 	} else {
 		asm ("movl %%ebp, %0" : "=r" (ebp) : );
 	}
@@ -99,15 +97,6 @@
 		((unsigned long)stack & (~(THREAD_SIZE - 1)));
 	print_context_stack(context, stack, ebp);
 
-	/*while (((long) stack & (THREAD_SIZE-1)) != 0) {
-		addr = *stack;
-		if (__kernel_text_address(addr)) {
-			printk("%08lx:	[<%08lx>]", (unsigned long) stack, addr);
-			print_symbol(" %s", addr);
-			printk("\n");
-		}
-		stack++;
-	}*/
 	printk("\n");
 }
 
diff --git a/arch/um/sys-i386/user-offsets.c b/arch/um/sys-i386/user-offsets.c
index 677fc26..26b6867 100644
--- a/arch/um/sys-i386/user-offsets.c
+++ b/arch/um/sys-i386/user-offsets.c
@@ -46,7 +46,7 @@
 	OFFSET(HOST_SC_FP_ST, _fpstate, _st);
 	OFFSET(HOST_SC_FXSR_ENV, _fpstate, _fxsr_env);
 
-	DEFINE_LONGS(HOST_FRAME_SIZE, FRAME_SIZE);
+	DEFINE(HOST_FRAME_SIZE, FRAME_SIZE);
 	DEFINE_LONGS(HOST_FP_SIZE, sizeof(struct user_i387_struct));
 	DEFINE_LONGS(HOST_XFP_SIZE, sizeof(struct user_fxsr_struct));
 
diff --git a/arch/x86_64/kernel/head.S b/arch/x86_64/kernel/head.S
index 4592bf2..b92e5f4 100644
--- a/arch/x86_64/kernel/head.S
+++ b/arch/x86_64/kernel/head.S
@@ -270,26 +270,26 @@
 .org 0x4000
 ENTRY(level2_ident_pgt)
 	/* 40MB for bootup. 	*/
-	.quad	0x0000000000000183
-	.quad	0x0000000000200183
-	.quad	0x0000000000400183
-	.quad	0x0000000000600183
-	.quad	0x0000000000800183
-	.quad	0x0000000000A00183
-	.quad	0x0000000000C00183
-	.quad	0x0000000000E00183
-	.quad	0x0000000001000183
-	.quad	0x0000000001200183
-	.quad	0x0000000001400183
-	.quad	0x0000000001600183
-	.quad	0x0000000001800183
-	.quad	0x0000000001A00183
-	.quad	0x0000000001C00183
-	.quad	0x0000000001E00183
-	.quad	0x0000000002000183
-	.quad	0x0000000002200183
-	.quad	0x0000000002400183
-	.quad	0x0000000002600183
+	.quad	0x0000000000000083
+	.quad	0x0000000000200083
+	.quad	0x0000000000400083
+	.quad	0x0000000000600083
+	.quad	0x0000000000800083
+	.quad	0x0000000000A00083
+	.quad	0x0000000000C00083
+	.quad	0x0000000000E00083
+	.quad	0x0000000001000083
+	.quad	0x0000000001200083
+	.quad	0x0000000001400083
+	.quad	0x0000000001600083
+	.quad	0x0000000001800083
+	.quad	0x0000000001A00083
+	.quad	0x0000000001C00083
+	.quad	0x0000000001E00083
+	.quad	0x0000000002000083
+	.quad	0x0000000002200083
+	.quad	0x0000000002400083
+	.quad	0x0000000002600083
 	/* Temporary mappings for the super early allocator in arch/x86_64/mm/init.c */
 	.globl temp_boot_pmds
 temp_boot_pmds:
diff --git a/drivers/atm/fore200e.c b/drivers/atm/fore200e.c
index 2bf723a..6f1a83c 100644
--- a/drivers/atm/fore200e.c
+++ b/drivers/atm/fore200e.c
@@ -178,14 +178,12 @@
 
 
 static void*
-fore200e_kmalloc(int size, int flags)
+fore200e_kmalloc(int size, unsigned int __nocast flags)
 {
-    void* chunk = kmalloc(size, flags);
+    void *chunk = kzalloc(size, flags);
 
-    if (chunk)
-	memset(chunk, 0x00, size);
-    else
-	printk(FORE200E "kmalloc() failed, requested size = %d, flags = 0x%x\n", size, flags);
+    if (!chunk)
+	printk(FORE200E "kmalloc() failed, requested size = %d, flags = 0x%x\n",			size, flags);
     
     return chunk;
 }
diff --git a/drivers/char/drm/drm_stub.c b/drivers/char/drm/drm_stub.c
index 95a976c..70458cb 100644
--- a/drivers/char/drm/drm_stub.c
+++ b/drivers/char/drm/drm_stub.c
@@ -47,7 +47,7 @@
 MODULE_PARM_DESC(debug, "Enable debug output");
 
 module_param_named(cards_limit, drm_cards_limit, int, 0444);
-module_param_named(debug, drm_debug, int, 0666);
+module_param_named(debug, drm_debug, int, 0600);
 
 drm_head_t **drm_heads;
 struct drm_sysfs_class *drm_class;
diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c
index bb0b3a8..1422285 100644
--- a/drivers/connector/connector.c
+++ b/drivers/connector/connector.c
@@ -69,7 +69,8 @@
  * a new message.
  *
  */
-int cn_netlink_send(struct cn_msg *msg, u32 __group, int gfp_mask)
+int cn_netlink_send(struct cn_msg *msg, u32 __group,
+		    unsigned int __nocast gfp_mask)
 {
 	struct cn_callback_entry *__cbq;
 	unsigned int size;
diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c
index ffbcd40..23a3f56 100644
--- a/drivers/infiniband/hw/mthca/mthca_main.c
+++ b/drivers/infiniband/hw/mthca/mthca_main.c
@@ -503,6 +503,25 @@
 	return err;
 }
 
+static void mthca_free_icms(struct mthca_dev *mdev)
+{
+	u8 status;
+
+	mthca_free_icm_table(mdev, mdev->mcg_table.table);
+	if (mdev->mthca_flags & MTHCA_FLAG_SRQ)
+		mthca_free_icm_table(mdev, mdev->srq_table.table);
+	mthca_free_icm_table(mdev, mdev->cq_table.table);
+	mthca_free_icm_table(mdev, mdev->qp_table.rdb_table);
+	mthca_free_icm_table(mdev, mdev->qp_table.eqp_table);
+	mthca_free_icm_table(mdev, mdev->qp_table.qp_table);
+	mthca_free_icm_table(mdev, mdev->mr_table.mpt_table);
+	mthca_free_icm_table(mdev, mdev->mr_table.mtt_table);
+	mthca_unmap_eq_icm(mdev);
+
+	mthca_UNMAP_ICM_AUX(mdev, &status);
+	mthca_free_icm(mdev, mdev->fw.arbel.aux_icm);
+}
+
 static int __devinit mthca_init_arbel(struct mthca_dev *mdev)
 {
 	struct mthca_dev_lim        dev_lim;
@@ -580,18 +599,7 @@
 	return 0;
 
 err_free_icm:
-	if (mdev->mthca_flags & MTHCA_FLAG_SRQ)
-		mthca_free_icm_table(mdev, mdev->srq_table.table);
-	mthca_free_icm_table(mdev, mdev->cq_table.table);
-	mthca_free_icm_table(mdev, mdev->qp_table.rdb_table);
-	mthca_free_icm_table(mdev, mdev->qp_table.eqp_table);
-	mthca_free_icm_table(mdev, mdev->qp_table.qp_table);
-	mthca_free_icm_table(mdev, mdev->mr_table.mpt_table);
-	mthca_free_icm_table(mdev, mdev->mr_table.mtt_table);
-	mthca_unmap_eq_icm(mdev);
-
-	mthca_UNMAP_ICM_AUX(mdev, &status);
-	mthca_free_icm(mdev, mdev->fw.arbel.aux_icm);
+	mthca_free_icms(mdev);
 
 err_stop_fw:
 	mthca_UNMAP_FA(mdev, &status);
@@ -611,18 +619,7 @@
 	mthca_CLOSE_HCA(mdev, 0, &status);
 
 	if (mthca_is_memfree(mdev)) {
-		if (mdev->mthca_flags & MTHCA_FLAG_SRQ)
-			mthca_free_icm_table(mdev, mdev->srq_table.table);
-		mthca_free_icm_table(mdev, mdev->cq_table.table);
-		mthca_free_icm_table(mdev, mdev->qp_table.rdb_table);
-		mthca_free_icm_table(mdev, mdev->qp_table.eqp_table);
-		mthca_free_icm_table(mdev, mdev->qp_table.qp_table);
-		mthca_free_icm_table(mdev, mdev->mr_table.mpt_table);
-		mthca_free_icm_table(mdev, mdev->mr_table.mtt_table);
-		mthca_unmap_eq_icm(mdev);
-
-		mthca_UNMAP_ICM_AUX(mdev, &status);
-		mthca_free_icm(mdev, mdev->fw.arbel.aux_icm);
+		mthca_free_icms(mdev);
 
 		mthca_UNMAP_FA(mdev, &status);
 		mthca_free_icm(mdev, mdev->fw.arbel.fw_icm);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 704f48e..6c5bf07 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -474,7 +474,7 @@
 	spin_unlock(&priv->lock);
 }
 
-static void path_lookup(struct sk_buff *skb, struct net_device *dev)
+static void ipoib_path_lookup(struct sk_buff *skb, struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(skb->dev);
 
@@ -569,7 +569,7 @@
 
 	if (skb->dst && skb->dst->neighbour) {
 		if (unlikely(!*to_ipoib_neigh(skb->dst->neighbour))) {
-			path_lookup(skb, dev);
+			ipoib_path_lookup(skb, dev);
 			goto out;
 		}
 
diff --git a/drivers/mfd/ucb1x00-core.c b/drivers/mfd/ucb1x00-core.c
index 10f6ce1..612564a 100644
--- a/drivers/mfd/ucb1x00-core.c
+++ b/drivers/mfd/ucb1x00-core.c
@@ -642,8 +642,6 @@
 module_init(ucb1x00_init);
 module_exit(ucb1x00_exit);
 
-EXPORT_SYMBOL(ucb1x00_class);
-
 EXPORT_SYMBOL(ucb1x00_io_set_dir);
 EXPORT_SYMBOL(ucb1x00_io_write);
 EXPORT_SYMBOL(ucb1x00_io_read);
diff --git a/drivers/mfd/ucb1x00.h b/drivers/mfd/ucb1x00.h
index 6b63264..9c9a647 100644
--- a/drivers/mfd/ucb1x00.h
+++ b/drivers/mfd/ucb1x00.h
@@ -106,8 +106,6 @@
 	void (*fn)(int, void *);
 };
 
-extern struct class ucb1x00_class;
-
 struct ucb1x00 {
 	spinlock_t		lock;
 	struct mcp		*mcp;
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 2a908c4..c748b0e 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -1655,7 +1655,7 @@
 
 config NET_POCKET
 	bool "Pocket and portable adapters"
-	depends on NET_ETHERNET && ISA
+	depends on NET_ETHERNET && PARPORT
 	---help---
 	  Cute little network (Ethernet) devices which attach to the parallel
 	  port ("pocket adapters"), commonly used with laptops. If you have
@@ -1679,7 +1679,7 @@
 
 config ATP
 	tristate "AT-LAN-TEC/RealTek pocket adapter support"
-	depends on NET_POCKET && ISA && X86
+	depends on NET_POCKET && PARPORT && X86
 	select CRC32
 	---help---
 	  This is a network (Ethernet) device which attaches to your parallel
@@ -1694,7 +1694,7 @@
 
 config DE600
 	tristate "D-Link DE600 pocket adapter support"
-	depends on NET_POCKET && ISA
+	depends on NET_POCKET && PARPORT
 	---help---
 	  This is a network (Ethernet) device which attaches to your parallel
 	  port. Read <file:Documentation/networking/DLINK.txt> as well as the
@@ -1709,7 +1709,7 @@
 
 config DE620
 	tristate "D-Link DE620 pocket adapter support"
-	depends on NET_POCKET && ISA
+	depends on NET_POCKET && PARPORT
 	---help---
 	  This is a network (Ethernet) device which attaches to your parallel
 	  port. Read <file:Documentation/networking/DLINK.txt> as well as the
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index bf81cd45..f0a5b77 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -487,6 +487,8 @@
  *	  * Added xmit_hash_policy_layer34()
  *	- Modified by Jay Vosburgh <fubar@us.ibm.com> to also support mode 4.
  *	  Set version to 2.6.3.
+ * 2005/09/26 - Jay Vosburgh <fubar@us.ibm.com>
+ *	- Removed backwards compatibility for old ifenslaves.  Version 2.6.4.
  */
 
 //#define BONDING_DEBUG 1
@@ -595,14 +597,7 @@
 static int bond_mode	= BOND_MODE_ROUNDROBIN;
 static int xmit_hashtype= BOND_XMIT_POLICY_LAYER2;
 static int lacp_fast	= 0;
-static int app_abi_ver	= 0;
-static int orig_app_abi_ver = -1; /* This is used to save the first ABI version
-				   * we receive from the application. Once set,
-				   * it won't be changed, and the module will
-				   * refuse to enslave/release interfaces if the
-				   * command comes from an application using
-				   * another ABI version.
-				   */
+
 struct bond_parm_tbl {
 	char *modename;
 	int mode;
@@ -1294,12 +1289,13 @@
 /*
  * Copy all the Multicast addresses from src to the bonding device dst
  */
-static int bond_mc_list_copy(struct dev_mc_list *mc_list, struct bonding *bond, int gpf_flag)
+static int bond_mc_list_copy(struct dev_mc_list *mc_list, struct bonding *bond,
+			     unsigned int __nocast gfp_flag)
 {
 	struct dev_mc_list *dmi, *new_dmi;
 
 	for (dmi = mc_list; dmi; dmi = dmi->next) {
-		new_dmi = kmalloc(sizeof(struct dev_mc_list), gpf_flag);
+		new_dmi = kmalloc(sizeof(struct dev_mc_list), gfp_flag);
 
 		if (!new_dmi) {
 			/* FIXME: Potential memory leak !!! */
@@ -1702,51 +1698,29 @@
 		}
 	}
 
-	if (app_abi_ver >= 1) {
-		/* The application is using an ABI, which requires the
-		 * slave interface to be closed.
-		 */
-		if ((slave_dev->flags & IFF_UP)) {
-			printk(KERN_ERR DRV_NAME
-			       ": Error: %s is up\n",
-			       slave_dev->name);
-			res = -EPERM;
-			goto err_undo_flags;
-		}
+	/*
+	 * Old ifenslave binaries are no longer supported.  These can
+	 * be identified with moderate accurary by the state of the slave:
+	 * the current ifenslave will set the interface down prior to
+	 * enslaving it; the old ifenslave will not.
+	 */
+	if ((slave_dev->flags & IFF_UP)) {
+		printk(KERN_ERR DRV_NAME ": %s is up. "
+		       "This may be due to an out of date ifenslave.\n",
+		       slave_dev->name);
+		res = -EPERM;
+		goto err_undo_flags;
+	}
 
-		if (slave_dev->set_mac_address == NULL) {
-			printk(KERN_ERR DRV_NAME
-			       ": Error: The slave device you specified does "
-			       "not support setting the MAC address.\n");
-			printk(KERN_ERR
-			       "Your kernel likely does not support slave "
-			       "devices.\n");
+	if (slave_dev->set_mac_address == NULL) {
+		printk(KERN_ERR DRV_NAME
+		       ": Error: The slave device you specified does "
+		       "not support setting the MAC address.\n");
+		printk(KERN_ERR
+		       "Your kernel likely does not support slave devices.\n");
 
-			res = -EOPNOTSUPP;
-			goto err_undo_flags;
-		}
-	} else {
-		/* The application is not using an ABI, which requires the
-		 * slave interface to be open.
-		 */
-		if (!(slave_dev->flags & IFF_UP)) {
-			printk(KERN_ERR DRV_NAME
-			       ": Error: %s is not running\n",
-			       slave_dev->name);
-			res = -EINVAL;
-			goto err_undo_flags;
-		}
-
-		if ((bond->params.mode == BOND_MODE_8023AD) ||
-		    (bond->params.mode == BOND_MODE_TLB)    ||
-		    (bond->params.mode == BOND_MODE_ALB)) {
-			printk(KERN_ERR DRV_NAME
-			       ": Error: to use %s mode, you must upgrade "
-			       "ifenslave.\n",
-			       bond_mode_name(bond->params.mode));
-			res = -EOPNOTSUPP;
-			goto err_undo_flags;
-		}
+		res = -EOPNOTSUPP;
+		goto err_undo_flags;
 	}
 
 	new_slave = kmalloc(sizeof(struct slave), GFP_KERNEL);
@@ -1762,41 +1736,36 @@
 	 */
 	new_slave->original_flags = slave_dev->flags;
 
-	if (app_abi_ver >= 1) {
-		/* save slave's original ("permanent") mac address for
-		 * modes that needs it, and for restoring it upon release,
-		 * and then set it to the master's address
-		 */
-		memcpy(new_slave->perm_hwaddr, slave_dev->dev_addr, ETH_ALEN);
+	/*
+	 * Save slave's original ("permanent") mac address for modes
+	 * that need it, and for restoring it upon release, and then
+	 * set it to the master's address
+	 */
+	memcpy(new_slave->perm_hwaddr, slave_dev->dev_addr, ETH_ALEN);
 
-		/* set slave to master's mac address
-		 * The application already set the master's
-		 * mac address to that of the first slave
-		 */
-		memcpy(addr.sa_data, bond_dev->dev_addr, bond_dev->addr_len);
-		addr.sa_family = slave_dev->type;
-		res = dev_set_mac_address(slave_dev, &addr);
-		if (res) {
-			dprintk("Error %d calling set_mac_address\n", res);
-			goto err_free;
-		}
+	/*
+	 * Set slave to master's mac address.  The application already
+	 * set the master's mac address to that of the first slave
+	 */
+	memcpy(addr.sa_data, bond_dev->dev_addr, bond_dev->addr_len);
+	addr.sa_family = slave_dev->type;
+	res = dev_set_mac_address(slave_dev, &addr);
+	if (res) {
+		dprintk("Error %d calling set_mac_address\n", res);
+		goto err_free;
+	}
 
-		/* open the slave since the application closed it */
-		res = dev_open(slave_dev);
-		if (res) {
-			dprintk("Openning slave %s failed\n", slave_dev->name);
-			goto err_restore_mac;
-		}
+	/* open the slave since the application closed it */
+	res = dev_open(slave_dev);
+	if (res) {
+		dprintk("Openning slave %s failed\n", slave_dev->name);
+		goto err_restore_mac;
 	}
 
 	res = netdev_set_master(slave_dev, bond_dev);
 	if (res) {
 		dprintk("Error %d calling netdev_set_master\n", res);
-		if (app_abi_ver < 1) {
-			goto err_free;
-		} else {
-			goto err_close;
-		}
+		goto err_close;
 	}
 
 	new_slave->dev = slave_dev;
@@ -1997,39 +1966,6 @@
 
 	write_unlock_bh(&bond->lock);
 
-	if (app_abi_ver < 1) {
-		/*
-		 * !!! This is to support old versions of ifenslave.
-		 * We can remove this in 2.5 because our ifenslave takes
-		 * care of this for us.
-		 * We check to see if the master has a mac address yet.
-		 * If not, we'll give it the mac address of our slave device.
-		 */
-		int ndx = 0;
-
-		for (ndx = 0; ndx < bond_dev->addr_len; ndx++) {
-			dprintk("Checking ndx=%d of bond_dev->dev_addr\n",
-				ndx);
-			if (bond_dev->dev_addr[ndx] != 0) {
-				dprintk("Found non-zero byte at ndx=%d\n",
-					ndx);
-				break;
-			}
-		}
-
-		if (ndx == bond_dev->addr_len) {
-			/*
-			 * We got all the way through the address and it was
-			 * all 0's.
-			 */
-			dprintk("%s doesn't have a MAC address yet.  \n",
-				bond_dev->name);
-			dprintk("Going to give assign it from %s.\n",
-				slave_dev->name);
-			bond_sethwaddr(bond_dev, slave_dev);
-		}
-	}
-
 	printk(KERN_INFO DRV_NAME
 	       ": %s: enslaving %s as a%s interface with a%s link.\n",
 	       bond_dev->name, slave_dev->name,
@@ -2227,12 +2163,10 @@
 	/* close slave before restoring its mac address */
 	dev_close(slave_dev);
 
-	if (app_abi_ver >= 1) {
-		/* restore original ("permanent") mac address */
-		memcpy(addr.sa_data, slave->perm_hwaddr, ETH_ALEN);
-		addr.sa_family = slave_dev->type;
-		dev_set_mac_address(slave_dev, &addr);
-	}
+	/* restore original ("permanent") mac address */
+	memcpy(addr.sa_data, slave->perm_hwaddr, ETH_ALEN);
+	addr.sa_family = slave_dev->type;
+	dev_set_mac_address(slave_dev, &addr);
 
 	/* restore the original state of the
 	 * IFF_NOARP flag that might have been
@@ -2320,12 +2254,10 @@
 		/* close slave before restoring its mac address */
 		dev_close(slave_dev);
 
-		if (app_abi_ver >= 1) {
-			/* restore original ("permanent") mac address*/
-			memcpy(addr.sa_data, slave->perm_hwaddr, ETH_ALEN);
-			addr.sa_family = slave_dev->type;
-			dev_set_mac_address(slave_dev, &addr);
-		}
+		/* restore original ("permanent") mac address*/
+		memcpy(addr.sa_data, slave->perm_hwaddr, ETH_ALEN);
+		addr.sa_family = slave_dev->type;
+		dev_set_mac_address(slave_dev, &addr);
 
 		/* restore the original state of the IFF_NOARP flag that might have
 		 * been set by bond_set_slave_inactive_flags()
@@ -2423,57 +2355,6 @@
 	return res;
 }
 
-static int bond_ethtool_ioctl(struct net_device *bond_dev, struct ifreq *ifr)
-{
-	struct ethtool_drvinfo info;
-	void __user *addr = ifr->ifr_data;
-	uint32_t cmd;
-
-	if (get_user(cmd, (uint32_t __user *)addr)) {
-		return -EFAULT;
-	}
-
-	switch (cmd) {
-	case ETHTOOL_GDRVINFO:
-		if (copy_from_user(&info, addr, sizeof(info))) {
-			return -EFAULT;
-		}
-
-		if (strcmp(info.driver, "ifenslave") == 0) {
-			int new_abi_ver;
-			char *endptr;
-
-			new_abi_ver = simple_strtoul(info.fw_version,
-						     &endptr, 0);
-			if (*endptr) {
-				printk(KERN_ERR DRV_NAME
-				       ": Error: got invalid ABI "
-				       "version from application\n");
-
-				return -EINVAL;
-			}
-
-			if (orig_app_abi_ver == -1) {
-				orig_app_abi_ver  = new_abi_ver;
-			}
-
-			app_abi_ver = new_abi_ver;
-		}
-
-		strncpy(info.driver,  DRV_NAME, 32);
-		strncpy(info.version, DRV_VERSION, 32);
-		snprintf(info.fw_version, 32, "%d", BOND_ABI_VERSION);
-
-		if (copy_to_user(addr, &info, sizeof(info))) {
-			return -EFAULT;
-		}
-
-		return 0;
-	default:
-		return -EOPNOTSUPP;
-	}
-}
-
 static int bond_info_query(struct net_device *bond_dev, struct ifbond *info)
 {
 	struct bonding *bond = bond_dev->priv;
@@ -3442,16 +3323,11 @@
 	seq_printf(seq, "Link Failure Count: %d\n",
 		   slave->link_failure_count);
 
-	if (app_abi_ver >= 1) {
-		seq_printf(seq,
-			   "Permanent HW addr: %02x:%02x:%02x:%02x:%02x:%02x\n",
-			   slave->perm_hwaddr[0],
-			   slave->perm_hwaddr[1],
-			   slave->perm_hwaddr[2],
-			   slave->perm_hwaddr[3],
-			   slave->perm_hwaddr[4],
-			   slave->perm_hwaddr[5]);
-	}
+	seq_printf(seq,
+		   "Permanent HW addr: %02x:%02x:%02x:%02x:%02x:%02x\n",
+		   slave->perm_hwaddr[0], slave->perm_hwaddr[1],
+		   slave->perm_hwaddr[2], slave->perm_hwaddr[3],
+		   slave->perm_hwaddr[4], slave->perm_hwaddr[5]);
 
 	if (bond->params.mode == BOND_MODE_8023AD) {
 		const struct aggregator *agg
@@ -4010,15 +3886,12 @@
 	struct ifslave k_sinfo;
 	struct ifslave __user *u_sinfo = NULL;
 	struct mii_ioctl_data *mii = NULL;
-	int prev_abi_ver = orig_app_abi_ver;
 	int res = 0;
 
 	dprintk("bond_ioctl: master=%s, cmd=%d\n",
 		bond_dev->name, cmd);
 
 	switch (cmd) {
-	case SIOCETHTOOL:
-		return bond_ethtool_ioctl(bond_dev, ifr);
 	case SIOCGMIIPHY:
 		mii = if_mii(ifr);
 		if (!mii) {
@@ -4090,21 +3963,6 @@
 		return -EPERM;
 	}
 
-	if (orig_app_abi_ver == -1) {
-		/* no orig_app_abi_ver was provided yet, so we'll use the
-		 * current one from now on, even if it's 0
-		 */
-		orig_app_abi_ver = app_abi_ver;
-
-	} else if (orig_app_abi_ver != app_abi_ver) {
-		printk(KERN_ERR DRV_NAME
-		       ": Error: already using ifenslave ABI version %d; to "
-		       "upgrade ifenslave to version %d, you must first "
-		       "reload bonding.\n",
-		       orig_app_abi_ver, app_abi_ver);
-		return -EINVAL;
-	}
-
 	slave_dev = dev_get_by_name(ifr->ifr_slave);
 
 	dprintk("slave_dev=%p: \n", slave_dev);
@@ -4137,14 +3995,6 @@
 		dev_put(slave_dev);
 	}
 
-	if (res < 0) {
-		/* The ioctl failed, so there's no point in changing the
-		 * orig_app_abi_ver. We'll restore it's value just in case
-		 * we've changed it earlier in this function.
-		 */
-		orig_app_abi_ver = prev_abi_ver;
-	}
-
 	return res;
 }
 
@@ -4578,9 +4428,18 @@
 	}
 }
 
+static void bond_ethtool_get_drvinfo(struct net_device *bond_dev,
+				    struct ethtool_drvinfo *drvinfo)
+{
+	strncpy(drvinfo->driver, DRV_NAME, 32);
+	strncpy(drvinfo->version, DRV_VERSION, 32);
+	snprintf(drvinfo->fw_version, 32, "%d", BOND_ABI_VERSION);
+}
+
 static struct ethtool_ops bond_ethtool_ops = {
 	.get_tx_csum		= ethtool_op_get_tx_csum,
 	.get_sg			= ethtool_op_get_sg,
+	.get_drvinfo		= bond_ethtool_get_drvinfo,
 };
 
 /*
diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h
index 3881969..bbf9da8 100644
--- a/drivers/net/bonding/bonding.h
+++ b/drivers/net/bonding/bonding.h
@@ -40,8 +40,8 @@
 #include "bond_3ad.h"
 #include "bond_alb.h"
 
-#define DRV_VERSION	"2.6.3"
-#define DRV_RELDATE	"June 8, 2005"
+#define DRV_VERSION	"2.6.4"
+#define DRV_RELDATE	"September 26, 2005"
 #define DRV_NAME	"bonding"
 #define DRV_DESCRIPTION	"Ethernet Channel Bonding Driver"
 
diff --git a/drivers/net/ibm_emac/ibm_emac_core.c b/drivers/net/ibm_emac/ibm_emac_core.c
index 0de3bb9..14e9b631 100644
--- a/drivers/net/ibm_emac/ibm_emac_core.c
+++ b/drivers/net/ibm_emac/ibm_emac_core.c
@@ -1875,6 +1875,9 @@
 		rc = -ENODEV;
 		goto bail;
 	}
+	
+	/* Disable any PHY features not supported by the platform */
+	ep->phy_mii.def->features &= ~emacdata->phy_feat_exc;
 
 	/* Setup initial PHY config & startup aneg */
 	if (ep->phy_mii.def->ops->init)
@@ -1882,6 +1885,34 @@
 	netif_carrier_off(ndev);
 	if (ep->phy_mii.def->features & SUPPORTED_Autoneg)
 		ep->want_autoneg = 1;
+	else {
+		ep->want_autoneg = 0;
+		
+		/* Select highest supported speed/duplex */
+		if (ep->phy_mii.def->features & SUPPORTED_1000baseT_Full) {
+			ep->phy_mii.speed = SPEED_1000;
+			ep->phy_mii.duplex = DUPLEX_FULL;
+		} else if (ep->phy_mii.def->features & 
+			   SUPPORTED_1000baseT_Half) {
+			ep->phy_mii.speed = SPEED_1000;
+			ep->phy_mii.duplex = DUPLEX_HALF;
+		} else if (ep->phy_mii.def->features & 
+			   SUPPORTED_100baseT_Full) {
+			ep->phy_mii.speed = SPEED_100;
+			ep->phy_mii.duplex = DUPLEX_FULL;
+		} else if (ep->phy_mii.def->features & 
+			   SUPPORTED_100baseT_Half) {
+			ep->phy_mii.speed = SPEED_100;
+			ep->phy_mii.duplex = DUPLEX_HALF;
+		} else if (ep->phy_mii.def->features & 
+			   SUPPORTED_10baseT_Full) {
+			ep->phy_mii.speed = SPEED_10;
+			ep->phy_mii.duplex = DUPLEX_FULL;
+		} else {
+			ep->phy_mii.speed = SPEED_10;
+			ep->phy_mii.duplex = DUPLEX_HALF;
+		}
+	}
 	emac_start_link(ep, NULL);
 
 	/* read the MAC Address */
diff --git a/drivers/net/ns83820.c b/drivers/net/ns83820.c
index e64df4d..83334db 100644
--- a/drivers/net/ns83820.c
+++ b/drivers/net/ns83820.c
@@ -584,7 +584,7 @@
 	return 0;
 }
 
-static inline int rx_refill(struct net_device *ndev, int gfp)
+static inline int rx_refill(struct net_device *ndev, unsigned int __nocast gfp)
 {
 	struct ns83820 *dev = PRIV(ndev);
 	unsigned i;
diff --git a/drivers/net/pcmcia/smc91c92_cs.c b/drivers/net/pcmcia/smc91c92_cs.c
index d652e1ed..c7cca84 100644
--- a/drivers/net/pcmcia/smc91c92_cs.c
+++ b/drivers/net/pcmcia/smc91c92_cs.c
@@ -1832,7 +1832,7 @@
 {
     struct dev_mc_list	*mc_addr;
 
-    for (mc_addr = addrs;  mc_addr && --count > 0;  mc_addr = mc_addr->next) {
+    for (mc_addr = addrs;  mc_addr && count-- > 0;  mc_addr = mc_addr->next) {
 	u_int position = ether_crc(6, mc_addr->dmi_addr);
 #ifndef final_version		/* Verify multicast address. */
 	if ((mc_addr->dmi_addr[0] & 1) == 0)
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
index fd398da..c2e6484 100644
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -2837,21 +2837,29 @@
 static int skge_set_mac_address(struct net_device *dev, void *p)
 {
 	struct skge_port *skge = netdev_priv(dev);
-	struct sockaddr *addr = p;
-	int err = 0;
+	struct skge_hw *hw = skge->hw;
+	unsigned port = skge->port;
+	const struct sockaddr *addr = p;
 
 	if (!is_valid_ether_addr(addr->sa_data))
 		return -EADDRNOTAVAIL;
 
-	skge_down(dev);
+	spin_lock_bh(&hw->phy_lock);
 	memcpy(dev->dev_addr, addr->sa_data, ETH_ALEN);
-	memcpy_toio(skge->hw->regs + B2_MAC_1 + skge->port*8,
+	memcpy_toio(hw->regs + B2_MAC_1 + port*8,
 		    dev->dev_addr, ETH_ALEN);
-	memcpy_toio(skge->hw->regs + B2_MAC_2 + skge->port*8,
+	memcpy_toio(hw->regs + B2_MAC_2 + port*8,
 		    dev->dev_addr, ETH_ALEN);
-	if (dev->flags & IFF_UP)
-		err = skge_up(dev);
-	return err;
+
+	if (hw->chip_id == CHIP_ID_GENESIS)
+		xm_outaddr(hw, port, XM_SA, dev->dev_addr);
+	else {
+		gma_set_addr(hw, port, GM_SRC_ADDR_1L, dev->dev_addr);
+		gma_set_addr(hw, port, GM_SRC_ADDR_2L, dev->dev_addr);
+	}
+	spin_unlock_bh(&hw->phy_lock);
+
+	return 0;
 }
 
 static const struct {
diff --git a/drivers/net/starfire.c b/drivers/net/starfire.c
index 88b89dc..efdb179 100644
--- a/drivers/net/starfire.c
+++ b/drivers/net/starfire.c
@@ -133,14 +133,18 @@
 	- finally added firmware (GPL'ed by Adaptec)
 	- removed compatibility code for 2.2.x
 
+	LK1.4.2.1 (Ion Badulescu)
+	- fixed 32/64 bit issues on i386 + CONFIG_HIGHMEM
+	- added 32-bit padding to outgoing skb's, removed previous workaround
+
 TODO:	- fix forced speed/duplexing code (broken a long time ago, when
 	somebody converted the driver to use the generic MII code)
 	- fix VLAN support
 */
 
 #define DRV_NAME	"starfire"
-#define DRV_VERSION	"1.03+LK1.4.2"
-#define DRV_RELDATE	"January 19, 2005"
+#define DRV_VERSION	"1.03+LK1.4.2.1"
+#define DRV_RELDATE	"October 3, 2005"
 
 #include <linux/config.h>
 #include <linux/version.h>
@@ -165,6 +169,14 @@
  * of length 1. If and when this is fixed, the #define below can be removed.
  */
 #define HAS_BROKEN_FIRMWARE
+
+/*
+ * If using the broken firmware, data must be padded to the next 32-bit boundary.
+ */
+#ifdef HAS_BROKEN_FIRMWARE
+#define PADDING_MASK 3
+#endif
+
 /*
  * Define this if using the driver with the zero-copy patch
  */
@@ -257,9 +269,10 @@
  * This SUCKS.
  * We need a much better method to determine if dma_addr_t is 64-bit.
  */
-#if (defined(__i386__) && defined(CONFIG_HIGHMEM) && (LINUX_VERSION_CODE > 0x20500 || defined(CONFIG_HIGHMEM64G))) || defined(__x86_64__) || defined (__ia64__) || defined(__mips64__) || (defined(__mips__) && defined(CONFIG_HIGHMEM) && defined(CONFIG_64BIT_PHYS_ADDR))
+#if (defined(__i386__) && defined(CONFIG_HIGHMEM64G)) || defined(__x86_64__) || defined (__ia64__) || defined(__mips64__) || (defined(__mips__) && defined(CONFIG_HIGHMEM) && defined(CONFIG_64BIT_PHYS_ADDR))
 /* 64-bit dma_addr_t */
 #define ADDR_64BITS	/* This chip uses 64 bit addresses. */
+#define netdrv_addr_t u64
 #define cpu_to_dma(x) cpu_to_le64(x)
 #define dma_to_cpu(x) le64_to_cpu(x)
 #define RX_DESC_Q_ADDR_SIZE RxDescQAddr64bit
@@ -268,6 +281,7 @@
 #define TX_COMPL_Q_ADDR_SIZE TxComplQAddr64bit
 #define RX_DESC_ADDR_SIZE RxDescAddr64bit
 #else  /* 32-bit dma_addr_t */
+#define netdrv_addr_t u32
 #define cpu_to_dma(x) cpu_to_le32(x)
 #define dma_to_cpu(x) le32_to_cpu(x)
 #define RX_DESC_Q_ADDR_SIZE RxDescQAddr32bit
@@ -1333,21 +1347,10 @@
 	}
 
 #if defined(ZEROCOPY) && defined(HAS_BROKEN_FIRMWARE)
-	{
-		int has_bad_length = 0;
-
-		if (skb_first_frag_len(skb) == 1)
-			has_bad_length = 1;
-		else {
-			for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
-				if (skb_shinfo(skb)->frags[i].size == 1) {
-					has_bad_length = 1;
-					break;
-				}
-		}
-
-		if (has_bad_length)
-			skb_checksum_help(skb, 0);
+	if (skb->ip_summed == CHECKSUM_HW) {
+		skb = skb_padto(skb, (skb->len + PADDING_MASK) & ~PADDING_MASK);
+		if (skb == NULL)
+			return NETDEV_TX_OK;
 	}
 #endif /* ZEROCOPY && HAS_BROKEN_FIRMWARE */
 
@@ -2127,13 +2130,12 @@
 #endif
 #endif
 
-#ifndef ADDR_64BITS
 	/* we can do this test only at run-time... sigh */
-	if (sizeof(dma_addr_t) == sizeof(u64)) {
-		printk("This driver has not been ported to this 64-bit architecture yet\n");
+	if (sizeof(dma_addr_t) != sizeof(netdrv_addr_t)) {
+		printk("This driver has dma_addr_t issues, please send email to maintainer\n");
 		return -ENODEV;
 	}
-#endif /* not ADDR_64BITS */
+
 	return pci_module_init (&starfire_driver);
 }
 
diff --git a/drivers/net/sungem.h b/drivers/net/sungem.h
index ff8ae5f..16edbb1 100644
--- a/drivers/net/sungem.h
+++ b/drivers/net/sungem.h
@@ -1035,7 +1035,8 @@
 			
 #define ALIGNED_RX_SKB_ADDR(addr) \
         ((((unsigned long)(addr) + (64UL - 1UL)) & ~(64UL - 1UL)) - (unsigned long)(addr))
-static __inline__ struct sk_buff *gem_alloc_skb(int size, int gfp_flags)
+static __inline__ struct sk_buff *gem_alloc_skb(int size,
+						unsigned int __nocast gfp_flags)
 {
 	struct sk_buff *skb = alloc_skb(size + 64, gfp_flags);
 
diff --git a/drivers/net/tokenring/ibmtr.c b/drivers/net/tokenring/ibmtr.c
index e7b0010..32057e6 100644
--- a/drivers/net/tokenring/ibmtr.c
+++ b/drivers/net/tokenring/ibmtr.c
@@ -531,7 +531,6 @@
 			if (!time_after(jiffies, timeout)) continue;
 			DPRINTK( "Hardware timeout during initialization.\n");
 			iounmap(t_mmio);
-			kfree(ti);
 			return -ENODEV;
 		}
 		ti->sram_phys =
@@ -645,7 +644,6 @@
 			DPRINTK("Unknown shared ram paging info %01X\n",
 							ti->shared_ram_paging);
 			iounmap(t_mmio); 
-			kfree(ti);
 			return -ENODEV;
 			break;
 		} /*end switch shared_ram_paging */
@@ -675,7 +673,6 @@
 			"driver limit (%05x), adapter not started.\n",
 			chk_base, ibmtr_mem_base + IBMTR_SHARED_RAM_SIZE);
 			iounmap(t_mmio);
-			kfree(ti);
 			return -ENODEV;
 		} else { /* seems cool, record what we have figured out */
 			ti->sram_base = new_base >> 12;
@@ -690,7 +687,6 @@
 		DPRINTK("Could not grab irq %d.  Halting Token Ring driver.\n",
 					irq);
 		iounmap(t_mmio);
-		kfree(ti);
 		return -ENODEV;
 	}
 	/*?? Now, allocate some of the PIO PORTs for this driver.. */
@@ -699,7 +695,6 @@
 		DPRINTK("Could not grab PIO range. Halting driver.\n");
 		free_irq(dev->irq, dev);
 		iounmap(t_mmio);
-		kfree(ti);
 		return -EBUSY;
 	}
 
diff --git a/drivers/net/tulip/21142.c b/drivers/net/tulip/21142.c
index 5db694c..683f14b 100644
--- a/drivers/net/tulip/21142.c
+++ b/drivers/net/tulip/21142.c
@@ -172,7 +172,7 @@
 			int i;
 			for (i = 0; i < tp->mtable->leafcount; i++)
 				if (tp->mtable->mleaf[i].media == dev->if_port) {
-					int startup = ! ((tp->chip_id == DC21143 && tp->revision == 65));
+					int startup = ! ((tp->chip_id == DC21143 && (tp->revision == 48 || tp->revision == 65)));
 					tp->cur_index = i;
 					tulip_select_media(dev, startup);
 					setup_done = 1;
diff --git a/drivers/net/wireless/orinoco.c b/drivers/net/wireless/orinoco.c
index 6deb7cc..cf3daaa 100644
--- a/drivers/net/wireless/orinoco.c
+++ b/drivers/net/wireless/orinoco.c
@@ -503,9 +503,14 @@
 		return 0;
 	}
 
-	/* Length of the packet body */
-	/* FIXME: what if the skb is smaller than this? */
-	len = max_t(int,skb->len - ETH_HLEN, ETH_ZLEN - ETH_HLEN);
+	/* Check packet length, pad short packets, round up odd length */
+	len = max_t(int, ALIGN(skb->len, 2), ETH_ZLEN);
+	if (skb->len < len) {
+		skb = skb_padto(skb, len);
+		if (skb == NULL)
+			goto fail;
+	}
+	len -= ETH_HLEN;
 
 	eh = (struct ethhdr *)skb->data;
 
@@ -557,8 +562,7 @@
 		p = skb->data;
 	}
 
-	/* Round up for odd length packets */
-	err = hermes_bap_pwrite(hw, USER_BAP, p, ALIGN(data_len, 2),
+	err = hermes_bap_pwrite(hw, USER_BAP, p, data_len,
 				txfid, data_off);
 	if (err) {
 		printk(KERN_ERR "%s: Error %d writing packet to BAP\n",
diff --git a/drivers/s390/net/qeth.h b/drivers/s390/net/qeth.h
index 2ad4797..9963479 100644
--- a/drivers/s390/net/qeth.h
+++ b/drivers/s390/net/qeth.h
@@ -686,6 +686,7 @@
 	__u32 pdu_hdr;
 	__u32 pdu_hdr_ack;
 	__u16 ipa;
+	__u32 pkt_seqno;
 };
 
 struct qeth_reply {
@@ -848,6 +849,7 @@
                                   "on interface %s", QETH_CARD_IFNAME(card));
                         return -ENOMEM;
                 }
+		kfree_skb(*skb);
                 *skb = new_skb;
 	}
 	return 0;
diff --git a/drivers/s390/net/qeth_main.c b/drivers/s390/net/qeth_main.c
index 71de834..bd28e24 100644
--- a/drivers/s390/net/qeth_main.c
+++ b/drivers/s390/net/qeth_main.c
@@ -511,7 +511,7 @@
 __qeth_set_offline(struct ccwgroup_device *cgdev, int recovery_mode)
 {
 	struct qeth_card *card = (struct qeth_card *) cgdev->dev.driver_data;
-	int rc = 0;
+	int rc = 0, rc2 = 0, rc3 = 0;
 	enum qeth_card_states recover_flag;
 
 	QETH_DBF_TEXT(setup, 3, "setoffl");
@@ -523,11 +523,13 @@
 			   CARD_BUS_ID(card));
 		return -ERESTARTSYS;
 	}
-	if ((rc = ccw_device_set_offline(CARD_DDEV(card))) ||
-	    (rc = ccw_device_set_offline(CARD_WDEV(card))) ||
-	    (rc = ccw_device_set_offline(CARD_RDEV(card)))) {
+	rc  = ccw_device_set_offline(CARD_DDEV(card));
+	rc2 = ccw_device_set_offline(CARD_WDEV(card));
+	rc3 = ccw_device_set_offline(CARD_RDEV(card));
+	if (!rc)
+		rc = (rc2) ? rc2 : rc3;
+	if (rc)
 		QETH_DBF_TEXT_(setup, 2, "1err%d", rc);
-	}
 	if (recover_flag == CARD_STATE_UP)
 		card->state = CARD_STATE_RECOVER;
 	qeth_notify_processes();
@@ -1046,6 +1048,7 @@
 	spin_lock_init(&card->vlanlock);
 	card->vlangrp = NULL;
 #endif
+	spin_lock_init(&card->lock);
 	spin_lock_init(&card->ip_lock);
 	spin_lock_init(&card->thread_mask_lock);
 	card->thread_start_mask = 0;
@@ -1626,16 +1629,6 @@
 	spin_unlock_irqrestore(&reply->card->lock, flags);
 }
 
-static void
-qeth_reset_ip_addresses(struct qeth_card *card)
-{
-	QETH_DBF_TEXT(trace, 2, "rstipadd");
-
-	qeth_clear_ip_list(card, 0, 1);
-	/* this function will also schedule the SET_IP_THREAD */
-	qeth_set_multicast_list(card->dev);
-}
-
 static struct qeth_ipa_cmd *
 qeth_check_ipa_data(struct qeth_card *card, struct qeth_cmd_buffer *iob)
 {
@@ -1664,9 +1657,8 @@
 					   "IP address reset.\n",
 					   QETH_CARD_IFNAME(card),
 					   card->info.chpid);
-				card->lan_online = 1;
 				netif_carrier_on(card->dev);
-				qeth_reset_ip_addresses(card);
+				qeth_schedule_recovery(card);
 				return NULL;
 			case IPA_CMD_REGISTER_LOCAL_ADDR:
 				QETH_DBF_TEXT(trace,3, "irla");
@@ -2387,6 +2379,7 @@
 		skb_pull(skb, VLAN_HLEN);
 	}
 #endif
+	*((__u32 *)skb->cb) = ++card->seqno.pkt_seqno;
 	return vlan_id;
 }
 
@@ -3014,7 +3007,7 @@
 			return -ENOMEM;
 		}
 		for(j = 0; j < QETH_MAX_BUFFER_ELEMENTS(card); ++j){
-			ptr = (void *) __get_free_page(GFP_KERNEL);
+			ptr = (void *) __get_free_page(GFP_KERNEL|GFP_DMA);
 			if (!ptr) {
 				while (j > 0)
 					free_page((unsigned long)
@@ -3058,7 +3051,8 @@
 	if (card->qdio.state == QETH_QDIO_ALLOCATED)
 		return 0;
 
-	card->qdio.in_q = kmalloc(sizeof(struct qeth_qdio_q), GFP_KERNEL);
+	card->qdio.in_q = kmalloc(sizeof(struct qeth_qdio_q), 
+				  GFP_KERNEL|GFP_DMA);
 	if (!card->qdio.in_q)
 		return - ENOMEM;
 	QETH_DBF_TEXT(setup, 2, "inq");
@@ -3083,7 +3077,7 @@
 	}
 	for (i = 0; i < card->qdio.no_out_queues; ++i){
 		card->qdio.out_qs[i] = kmalloc(sizeof(struct qeth_qdio_out_q),
-					       GFP_KERNEL);
+					       GFP_KERNEL|GFP_DMA);
 		if (!card->qdio.out_qs[i]){
 			while (i > 0)
 				kfree(card->qdio.out_qs[--i]);
@@ -6470,6 +6464,9 @@
 	if (cmd->hdr.prot_version == QETH_PROT_IPV4) {
 		card->options.ipa4.supported_funcs = cmd->hdr.ipa_supported;
 		card->options.ipa4.enabled_funcs = cmd->hdr.ipa_enabled;
+		/* Disable IPV6 support hard coded for Hipersockets */
+		if(card->info.type == QETH_CARD_TYPE_IQD)
+			card->options.ipa4.supported_funcs &= ~IPA_IPV6;
 	} else {
 #ifdef CONFIG_QETH_IPV6
 		card->options.ipa6.supported_funcs = cmd->hdr.ipa_supported;
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index 20019b8..be96cb7 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -521,6 +521,14 @@
 
 	  If unsure, say N.
 
+config SCSI_SATA_SIL24
+	tristate "Silicon Image 3124/3132 SATA support"
+	depends on SCSI_SATA && PCI && EXPERIMENTAL
+	help
+	  This option enables support for Silicon Image 3124/3132 Serial ATA.
+
+	  If unsure, say N.
+
 config SCSI_SATA_SIS
 	tristate "SiS 964/180 SATA support"
 	depends on SCSI_SATA && PCI && EXPERIMENTAL
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index 48529d1..e2e3d86 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -130,6 +130,7 @@
 obj-$(CONFIG_SCSI_SATA_PROMISE)	+= libata.o sata_promise.o
 obj-$(CONFIG_SCSI_SATA_QSTOR)	+= libata.o sata_qstor.o
 obj-$(CONFIG_SCSI_SATA_SIL)	+= libata.o sata_sil.o
+obj-$(CONFIG_SCSI_SATA_SIL24)	+= libata.o sata_sil24.o
 obj-$(CONFIG_SCSI_SATA_VIA)	+= libata.o sata_via.o
 obj-$(CONFIG_SCSI_SATA_VITESSE)	+= libata.o sata_vsc.o
 obj-$(CONFIG_SCSI_SATA_SIS)	+= libata.o sata_sis.o
diff --git a/drivers/scsi/ahci.c b/drivers/scsi/ahci.c
index 6e4bb36..f0d8f89 100644
--- a/drivers/scsi/ahci.c
+++ b/drivers/scsi/ahci.c
@@ -680,17 +680,36 @@
 
         for (i = 0; i < host_set->n_ports; i++) {
 		struct ata_port *ap;
-		u32 tmp;
 
-		VPRINTK("port %u\n", i);
+		if (!(irq_stat & (1 << i)))
+			continue;
+
 		ap = host_set->ports[i];
-		tmp = irq_stat & (1 << i);
-		if (tmp && ap) {
+		if (ap) {
 			struct ata_queued_cmd *qc;
 			qc = ata_qc_from_tag(ap, ap->active_tag);
-			if (ahci_host_intr(ap, qc))
-				irq_ack |= (1 << i);
+			if (!ahci_host_intr(ap, qc))
+				if (ata_ratelimit()) {
+					struct pci_dev *pdev =
+					  to_pci_dev(ap->host_set->dev);
+					printk(KERN_WARNING
+					  "ahci(%s): unhandled interrupt on port %u\n",
+					  pci_name(pdev), i);
+				}
+
+			VPRINTK("port %u\n", i);
+		} else {
+			VPRINTK("port %u (no irq)\n", i);
+			if (ata_ratelimit()) {
+				struct pci_dev *pdev =
+				  to_pci_dev(ap->host_set->dev);
+				printk(KERN_WARNING
+				  "ahci(%s): interrupt on disabled port %u\n",
+				  pci_name(pdev), i);
+			}
 		}
+
+		irq_ack |= (1 << i);
 	}
 
 	if (irq_ack) {
diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index 943b44c..9aa9308 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -48,6 +48,7 @@
 #include <linux/completion.h>
 #include <linux/suspend.h>
 #include <linux/workqueue.h>
+#include <linux/jiffies.h>
 #include <scsi/scsi.h>
 #include "scsi.h"
 #include "scsi_priv.h"
@@ -62,6 +63,7 @@
 static unsigned int ata_busy_sleep (struct ata_port *ap,
 				    unsigned long tmout_pat,
 			    	    unsigned long tmout);
+static void ata_dev_init_params(struct ata_port *ap, struct ata_device *dev);
 static void ata_set_mode(struct ata_port *ap);
 static void ata_dev_set_xfermode(struct ata_port *ap, struct ata_device *dev);
 static unsigned int ata_get_mode_mask(struct ata_port *ap, int shift);
@@ -69,7 +71,6 @@
 static int ata_choose_xfer_mode(struct ata_port *ap,
 				u8 *xfer_mode_out,
 				unsigned int *xfer_shift_out);
-static int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat);
 static void __ata_qc_complete(struct ata_queued_cmd *qc);
 
 static unsigned int ata_unique_id = 1;
@@ -1131,7 +1132,7 @@
 static void ata_dev_identify(struct ata_port *ap, unsigned int device)
 {
 	struct ata_device *dev = &ap->device[device];
-	unsigned int i;
+	unsigned int major_version;
 	u16 tmp;
 	unsigned long xfer_modes;
 	u8 status;
@@ -1229,9 +1230,9 @@
 	 * common ATA, ATAPI feature tests
 	 */
 
-	/* we require LBA and DMA support (bits 8 & 9 of word 49) */
-	if (!ata_id_has_dma(dev->id) || !ata_id_has_lba(dev->id)) {
-		printk(KERN_DEBUG "ata%u: no dma/lba\n", ap->id);
+	/* we require DMA support (bits 8 of word 49) */
+	if (!ata_id_has_dma(dev->id)) {
+		printk(KERN_DEBUG "ata%u: no dma\n", ap->id);
 		goto err_out_nosup;
 	}
 
@@ -1251,32 +1252,69 @@
 		if (!ata_id_is_ata(dev->id))	/* sanity check */
 			goto err_out_nosup;
 
+		/* get major version */
 		tmp = dev->id[ATA_ID_MAJOR_VER];
-		for (i = 14; i >= 1; i--)
-			if (tmp & (1 << i))
+		for (major_version = 14; major_version >= 1; major_version--)
+			if (tmp & (1 << major_version))
 				break;
 
-		/* we require at least ATA-3 */
-		if (i < 3) {
-			printk(KERN_DEBUG "ata%u: no ATA-3\n", ap->id);
-			goto err_out_nosup;
-		}
+		/*
+		 * The exact sequence expected by certain pre-ATA4 drives is:
+		 * SRST RESET
+		 * IDENTIFY
+		 * INITIALIZE DEVICE PARAMETERS
+		 * anything else..
+		 * Some drives were very specific about that exact sequence.
+		 */
+		if (major_version < 4 || (!ata_id_has_lba(dev->id)))
+			ata_dev_init_params(ap, dev);
 
-		if (ata_id_has_lba48(dev->id)) {
-			dev->flags |= ATA_DFLAG_LBA48;
-			dev->n_sectors = ata_id_u64(dev->id, 100);
-		} else {
-			dev->n_sectors = ata_id_u32(dev->id, 60);
+		if (ata_id_has_lba(dev->id)) {
+			dev->flags |= ATA_DFLAG_LBA;
+
+			if (ata_id_has_lba48(dev->id)) {
+				dev->flags |= ATA_DFLAG_LBA48;
+				dev->n_sectors = ata_id_u64(dev->id, 100);
+			} else {
+				dev->n_sectors = ata_id_u32(dev->id, 60);
+			}
+
+			/* print device info to dmesg */
+			printk(KERN_INFO "ata%u: dev %u ATA-%d, max %s, %Lu sectors:%s\n",
+			       ap->id, device,
+			       major_version,
+			       ata_mode_string(xfer_modes),
+			       (unsigned long long)dev->n_sectors,
+			       dev->flags & ATA_DFLAG_LBA48 ? " LBA48" : " LBA");
+		} else { 
+			/* CHS */
+
+			/* Default translation */
+			dev->cylinders	= dev->id[1];
+			dev->heads	= dev->id[3];
+			dev->sectors	= dev->id[6];
+			dev->n_sectors	= dev->cylinders * dev->heads * dev->sectors;
+
+			if (ata_id_current_chs_valid(dev->id)) {
+				/* Current CHS translation is valid. */
+				dev->cylinders = dev->id[54];
+				dev->heads     = dev->id[55];
+				dev->sectors   = dev->id[56];
+				
+				dev->n_sectors = ata_id_u32(dev->id, 57);
+			}
+
+			/* print device info to dmesg */
+			printk(KERN_INFO "ata%u: dev %u ATA-%d, max %s, %Lu sectors: CHS %d/%d/%d\n",
+			       ap->id, device,
+			       major_version,
+			       ata_mode_string(xfer_modes),
+			       (unsigned long long)dev->n_sectors,
+			       (int)dev->cylinders, (int)dev->heads, (int)dev->sectors);
+
 		}
 
 		ap->host->max_cmd_len = 16;
-
-		/* print device info to dmesg */
-		printk(KERN_INFO "ata%u: dev %u ATA, max %s, %Lu sectors:%s\n",
-		       ap->id, device,
-		       ata_mode_string(xfer_modes),
-		       (unsigned long long)dev->n_sectors,
-		       dev->flags & ATA_DFLAG_LBA48 ? " lba48" : "");
 	}
 
 	/* ATAPI-specific feature tests */
@@ -2144,6 +2182,54 @@
 }
 
 /**
+ *	ata_dev_init_params - Issue INIT DEV PARAMS command
+ *	@ap: Port associated with device @dev
+ *	@dev: Device to which command will be sent
+ *
+ *	LOCKING:
+ */
+
+static void ata_dev_init_params(struct ata_port *ap, struct ata_device *dev)
+{
+	DECLARE_COMPLETION(wait);
+	struct ata_queued_cmd *qc;
+	int rc;
+	unsigned long flags;
+	u16 sectors = dev->id[6];
+	u16 heads   = dev->id[3];
+
+	/* Number of sectors per track 1-255. Number of heads 1-16 */
+	if (sectors < 1 || sectors > 255 || heads < 1 || heads > 16)
+		return;
+
+	/* set up init dev params taskfile */
+	DPRINTK("init dev params \n");
+
+	qc = ata_qc_new_init(ap, dev);
+	BUG_ON(qc == NULL);
+
+	qc->tf.command = ATA_CMD_INIT_DEV_PARAMS;
+	qc->tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
+	qc->tf.protocol = ATA_PROT_NODATA;
+	qc->tf.nsect = sectors;
+	qc->tf.device |= (heads - 1) & 0x0f; /* max head = num. of heads - 1 */
+
+	qc->waiting = &wait;
+	qc->complete_fn = ata_qc_complete_noop;
+
+	spin_lock_irqsave(&ap->host_set->lock, flags);
+	rc = ata_qc_issue(qc);
+	spin_unlock_irqrestore(&ap->host_set->lock, flags);
+
+	if (rc)
+		ata_port_disable(ap);
+	else
+		wait_for_completion(&wait);
+
+	DPRINTK("EXIT\n");
+}
+
+/**
  *	ata_sg_clean - Unmap DMA memory associated with command
  *	@qc: Command containing DMA memory to be released
  *
@@ -2507,20 +2593,20 @@
 static unsigned long ata_pio_poll(struct ata_port *ap)
 {
 	u8 status;
-	unsigned int poll_state = PIO_ST_UNKNOWN;
-	unsigned int reg_state = PIO_ST_UNKNOWN;
-	const unsigned int tmout_state = PIO_ST_TMOUT;
+	unsigned int poll_state = HSM_ST_UNKNOWN;
+	unsigned int reg_state = HSM_ST_UNKNOWN;
+	const unsigned int tmout_state = HSM_ST_TMOUT;
 
-	switch (ap->pio_task_state) {
-	case PIO_ST:
-	case PIO_ST_POLL:
-		poll_state = PIO_ST_POLL;
-		reg_state = PIO_ST;
+	switch (ap->hsm_task_state) {
+	case HSM_ST:
+	case HSM_ST_POLL:
+		poll_state = HSM_ST_POLL;
+		reg_state = HSM_ST;
 		break;
-	case PIO_ST_LAST:
-	case PIO_ST_LAST_POLL:
-		poll_state = PIO_ST_LAST_POLL;
-		reg_state = PIO_ST_LAST;
+	case HSM_ST_LAST:
+	case HSM_ST_LAST_POLL:
+		poll_state = HSM_ST_LAST_POLL;
+		reg_state = HSM_ST_LAST;
 		break;
 	default:
 		BUG();
@@ -2530,14 +2616,14 @@
 	status = ata_chk_status(ap);
 	if (status & ATA_BUSY) {
 		if (time_after(jiffies, ap->pio_task_timeout)) {
-			ap->pio_task_state = tmout_state;
+			ap->hsm_task_state = tmout_state;
 			return 0;
 		}
-		ap->pio_task_state = poll_state;
+		ap->hsm_task_state = poll_state;
 		return ATA_SHORT_PAUSE;
 	}
 
-	ap->pio_task_state = reg_state;
+	ap->hsm_task_state = reg_state;
 	return 0;
 }
 
@@ -2562,14 +2648,14 @@
 	 * we enter, BSY will be cleared in a chk-status or two.  If not,
 	 * the drive is probably seeking or something.  Snooze for a couple
 	 * msecs, then chk-status again.  If still busy, fall back to
-	 * PIO_ST_POLL state.
+	 * HSM_ST_POLL state.
 	 */
 	drv_stat = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 10);
 	if (drv_stat & (ATA_BUSY | ATA_DRQ)) {
 		msleep(2);
 		drv_stat = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 10);
 		if (drv_stat & (ATA_BUSY | ATA_DRQ)) {
-			ap->pio_task_state = PIO_ST_LAST_POLL;
+			ap->hsm_task_state = HSM_ST_LAST_POLL;
 			ap->pio_task_timeout = jiffies + ATA_TMOUT_PIO;
 			return 0;
 		}
@@ -2577,14 +2663,14 @@
 
 	drv_stat = ata_wait_idle(ap);
 	if (!ata_ok(drv_stat)) {
-		ap->pio_task_state = PIO_ST_ERR;
+		ap->hsm_task_state = HSM_ST_ERR;
 		return 0;
 	}
 
 	qc = ata_qc_from_tag(ap, ap->active_tag);
 	assert(qc != NULL);
 
-	ap->pio_task_state = PIO_ST_IDLE;
+	ap->hsm_task_state = HSM_ST_IDLE;
 
 	ata_poll_qc_complete(qc, drv_stat);
 
@@ -2744,7 +2830,7 @@
 	unsigned char *buf;
 
 	if (qc->cursect == (qc->nsect - 1))
-		ap->pio_task_state = PIO_ST_LAST;
+		ap->hsm_task_state = HSM_ST_LAST;
 
 	page = sg[qc->cursg].page;
 	offset = sg[qc->cursg].offset + qc->cursg_ofs * ATA_SECT_SIZE;
@@ -2794,7 +2880,7 @@
 	unsigned int offset, count;
 
 	if (qc->curbytes + bytes >= qc->nbytes)
-		ap->pio_task_state = PIO_ST_LAST;
+		ap->hsm_task_state = HSM_ST_LAST;
 
 next_sg:
 	if (unlikely(qc->cursg >= qc->n_elem)) {
@@ -2816,7 +2902,7 @@
 		for (i = 0; i < words; i++)
 			ata_data_xfer(ap, (unsigned char*)pad_buf, 2, do_write);
 
-		ap->pio_task_state = PIO_ST_LAST;
+		ap->hsm_task_state = HSM_ST_LAST;
 		return;
 	}
 
@@ -2897,7 +2983,7 @@
 err_out:
 	printk(KERN_INFO "ata%u: dev %u: ATAPI check failed\n",
 	      ap->id, dev->devno);
-	ap->pio_task_state = PIO_ST_ERR;
+	ap->hsm_task_state = HSM_ST_ERR;
 }
 
 /**
@@ -2919,14 +3005,14 @@
 	 * a chk-status or two.  If not, the drive is probably seeking
 	 * or something.  Snooze for a couple msecs, then
 	 * chk-status again.  If still busy, fall back to
-	 * PIO_ST_POLL state.
+	 * HSM_ST_POLL state.
 	 */
 	status = ata_busy_wait(ap, ATA_BUSY, 5);
 	if (status & ATA_BUSY) {
 		msleep(2);
 		status = ata_busy_wait(ap, ATA_BUSY, 10);
 		if (status & ATA_BUSY) {
-			ap->pio_task_state = PIO_ST_POLL;
+			ap->hsm_task_state = HSM_ST_POLL;
 			ap->pio_task_timeout = jiffies + ATA_TMOUT_PIO;
 			return;
 		}
@@ -2938,7 +3024,7 @@
 	if (is_atapi_taskfile(&qc->tf)) {
 		/* no more data to transfer or unsupported ATAPI command */
 		if ((status & ATA_DRQ) == 0) {
-			ap->pio_task_state = PIO_ST_LAST;
+			ap->hsm_task_state = HSM_ST_LAST;
 			return;
 		}
 
@@ -2946,7 +3032,7 @@
 	} else {
 		/* handle BSY=0, DRQ=0 as error */
 		if ((status & ATA_DRQ) == 0) {
-			ap->pio_task_state = PIO_ST_ERR;
+			ap->hsm_task_state = HSM_ST_ERR;
 			return;
 		}
 
@@ -2966,7 +3052,7 @@
 	printk(KERN_WARNING "ata%u: PIO error, drv_stat 0x%x\n",
 	       ap->id, drv_stat);
 
-	ap->pio_task_state = PIO_ST_IDLE;
+	ap->hsm_task_state = HSM_ST_IDLE;
 
 	ata_poll_qc_complete(qc, drv_stat | ATA_ERR);
 }
@@ -2981,25 +3067,25 @@
 	timeout = 0;
 	qc_completed = 0;
 
-	switch (ap->pio_task_state) {
-	case PIO_ST_IDLE:
+	switch (ap->hsm_task_state) {
+	case HSM_ST_IDLE:
 		return;
 
-	case PIO_ST:
+	case HSM_ST:
 		ata_pio_block(ap);
 		break;
 
-	case PIO_ST_LAST:
+	case HSM_ST_LAST:
 		qc_completed = ata_pio_complete(ap);
 		break;
 
-	case PIO_ST_POLL:
-	case PIO_ST_LAST_POLL:
+	case HSM_ST_POLL:
+	case HSM_ST_LAST_POLL:
 		timeout = ata_pio_poll(ap);
 		break;
 
-	case PIO_ST_TMOUT:
-	case PIO_ST_ERR:
+	case HSM_ST_TMOUT:
+	case HSM_ST_ERR:
 		ata_pio_error(ap);
 		return;
 	}
@@ -3010,52 +3096,6 @@
 		goto fsm_start;
 }
 
-static void atapi_request_sense(struct ata_port *ap, struct ata_device *dev,
-				struct scsi_cmnd *cmd)
-{
-	DECLARE_COMPLETION(wait);
-	struct ata_queued_cmd *qc;
-	unsigned long flags;
-	int rc;
-
-	DPRINTK("ATAPI request sense\n");
-
-	qc = ata_qc_new_init(ap, dev);
-	BUG_ON(qc == NULL);
-
-	/* FIXME: is this needed? */
-	memset(cmd->sense_buffer, 0, sizeof(cmd->sense_buffer));
-
-	ata_sg_init_one(qc, cmd->sense_buffer, sizeof(cmd->sense_buffer));
-	qc->dma_dir = DMA_FROM_DEVICE;
-
-	memset(&qc->cdb, 0, ap->cdb_len);
-	qc->cdb[0] = REQUEST_SENSE;
-	qc->cdb[4] = SCSI_SENSE_BUFFERSIZE;
-
-	qc->tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
-	qc->tf.command = ATA_CMD_PACKET;
-
-	qc->tf.protocol = ATA_PROT_ATAPI;
-	qc->tf.lbam = (8 * 1024) & 0xff;
-	qc->tf.lbah = (8 * 1024) >> 8;
-	qc->nbytes = SCSI_SENSE_BUFFERSIZE;
-
-	qc->waiting = &wait;
-	qc->complete_fn = ata_qc_complete_noop;
-
-	spin_lock_irqsave(&ap->host_set->lock, flags);
-	rc = ata_qc_issue(qc);
-	spin_unlock_irqrestore(&ap->host_set->lock, flags);
-
-	if (rc)
-		ata_port_disable(ap);
-	else
-		wait_for_completion(&wait);
-
-	DPRINTK("EXIT\n");
-}
-
 /**
  *	ata_qc_timeout - Handle timeout of queued command
  *	@qc: Command that timed out
@@ -3173,14 +3213,14 @@
 	DPRINTK("ENTER\n");
 
 	qc = ata_qc_from_tag(ap, ap->active_tag);
-	if (!qc) {
+	if (qc)
+		ata_qc_timeout(qc);
+	else {
 		printk(KERN_ERR "ata%u: BUG: timeout without command\n",
 		       ap->id);
 		goto out;
 	}
 
-	ata_qc_timeout(qc);
-
 out:
 	DPRINTK("EXIT\n");
 }
@@ -3238,14 +3278,18 @@
 
 		ata_tf_init(ap, &qc->tf, dev->devno);
 
-		if (dev->flags & ATA_DFLAG_LBA48)
-			qc->tf.flags |= ATA_TFLAG_LBA48;
+		if (dev->flags & ATA_DFLAG_LBA) {
+			qc->tf.flags |= ATA_TFLAG_LBA;
+
+			if (dev->flags & ATA_DFLAG_LBA48)
+				qc->tf.flags |= ATA_TFLAG_LBA48;
+		}
 	}
 
 	return qc;
 }
 
-static int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat)
+int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat)
 {
 	return 0;
 }
@@ -3442,7 +3486,7 @@
 	case ATA_PROT_PIO: /* load tf registers, initiate polling pio */
 		ata_qc_set_polling(qc);
 		ata_tf_to_host_nolock(ap, &qc->tf);
-		ap->pio_task_state = PIO_ST;
+		ap->hsm_task_state = HSM_ST;
 		queue_work(ata_wq, &ap->pio_task);
 		break;
 
@@ -3668,7 +3712,7 @@
 		void __iomem *mmio = (void __iomem *) ap->ioaddr.bmdma_addr;
 		host_stat = readb(mmio + ATA_DMA_STATUS);
 	} else
-	host_stat = inb(ap->ioaddr.bmdma_addr + ATA_DMA_STATUS);
+		host_stat = inb(ap->ioaddr.bmdma_addr + ATA_DMA_STATUS);
 	return host_stat;
 }
 
@@ -3888,7 +3932,7 @@
 		ata_data_xfer(ap, qc->cdb, ap->cdb_len, 1);
 
 		/* PIO commands are handled by polling */
-		ap->pio_task_state = PIO_ST;
+		ap->hsm_task_state = HSM_ST;
 		queue_work(ata_wq, &ap->pio_task);
 	}
 
@@ -4202,7 +4246,7 @@
 	for (i = 0; i < count; i++) {
 		struct ata_port *ap = host_set->ports[i];
 
-		scsi_scan_host(ap->host);
+		ata_scsi_scan_host(ap);
 	}
 
 	dev_set_drvdata(dev, host_set);
@@ -4362,85 +4406,87 @@
  *	ata_pci_init_native_mode - Initialize native-mode driver
  *	@pdev:  pci device to be initialized
  *	@port:  array[2] of pointers to port info structures.
+ *	@ports: bitmap of ports present
  *
  *	Utility function which allocates and initializes an
  *	ata_probe_ent structure for a standard dual-port
  *	PIO-based IDE controller.  The returned ata_probe_ent
  *	structure can be passed to ata_device_add().  The returned
  *	ata_probe_ent structure should then be freed with kfree().
+ *
+ *	The caller need only pass the address of the primary port, the
+ *	secondary will be deduced automatically. If the device has non
+ *	standard secondary port mappings this function can be called twice,
+ *	once for each interface.
  */
 
 struct ata_probe_ent *
-ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port)
+ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int ports)
 {
 	struct ata_probe_ent *probe_ent =
 		ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[0]);
+	int p = 0;
+
 	if (!probe_ent)
 		return NULL;
 
-	probe_ent->n_ports = 2;
 	probe_ent->irq = pdev->irq;
 	probe_ent->irq_flags = SA_SHIRQ;
 
-	probe_ent->port[0].cmd_addr = pci_resource_start(pdev, 0);
-	probe_ent->port[0].altstatus_addr =
-	probe_ent->port[0].ctl_addr =
-		pci_resource_start(pdev, 1) | ATA_PCI_CTL_OFS;
-	probe_ent->port[0].bmdma_addr = pci_resource_start(pdev, 4);
+	if (ports & ATA_PORT_PRIMARY) {
+		probe_ent->port[p].cmd_addr = pci_resource_start(pdev, 0);
+		probe_ent->port[p].altstatus_addr =
+		probe_ent->port[p].ctl_addr =
+			pci_resource_start(pdev, 1) | ATA_PCI_CTL_OFS;
+		probe_ent->port[p].bmdma_addr = pci_resource_start(pdev, 4);
+		ata_std_ports(&probe_ent->port[p]);
+		p++;
+	}
 
-	probe_ent->port[1].cmd_addr = pci_resource_start(pdev, 2);
-	probe_ent->port[1].altstatus_addr =
-	probe_ent->port[1].ctl_addr =
-		pci_resource_start(pdev, 3) | ATA_PCI_CTL_OFS;
-	probe_ent->port[1].bmdma_addr = pci_resource_start(pdev, 4) + 8;
+	if (ports & ATA_PORT_SECONDARY) {
+		probe_ent->port[p].cmd_addr = pci_resource_start(pdev, 2);
+		probe_ent->port[p].altstatus_addr =
+		probe_ent->port[p].ctl_addr =
+			pci_resource_start(pdev, 3) | ATA_PCI_CTL_OFS;
+		probe_ent->port[p].bmdma_addr = pci_resource_start(pdev, 4) + 8;
+		ata_std_ports(&probe_ent->port[p]);
+		p++;
+	}
 
-	ata_std_ports(&probe_ent->port[0]);
-	ata_std_ports(&probe_ent->port[1]);
-
+	probe_ent->n_ports = p;
 	return probe_ent;
 }
 
-static struct ata_probe_ent *
-ata_pci_init_legacy_mode(struct pci_dev *pdev, struct ata_port_info **port,
-    struct ata_probe_ent **ppe2)
+static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev, struct ata_port_info **port, int port_num)
 {
-	struct ata_probe_ent *probe_ent, *probe_ent2;
+	struct ata_probe_ent *probe_ent;
 
 	probe_ent = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[0]);
 	if (!probe_ent)
 		return NULL;
-	probe_ent2 = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[1]);
-	if (!probe_ent2) {
-		kfree(probe_ent);
-		return NULL;
-	}
 
-	probe_ent->n_ports = 1;
-	probe_ent->irq = 14;
-
-	probe_ent->hard_port_no = 0;
+	
 	probe_ent->legacy_mode = 1;
+	probe_ent->n_ports = 1;
+	probe_ent->hard_port_no = port_num;
 
-	probe_ent2->n_ports = 1;
-	probe_ent2->irq = 15;
-
-	probe_ent2->hard_port_no = 1;
-	probe_ent2->legacy_mode = 1;
-
-	probe_ent->port[0].cmd_addr = 0x1f0;
-	probe_ent->port[0].altstatus_addr =
-	probe_ent->port[0].ctl_addr = 0x3f6;
-	probe_ent->port[0].bmdma_addr = pci_resource_start(pdev, 4);
-
-	probe_ent2->port[0].cmd_addr = 0x170;
-	probe_ent2->port[0].altstatus_addr =
-	probe_ent2->port[0].ctl_addr = 0x376;
-	probe_ent2->port[0].bmdma_addr = pci_resource_start(pdev, 4)+8;
-
+	switch(port_num)
+	{
+		case 0:
+			probe_ent->irq = 14;
+			probe_ent->port[0].cmd_addr = 0x1f0;
+			probe_ent->port[0].altstatus_addr =
+			probe_ent->port[0].ctl_addr = 0x3f6;
+			break;
+		case 1:
+			probe_ent->irq = 15;
+			probe_ent->port[0].cmd_addr = 0x170;
+			probe_ent->port[0].altstatus_addr =
+			probe_ent->port[0].ctl_addr = 0x376;
+			break;
+	}
+	probe_ent->port[0].bmdma_addr = pci_resource_start(pdev, 4) + 8 * port_num;
 	ata_std_ports(&probe_ent->port[0]);
-	ata_std_ports(&probe_ent2->port[0]);
-
-	*ppe2 = probe_ent2;
 	return probe_ent;
 }
 
@@ -4469,7 +4515,7 @@
 int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 		      unsigned int n_ports)
 {
-	struct ata_probe_ent *probe_ent, *probe_ent2 = NULL;
+	struct ata_probe_ent *probe_ent = NULL, *probe_ent2 = NULL;
 	struct ata_port_info *port[2];
 	u8 tmp8, mask;
 	unsigned int legacy_mode = 0;
@@ -4486,7 +4532,7 @@
 
 	if ((port[0]->host_flags & ATA_FLAG_NO_LEGACY) == 0
 	    && (pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		/* TODO: support transitioning to native mode? */
+		/* TODO: What if one channel is in native mode ... */
 		pci_read_config_byte(pdev, PCI_CLASS_PROG, &tmp8);
 		mask = (1 << 2) | (1 << 0);
 		if ((tmp8 & mask) != mask)
@@ -4494,11 +4540,20 @@
 	}
 
 	/* FIXME... */
-	if ((!legacy_mode) && (n_ports > 1)) {
-		printk(KERN_ERR "ata: BUG: native mode, n_ports > 1\n");
-		return -EINVAL;
+	if ((!legacy_mode) && (n_ports > 2)) {
+		printk(KERN_ERR "ata: BUG: native mode, n_ports > 2\n");
+		n_ports = 2;
+		/* For now */
 	}
 
+	/* FIXME: Really for ATA it isn't safe because the device may be
+	   multi-purpose and we want to leave it alone if it was already
+	   enabled. Secondly for shared use as Arjan says we want refcounting
+	   
+	   Checking dev->is_enabled is insufficient as this is not set at
+	   boot for the primary video which is BIOS enabled
+         */
+         
 	rc = pci_enable_device(pdev);
 	if (rc)
 		return rc;
@@ -4509,6 +4564,7 @@
 		goto err_out;
 	}
 
+	/* FIXME: Should use platform specific mappers for legacy port ranges */
 	if (legacy_mode) {
 		if (!request_region(0x1f0, 8, "libata")) {
 			struct resource *conflict, res;
@@ -4553,10 +4609,17 @@
 		goto err_out_regions;
 
 	if (legacy_mode) {
-		probe_ent = ata_pci_init_legacy_mode(pdev, port, &probe_ent2);
-	} else
-		probe_ent = ata_pci_init_native_mode(pdev, port);
-	if (!probe_ent) {
+		if (legacy_mode & (1 << 0))
+			probe_ent = ata_pci_init_legacy_port(pdev, port, 0);
+		if (legacy_mode & (1 << 1))
+			probe_ent2 = ata_pci_init_legacy_port(pdev, port, 1);
+	} else {
+		if (n_ports == 2)
+			probe_ent = ata_pci_init_native_mode(pdev, port, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
+		else
+			probe_ent = ata_pci_init_native_mode(pdev, port, ATA_PORT_PRIMARY);
+	}
+	if (!probe_ent && !probe_ent2) {
 		rc = -ENOMEM;
 		goto err_out_regions;
 	}
@@ -4668,6 +4731,27 @@
 module_init(ata_init);
 module_exit(ata_exit);
 
+static unsigned long ratelimit_time;
+static spinlock_t ata_ratelimit_lock = SPIN_LOCK_UNLOCKED;
+
+int ata_ratelimit(void)
+{
+	int rc;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ata_ratelimit_lock, flags);
+
+	if (time_after(jiffies, ratelimit_time)) {
+		rc = 1;
+		ratelimit_time = jiffies + (HZ/5);
+	} else
+		rc = 0;
+
+	spin_unlock_irqrestore(&ata_ratelimit_lock, flags);
+
+	return rc;
+}
+
 /*
  * libata is essentially a library of internal helper functions for
  * low-level ATA host controller drivers.  As such, the API/ABI is
@@ -4709,6 +4793,7 @@
 EXPORT_SYMBOL_GPL(__sata_phy_reset);
 EXPORT_SYMBOL_GPL(ata_bus_reset);
 EXPORT_SYMBOL_GPL(ata_port_disable);
+EXPORT_SYMBOL_GPL(ata_ratelimit);
 EXPORT_SYMBOL_GPL(ata_scsi_ioctl);
 EXPORT_SYMBOL_GPL(ata_scsi_queuecmd);
 EXPORT_SYMBOL_GPL(ata_scsi_error);
diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c
index ee3f105..4cf43de 100644
--- a/drivers/scsi/libata-scsi.c
+++ b/drivers/scsi/libata-scsi.c
@@ -49,6 +49,14 @@
 ata_scsi_find_dev(struct ata_port *ap, struct scsi_device *scsidev);
 
 
+static void ata_scsi_invalid_field(struct scsi_cmnd *cmd,
+				   void (*done)(struct scsi_cmnd *))
+{
+	ata_scsi_set_sense(cmd, ILLEGAL_REQUEST, 0x24, 0x0);
+	/* "Invalid field in cbd" */
+	done(cmd);
+}
+
 /**
  *	ata_std_bios_param - generic bios head/sector/cylinder calculator used by sd.
  *	@sdev: SCSI device for which BIOS geometry is to be determined
@@ -182,7 +190,6 @@
 {
 	struct scsi_cmnd *cmd = qc->scsicmd;
 	u8 err = 0;
-	unsigned char *sb = cmd->sense_buffer;
 	/* Based on the 3ware driver translation table */
 	static unsigned char sense_table[][4] = {
 		/* BBD|ECC|ID|MAR */
@@ -225,8 +232,6 @@
 	};
 	int i = 0;
 
-	cmd->result = SAM_STAT_CHECK_CONDITION;
-
 	/*
 	 *	Is this an error we can process/parse
 	 */
@@ -281,11 +286,9 @@
 		/* Look for best matches first */
 		if((sense_table[i][0] & err) == sense_table[i][0])
 		{
-			sb[0] = 0x70;
-			sb[2] = sense_table[i][1];
-			sb[7] = 0x0a;
-			sb[12] = sense_table[i][2];
-			sb[13] = sense_table[i][3];
+			ata_scsi_set_sense(cmd, sense_table[i][1] /* sk */,
+					   sense_table[i][2] /* asc */,
+					   sense_table[i][3] /* ascq */ );
 			return;
 		}
 		i++;
@@ -300,11 +303,9 @@
 	{
 		if(stat_table[i][0] & drv_stat)
 		{
-			sb[0] = 0x70;
-			sb[2] = stat_table[i][1];
-			sb[7] = 0x0a;
-			sb[12] = stat_table[i][2];
-			sb[13] = stat_table[i][3];
+			ata_scsi_set_sense(cmd, sense_table[i][1] /* sk */,
+					   sense_table[i][2] /* asc */,
+					   sense_table[i][3] /* ascq */ );
 			return;
 		}
 		i++;
@@ -313,15 +314,12 @@
 	printk(KERN_ERR "ata%u: called with no error (%02X)!\n", qc->ap->id, drv_stat);
 	/* additional-sense-code[-qualifier] */
 
-	sb[0] = 0x70;
-	sb[2] = MEDIUM_ERROR;
-	sb[7] = 0x0A;
 	if (cmd->sc_data_direction == DMA_FROM_DEVICE) {
-		sb[12] = 0x11; /* "unrecovered read error" */
-		sb[13] = 0x04;
+		ata_scsi_set_sense(cmd, MEDIUM_ERROR, 0x11, 0x4);
+		/* "unrecovered read error" */
 	} else {
-		sb[12] = 0x0C; /* "write error -             */
-		sb[13] = 0x02; /*  auto-reallocation failed" */
+		ata_scsi_set_sense(cmd, MEDIUM_ERROR, 0xc, 0x2);
+		/* "write error - auto-reallocation failed" */
 	}
 }
 
@@ -440,15 +438,26 @@
 		;	/* ignore IMMED bit, violates sat-r05 */
 	}
 	if (scsicmd[4] & 0x2)
-		return 1;	/* LOEJ bit set not supported */
+		goto invalid_fld;       /* LOEJ bit set not supported */
 	if (((scsicmd[4] >> 4) & 0xf) != 0)
-		return 1;	/* power conditions not supported */
+		goto invalid_fld;       /* power conditions not supported */
 	if (scsicmd[4] & 0x1) {
 		tf->nsect = 1;	/* 1 sector, lba=0 */
-		tf->lbah = 0x0;
-		tf->lbam = 0x0;
-		tf->lbal = 0x0;
-		tf->device |= ATA_LBA;
+
+		if (qc->dev->flags & ATA_DFLAG_LBA) {
+			qc->tf.flags |= ATA_TFLAG_LBA;
+
+			tf->lbah = 0x0;
+			tf->lbam = 0x0;
+			tf->lbal = 0x0;
+			tf->device |= ATA_LBA;
+		} else {
+			/* CHS */
+			tf->lbal = 0x1; /* sect */
+			tf->lbam = 0x0; /* cyl low */
+			tf->lbah = 0x0; /* cyl high */
+		}
+
 		tf->command = ATA_CMD_VERIFY;	/* READ VERIFY */
 	} else {
 		tf->nsect = 0;	/* time period value (0 implies now) */
@@ -463,6 +472,11 @@
 	 */
 
 	return 0;
+
+invalid_fld:
+	ata_scsi_set_sense(qc->scsicmd, ILLEGAL_REQUEST, 0x24, 0x0);
+	/* "Invalid field in cbd" */
+	return 1;
 }
 
 
@@ -498,6 +512,99 @@
 }
 
 /**
+ *	scsi_6_lba_len - Get LBA and transfer length
+ *	@scsicmd: SCSI command to translate
+ *
+ *	Calculate LBA and transfer length for 6-byte commands.
+ *
+ *	RETURNS:
+ *	@plba: the LBA
+ *	@plen: the transfer length
+ */
+
+static void scsi_6_lba_len(u8 *scsicmd, u64 *plba, u32 *plen)
+{
+	u64 lba = 0;
+	u32 len = 0;
+
+	VPRINTK("six-byte command\n");
+
+	lba |= ((u64)scsicmd[2]) << 8;
+	lba |= ((u64)scsicmd[3]);
+
+	len |= ((u32)scsicmd[4]);
+
+	*plba = lba;
+	*plen = len;
+}
+
+/**
+ *	scsi_10_lba_len - Get LBA and transfer length
+ *	@scsicmd: SCSI command to translate
+ *
+ *	Calculate LBA and transfer length for 10-byte commands.
+ *
+ *	RETURNS:
+ *	@plba: the LBA
+ *	@plen: the transfer length
+ */
+
+static void scsi_10_lba_len(u8 *scsicmd, u64 *plba, u32 *plen)
+{
+	u64 lba = 0;
+	u32 len = 0;
+
+	VPRINTK("ten-byte command\n");
+
+	lba |= ((u64)scsicmd[2]) << 24;
+	lba |= ((u64)scsicmd[3]) << 16;
+	lba |= ((u64)scsicmd[4]) << 8;
+	lba |= ((u64)scsicmd[5]);
+
+	len |= ((u32)scsicmd[7]) << 8;
+	len |= ((u32)scsicmd[8]);
+
+	*plba = lba;
+	*plen = len;
+}
+
+/**
+ *	scsi_16_lba_len - Get LBA and transfer length
+ *	@scsicmd: SCSI command to translate
+ *
+ *	Calculate LBA and transfer length for 16-byte commands.
+ *
+ *	RETURNS:
+ *	@plba: the LBA
+ *	@plen: the transfer length
+ */
+
+static void scsi_16_lba_len(u8 *scsicmd, u64 *plba, u32 *plen)
+{
+	u64 lba = 0;
+	u32 len = 0;
+
+	VPRINTK("sixteen-byte command\n");
+
+	lba |= ((u64)scsicmd[2]) << 56;
+	lba |= ((u64)scsicmd[3]) << 48;
+	lba |= ((u64)scsicmd[4]) << 40;
+	lba |= ((u64)scsicmd[5]) << 32;
+	lba |= ((u64)scsicmd[6]) << 24;
+	lba |= ((u64)scsicmd[7]) << 16;
+	lba |= ((u64)scsicmd[8]) << 8;
+	lba |= ((u64)scsicmd[9]);
+
+	len |= ((u32)scsicmd[10]) << 24;
+	len |= ((u32)scsicmd[11]) << 16;
+	len |= ((u32)scsicmd[12]) << 8;
+	len |= ((u32)scsicmd[13]);
+
+	*plba = lba;
+	*plen = len;
+}
+
+/**
  *	ata_scsi_verify_xlat - Translate SCSI VERIFY command into an ATA one
  *	@qc: Storage for translated ATA taskfile
  *	@scsicmd: SCSI command to translate
@@ -514,79 +621,102 @@
 static unsigned int ata_scsi_verify_xlat(struct ata_queued_cmd *qc, u8 *scsicmd)
 {
 	struct ata_taskfile *tf = &qc->tf;
+	struct ata_device *dev = qc->dev;
+	unsigned int lba   = tf->flags & ATA_TFLAG_LBA;
 	unsigned int lba48 = tf->flags & ATA_TFLAG_LBA48;
 	u64 dev_sectors = qc->dev->n_sectors;
-	u64 sect = 0;
-	u32 n_sect = 0;
+	u64 block;
+	u32 n_block;
 
 	tf->flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
 	tf->protocol = ATA_PROT_NODATA;
-	tf->device |= ATA_LBA;
 
-	if (scsicmd[0] == VERIFY) {
-		sect |= ((u64)scsicmd[2]) << 24;
-		sect |= ((u64)scsicmd[3]) << 16;
-		sect |= ((u64)scsicmd[4]) << 8;
-		sect |= ((u64)scsicmd[5]);
-
-		n_sect |= ((u32)scsicmd[7]) << 8;
-		n_sect |= ((u32)scsicmd[8]);
-	}
-
-	else if (scsicmd[0] == VERIFY_16) {
-		sect |= ((u64)scsicmd[2]) << 56;
-		sect |= ((u64)scsicmd[3]) << 48;
-		sect |= ((u64)scsicmd[4]) << 40;
-		sect |= ((u64)scsicmd[5]) << 32;
-		sect |= ((u64)scsicmd[6]) << 24;
-		sect |= ((u64)scsicmd[7]) << 16;
-		sect |= ((u64)scsicmd[8]) << 8;
-		sect |= ((u64)scsicmd[9]);
-
-		n_sect |= ((u32)scsicmd[10]) << 24;
-		n_sect |= ((u32)scsicmd[11]) << 16;
-		n_sect |= ((u32)scsicmd[12]) << 8;
-		n_sect |= ((u32)scsicmd[13]);
-	}
-
+	if (scsicmd[0] == VERIFY)
+		scsi_10_lba_len(scsicmd, &block, &n_block);
+	else if (scsicmd[0] == VERIFY_16)
+		scsi_16_lba_len(scsicmd, &block, &n_block);
 	else
-		return 1;
+		goto invalid_fld;
 
-	if (!n_sect)
-		return 1;
-	if (sect >= dev_sectors)
-		return 1;
-	if ((sect + n_sect) > dev_sectors)
-		return 1;
+	if (!n_block)
+		goto nothing_to_do;
+	if (block >= dev_sectors)
+		goto out_of_range;
+	if ((block + n_block) > dev_sectors)
+		goto out_of_range;
 	if (lba48) {
-		if (n_sect > (64 * 1024))
-			return 1;
+		if (n_block > (64 * 1024))
+			goto invalid_fld;
 	} else {
-		if (n_sect > 256)
-			return 1;
+		if (n_block > 256)
+			goto invalid_fld;
 	}
 
-	if (lba48) {
-		tf->command = ATA_CMD_VERIFY_EXT;
+	if (lba) {
+		if (lba48) {
+			tf->command = ATA_CMD_VERIFY_EXT;
 
-		tf->hob_nsect = (n_sect >> 8) & 0xff;
+			tf->hob_nsect = (n_block >> 8) & 0xff;
 
-		tf->hob_lbah = (sect >> 40) & 0xff;
-		tf->hob_lbam = (sect >> 32) & 0xff;
-		tf->hob_lbal = (sect >> 24) & 0xff;
+			tf->hob_lbah = (block >> 40) & 0xff;
+			tf->hob_lbam = (block >> 32) & 0xff;
+			tf->hob_lbal = (block >> 24) & 0xff;
+		} else {
+			tf->command = ATA_CMD_VERIFY;
+
+			tf->device |= (block >> 24) & 0xf;
+		}
+
+		tf->nsect = n_block & 0xff;
+
+		tf->lbah = (block >> 16) & 0xff;
+		tf->lbam = (block >> 8) & 0xff;
+		tf->lbal = block & 0xff;
+
+		tf->device |= ATA_LBA;
 	} else {
+		/* CHS */
+		u32 sect, head, cyl, track;
+
+		/* Convert LBA to CHS */
+		track = (u32)block / dev->sectors;
+		cyl   = track / dev->heads;
+		head  = track % dev->heads;
+		sect  = (u32)block % dev->sectors + 1;
+
+		DPRINTK("block %u track %u cyl %u head %u sect %u\n",
+			(u32)block, track, cyl, head, sect);
+		
+		/* Check whether the converted CHS can fit. 
+		   Cylinder: 0-65535 
+		   Head: 0-15
+		   Sector: 1-255*/
+		if ((cyl >> 16) || (head >> 4) || (sect >> 8) || (!sect)) 
+			goto out_of_range;
+		
 		tf->command = ATA_CMD_VERIFY;
-
-		tf->device |= (sect >> 24) & 0xf;
+		tf->nsect = n_block & 0xff; /* Sector count 0 means 256 sectors */
+		tf->lbal = sect;
+		tf->lbam = cyl;
+		tf->lbah = cyl >> 8;
+		tf->device |= head;
 	}
 
-	tf->nsect = n_sect & 0xff;
-
-	tf->lbah = (sect >> 16) & 0xff;
-	tf->lbam = (sect >> 8) & 0xff;
-	tf->lbal = sect & 0xff;
-
 	return 0;
+
+invalid_fld:
+	ata_scsi_set_sense(qc->scsicmd, ILLEGAL_REQUEST, 0x24, 0x0);
+	/* "Invalid field in cbd" */
+	return 1;
+
+out_of_range:
+	ata_scsi_set_sense(qc->scsicmd, ILLEGAL_REQUEST, 0x21, 0x0);
+	/* "Logical Block Address out of range" */
+	return 1;
+
+nothing_to_do:
+	qc->scsicmd->result = SAM_STAT_GOOD;
+	return 1;
 }
 
 /**
@@ -612,11 +742,14 @@
 static unsigned int ata_scsi_rw_xlat(struct ata_queued_cmd *qc, u8 *scsicmd)
 {
 	struct ata_taskfile *tf = &qc->tf;
+	struct ata_device *dev = qc->dev;
+	unsigned int lba   = tf->flags & ATA_TFLAG_LBA;
 	unsigned int lba48 = tf->flags & ATA_TFLAG_LBA48;
+	u64 block;
+	u32 n_block;
 
 	tf->flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
 	tf->protocol = qc->dev->xfer_protocol;
-	tf->device |= ATA_LBA;
 
 	if (scsicmd[0] == READ_10 || scsicmd[0] == READ_6 ||
 	    scsicmd[0] == READ_16) {
@@ -626,89 +759,115 @@
 		tf->flags |= ATA_TFLAG_WRITE;
 	}
 
-	if (scsicmd[0] == READ_10 || scsicmd[0] == WRITE_10) {
+	/* Calculate the SCSI LBA and transfer length. */
+	switch (scsicmd[0]) {
+	case READ_10:
+	case WRITE_10:
+		scsi_10_lba_len(scsicmd, &block, &n_block);
+		break;
+	case READ_6:
+	case WRITE_6:
+		scsi_6_lba_len(scsicmd, &block, &n_block);
+
+		/* for 6-byte r/w commands, transfer length 0
+		 * means 256 blocks of data, not 0 block.
+		 */
+		if (!n_block)
+			n_block = 256;
+		break;
+	case READ_16:
+	case WRITE_16:
+		scsi_16_lba_len(scsicmd, &block, &n_block);
+		break;
+	default:
+		DPRINTK("no-byte command\n");
+		goto invalid_fld;
+	}
+
+	/* Check and compose ATA command */
+	if (!n_block)
+		/* For 10-byte and 16-byte SCSI R/W commands, transfer
+		 * length 0 means transfer 0 block of data.
+		 * However, for ATA R/W commands, sector count 0 means
+		 * 256 or 65536 sectors, not 0 sectors as in SCSI.
+		 */
+		goto nothing_to_do;
+
+	if (lba) {
 		if (lba48) {
-			tf->hob_nsect = scsicmd[7];
-			tf->hob_lbal = scsicmd[2];
+			/* The request -may- be too large for LBA48. */
+			if ((block >> 48) || (n_block > 65536))
+				goto out_of_range;
 
-			qc->nsect = ((unsigned int)scsicmd[7] << 8) |
-					scsicmd[8];
-		} else {
-			/* if we don't support LBA48 addressing, the request
-			 * -may- be too large. */
-			if ((scsicmd[2] & 0xf0) || scsicmd[7])
-				return 1;
+			tf->hob_nsect = (n_block >> 8) & 0xff;
 
-			/* stores LBA27:24 in lower 4 bits of device reg */
-			tf->device |= scsicmd[2];
+			tf->hob_lbah = (block >> 40) & 0xff;
+			tf->hob_lbam = (block >> 32) & 0xff;
+			tf->hob_lbal = (block >> 24) & 0xff;
+		} else { 
+			/* LBA28 */
 
-			qc->nsect = scsicmd[8];
+			/* The request -may- be too large for LBA28. */
+			if ((block >> 28) || (n_block > 256))
+				goto out_of_range;
+
+			tf->device |= (block >> 24) & 0xf;
 		}
 
-		tf->nsect = scsicmd[8];
-		tf->lbal = scsicmd[5];
-		tf->lbam = scsicmd[4];
-		tf->lbah = scsicmd[3];
+		qc->nsect = n_block;
+		tf->nsect = n_block & 0xff;
 
-		VPRINTK("ten-byte command\n");
-		if (qc->nsect == 0) /* we don't support length==0 cmds */
-			return 1;
-		return 0;
+		tf->lbah = (block >> 16) & 0xff;
+		tf->lbam = (block >> 8) & 0xff;
+		tf->lbal = block & 0xff;
+
+		tf->device |= ATA_LBA;
+	} else { 
+		/* CHS */
+		u32 sect, head, cyl, track;
+
+		/* The request -may- be too large for CHS addressing. */
+		if ((block >> 28) || (n_block > 256))
+			goto out_of_range;
+
+		/* Convert LBA to CHS */
+		track = (u32)block / dev->sectors;
+		cyl   = track / dev->heads;
+		head  = track % dev->heads;
+		sect  = (u32)block % dev->sectors + 1;
+
+		DPRINTK("block %u track %u cyl %u head %u sect %u\n",
+			(u32)block, track, cyl, head, sect);
+
+		/* Check whether the converted CHS can fit. 
+		   Cylinder: 0-65535 
+		   Head: 0-15
+		   Sector: 1-255*/
+		if ((cyl >> 16) || (head >> 4) || (sect >> 8) || (!sect))
+			goto out_of_range;
+
+		qc->nsect = n_block;
+		tf->nsect = n_block & 0xff; /* Sector count 0 means 256 sectors */
+		tf->lbal = sect;
+		tf->lbam = cyl;
+		tf->lbah = cyl >> 8;
+		tf->device |= head;
 	}
 
-	if (scsicmd[0] == READ_6 || scsicmd[0] == WRITE_6) {
-		qc->nsect = tf->nsect = scsicmd[4];
-		if (!qc->nsect) {
-			qc->nsect = 256;
-			if (lba48)
-				tf->hob_nsect = 1;
-		}
+	return 0;
 
-		tf->lbal = scsicmd[3];
-		tf->lbam = scsicmd[2];
-		tf->lbah = scsicmd[1] & 0x1f; /* mask out reserved bits */
+invalid_fld:
+	ata_scsi_set_sense(qc->scsicmd, ILLEGAL_REQUEST, 0x24, 0x0);
+	/* "Invalid field in cbd" */
+	return 1;
 
-		VPRINTK("six-byte command\n");
-		return 0;
-	}
+out_of_range:
+	ata_scsi_set_sense(qc->scsicmd, ILLEGAL_REQUEST, 0x21, 0x0);
+	/* "Logical Block Address out of range" */
+	return 1;
 
-	if (scsicmd[0] == READ_16 || scsicmd[0] == WRITE_16) {
-		/* rule out impossible LBAs and sector counts */
-		if (scsicmd[2] || scsicmd[3] || scsicmd[10] || scsicmd[11])
-			return 1;
-
-		if (lba48) {
-			tf->hob_nsect = scsicmd[12];
-			tf->hob_lbal = scsicmd[6];
-			tf->hob_lbam = scsicmd[5];
-			tf->hob_lbah = scsicmd[4];
-
-			qc->nsect = ((unsigned int)scsicmd[12] << 8) |
-					scsicmd[13];
-		} else {
-			/* once again, filter out impossible non-zero values */
-			if (scsicmd[4] || scsicmd[5] || scsicmd[12] ||
-			    (scsicmd[6] & 0xf0))
-				return 1;
-
-			/* stores LBA27:24 in lower 4 bits of device reg */
-			tf->device |= scsicmd[6];
-
-			qc->nsect = scsicmd[13];
-		}
-
-		tf->nsect = scsicmd[13];
-		tf->lbal = scsicmd[9];
-		tf->lbam = scsicmd[8];
-		tf->lbah = scsicmd[7];
-
-		VPRINTK("sixteen-byte command\n");
-		if (qc->nsect == 0) /* we don't support length==0 cmds */
-			return 1;
-		return 0;
-	}
-
-	DPRINTK("no-byte command\n");
+nothing_to_do:
+	qc->scsicmd->result = SAM_STAT_GOOD;
 	return 1;
 }
 
@@ -741,6 +900,12 @@
  *	This function sets up an ata_queued_cmd structure for the
  *	SCSI command, and sends that ata_queued_cmd to the hardware.
  *
+ *	The xlat_func argument (actor) returns 0 if ready to execute
+ *	ATA command, else 1 to finish translation. If 1 is returned
+ *	then cmd->result (and possibly cmd->sense_buffer) are assumed
+ *	to be set reflecting an error condition or clean (early)
+ *	termination.
+ *
  *	LOCKING:
  *	spin_lock_irqsave(host_set lock)
  */
@@ -757,7 +922,7 @@
 
 	qc = ata_scsi_qc_new(ap, dev, cmd, done);
 	if (!qc)
-		return;
+		goto err_mem;
 
 	/* data is present; dma-map it */
 	if (cmd->sc_data_direction == DMA_FROM_DEVICE ||
@@ -765,7 +930,7 @@
 		if (unlikely(cmd->request_bufflen < 1)) {
 			printk(KERN_WARNING "ata%u(%u): WARNING: zero len r/w req\n",
 			       ap->id, dev->devno);
-			goto err_out;
+			goto err_did;
 		}
 
 		if (cmd->use_sg)
@@ -780,19 +945,28 @@
 	qc->complete_fn = ata_scsi_qc_complete;
 
 	if (xlat_func(qc, scsicmd))
-		goto err_out;
+		goto early_finish;
 
 	/* select device, send command to hardware */
 	if (ata_qc_issue(qc))
-		goto err_out;
+		goto err_did;
 
 	VPRINTK("EXIT\n");
 	return;
 
-err_out:
+early_finish:
+        ata_qc_free(qc);
+	done(cmd);
+	DPRINTK("EXIT - early finish (good or error)\n");
+	return;
+
+err_did:
 	ata_qc_free(qc);
-	ata_bad_cdb(cmd, done);
-	DPRINTK("EXIT - badcmd\n");
+err_mem:
+	cmd->result = (DID_ERROR << 16);
+	done(cmd);
+	DPRINTK("EXIT - internal\n");
+	return;
 }
 
 /**
@@ -859,7 +1033,8 @@
  *	Mapping the response buffer, calling the command's handler,
  *	and handling the handler's return value.  This return value
  *	indicates whether the handler wishes the SCSI command to be
- *	completed successfully, or not.
+ *	completed successfully (0), or not (in which case cmd->result
+ *	and sense buffer are assumed to be set).
  *
  *	LOCKING:
  *	spin_lock_irqsave(host_set lock)
@@ -878,12 +1053,9 @@
 	rc = actor(args, rbuf, buflen);
 	ata_scsi_rbuf_put(cmd, rbuf);
 
-	if (rc)
-		ata_bad_cdb(cmd, args->done);
-	else {
+	if (rc == 0)
 		cmd->result = SAM_STAT_GOOD;
-		args->done(cmd);
-	}
+	args->done(cmd);
 }
 
 /**
@@ -1189,8 +1361,16 @@
 	 * in the same manner)
 	 */
 	page_control = scsicmd[2] >> 6;
-	if ((page_control != 0) && (page_control != 3))
-		return 1;
+	switch (page_control) {
+	case 0: /* current */
+		break;  /* supported */
+	case 3: /* saved */
+		goto saving_not_supp;
+	case 1: /* changeable */
+	case 2: /* defaults */
+	default:
+		goto invalid_fld;
+	}
 
 	if (six_byte)
 		output_len = 4;
@@ -1221,7 +1401,7 @@
 		break;
 
 	default:		/* invalid page code */
-		return 1;
+		goto invalid_fld;
 	}
 
 	if (six_byte) {
@@ -1234,6 +1414,16 @@
 	}
 
 	return 0;
+
+invalid_fld:
+	ata_scsi_set_sense(args->cmd, ILLEGAL_REQUEST, 0x24, 0x0);
+	/* "Invalid field in cbd" */
+	return 1;
+
+saving_not_supp:
+	ata_scsi_set_sense(args->cmd, ILLEGAL_REQUEST, 0x39, 0x0);
+	 /* "Saving parameters not supported" */
+	return 1;
 }
 
 /**
@@ -1256,10 +1446,20 @@
 
 	VPRINTK("ENTER\n");
 
-	if (ata_id_has_lba48(args->id))
-		n_sectors = ata_id_u64(args->id, 100);
-	else
-		n_sectors = ata_id_u32(args->id, 60);
+	if (ata_id_has_lba(args->id)) {
+		if (ata_id_has_lba48(args->id))
+			n_sectors = ata_id_u64(args->id, 100);
+		else
+			n_sectors = ata_id_u32(args->id, 60);
+	} else {
+		/* CHS default translation */
+		n_sectors = args->id[1] * args->id[3] * args->id[6];
+
+		if (ata_id_current_chs_valid(args->id))
+			/* CHS current translation */
+			n_sectors = ata_id_u32(args->id, 57);
+	}
+
 	n_sectors--;		/* ATA TotalUserSectors - 1 */
 
 	if (args->cmd->cmnd[0] == READ_CAPACITY) {
@@ -1323,6 +1523,34 @@
 }
 
 /**
+ *	ata_scsi_set_sense - Set SCSI sense data and status
+ *	@cmd: SCSI request to be handled
+ *	@sk: SCSI-defined sense key
+ *	@asc: SCSI-defined additional sense code
+ *	@ascq: SCSI-defined additional sense code qualifier
+ *
+ *	Helper function that builds a valid fixed format, current
+ *	response code and the given sense key (sk), additional sense
+ *	code (asc) and additional sense code qualifier (ascq) with
+ *	a SCSI command status of %SAM_STAT_CHECK_CONDITION and
+ *	DRIVER_SENSE set in the upper bits of scsi_cmnd::result .
+ *
+ *	LOCKING:
+ *	Not required
+ */
+
+void ata_scsi_set_sense(struct scsi_cmnd *cmd, u8 sk, u8 asc, u8 ascq)
+{
+	cmd->result = (DRIVER_SENSE << 24) | SAM_STAT_CHECK_CONDITION;
+
+	cmd->sense_buffer[0] = 0x70;	/* fixed format, current */
+	cmd->sense_buffer[2] = sk;
+	cmd->sense_buffer[7] = 18 - 8;	/* additional sense length */
+	cmd->sense_buffer[12] = asc;
+	cmd->sense_buffer[13] = ascq;
+}
+
+/**
  *	ata_scsi_badcmd - End a SCSI request with an error
  *	@cmd: SCSI request to be handled
  *	@done: SCSI command completion function
@@ -1340,30 +1568,84 @@
 void ata_scsi_badcmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *), u8 asc, u8 ascq)
 {
 	DPRINTK("ENTER\n");
-	cmd->result = SAM_STAT_CHECK_CONDITION;
-
-	cmd->sense_buffer[0] = 0x70;
-	cmd->sense_buffer[2] = ILLEGAL_REQUEST;
-	cmd->sense_buffer[7] = 14 - 8;	/* addnl. sense len. FIXME: correct? */
-	cmd->sense_buffer[12] = asc;
-	cmd->sense_buffer[13] = ascq;
+	ata_scsi_set_sense(cmd, ILLEGAL_REQUEST, asc, ascq);
 
 	done(cmd);
 }
 
+void atapi_request_sense(struct ata_port *ap, struct ata_device *dev,
+			 struct scsi_cmnd *cmd)
+{
+	DECLARE_COMPLETION(wait);
+	struct ata_queued_cmd *qc;
+	unsigned long flags;
+	int rc;
+
+	DPRINTK("ATAPI request sense\n");
+
+	qc = ata_qc_new_init(ap, dev);
+	BUG_ON(qc == NULL);
+
+	/* FIXME: is this needed? */
+	memset(cmd->sense_buffer, 0, sizeof(cmd->sense_buffer));
+
+	ata_sg_init_one(qc, cmd->sense_buffer, sizeof(cmd->sense_buffer));
+	qc->dma_dir = DMA_FROM_DEVICE;
+
+	memset(&qc->cdb, 0, ap->cdb_len);
+	qc->cdb[0] = REQUEST_SENSE;
+	qc->cdb[4] = SCSI_SENSE_BUFFERSIZE;
+
+	qc->tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
+	qc->tf.command = ATA_CMD_PACKET;
+
+	qc->tf.protocol = ATA_PROT_ATAPI;
+	qc->tf.lbam = (8 * 1024) & 0xff;
+	qc->tf.lbah = (8 * 1024) >> 8;
+	qc->nbytes = SCSI_SENSE_BUFFERSIZE;
+
+	qc->waiting = &wait;
+	qc->complete_fn = ata_qc_complete_noop;
+
+	spin_lock_irqsave(&ap->host_set->lock, flags);
+	rc = ata_qc_issue(qc);
+	spin_unlock_irqrestore(&ap->host_set->lock, flags);
+
+	if (rc)
+		ata_port_disable(ap);
+	else
+		wait_for_completion(&wait);
+
+	DPRINTK("EXIT\n");
+}
+
 static int atapi_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat)
 {
 	struct scsi_cmnd *cmd = qc->scsicmd;
 
-	if (unlikely(drv_stat & (ATA_ERR | ATA_BUSY | ATA_DRQ))) {
+	VPRINTK("ENTER, drv_stat == 0x%x\n", drv_stat);
+
+	if (unlikely(drv_stat & (ATA_BUSY | ATA_DRQ)))
+		ata_to_sense_error(qc, drv_stat);
+
+	else if (unlikely(drv_stat & ATA_ERR)) {
 		DPRINTK("request check condition\n");
 
+		/* FIXME: command completion with check condition
+		 * but no sense causes the error handler to run,
+		 * which then issues REQUEST SENSE, fills in the sense 
+		 * buffer, and completes the command (for the second
+		 * time).  We need to issue REQUEST SENSE some other
+		 * way, to avoid completing the command twice.
+		 */
 		cmd->result = SAM_STAT_CHECK_CONDITION;
 
 		qc->scsidone(cmd);
 
 		return 1;
-	} else {
+	}
+
+	else {
 		u8 *scsicmd = cmd->cmnd;
 
 		if (scsicmd[0] == INQUIRY) {
@@ -1371,15 +1653,30 @@
 			unsigned int buflen;
 
 			buflen = ata_scsi_rbuf_get(cmd, &buf);
-			buf[2] = 0x5;
-			buf[3] = (buf[3] & 0xf0) | 2;
+
+	/* ATAPI devices typically report zero for their SCSI version,
+	 * and sometimes deviate from the spec WRT response data
+	 * format.  If SCSI version is reported as zero like normal,
+	 * then we make the following fixups:  1) Fake MMC-5 version,
+	 * to indicate to the Linux scsi midlayer this is a modern
+	 * device.  2) Ensure response data format / ATAPI information
+	 * are always correct.
+	 */
+	/* FIXME: do we ever override EVPD pages and the like, with
+	 * this code?
+	 */
+			if (buf[2] == 0) {
+				buf[2] = 0x5;
+				buf[3] = 0x32;
+			}
+
 			ata_scsi_rbuf_put(cmd, buf);
 		}
+
 		cmd->result = SAM_STAT_GOOD;
 	}
 
 	qc->scsidone(cmd);
-
 	return 0;
 }
 /**
@@ -1640,7 +1937,7 @@
 
 		case INQUIRY:
 			if (scsicmd[1] & 2)	           /* is CmdDt set?  */
-				ata_bad_cdb(cmd, done);
+				ata_scsi_invalid_field(cmd, done);
 			else if ((scsicmd[1] & 1) == 0)    /* is EVPD clear? */
 				ata_scsi_rbuf_fill(&args, ata_scsiop_inq_std);
 			else if (scsicmd[2] == 0x00)
@@ -1650,7 +1947,7 @@
 			else if (scsicmd[2] == 0x83)
 				ata_scsi_rbuf_fill(&args, ata_scsiop_inq_83);
 			else
-				ata_bad_cdb(cmd, done);
+				ata_scsi_invalid_field(cmd, done);
 			break;
 
 		case MODE_SENSE:
@@ -1660,7 +1957,7 @@
 
 		case MODE_SELECT:	/* unconditionally return */
 		case MODE_SELECT_10:	/* bad-field-in-cdb */
-			ata_bad_cdb(cmd, done);
+			ata_scsi_invalid_field(cmd, done);
 			break;
 
 		case READ_CAPACITY:
@@ -1671,7 +1968,7 @@
 			if ((scsicmd[1] & 0x1f) == SAI_READ_CAPACITY_16)
 				ata_scsi_rbuf_fill(&args, ata_scsiop_read_cap);
 			else
-				ata_bad_cdb(cmd, done);
+				ata_scsi_invalid_field(cmd, done);
 			break;
 
 		case REPORT_LUNS:
@@ -1683,8 +1980,26 @@
 
 		/* all other commands */
 		default:
-			ata_bad_scsiop(cmd, done);
+			ata_scsi_set_sense(cmd, ILLEGAL_REQUEST, 0x20, 0x0);
+			/* "Invalid command operation code" */
+			done(cmd);
 			break;
 	}
 }
 
+void ata_scsi_scan_host(struct ata_port *ap)
+{
+	struct ata_device *dev;
+	unsigned int i;
+
+	if (ap->flags & ATA_FLAG_PORT_DISABLED)
+		return;
+
+	for (i = 0; i < ATA_MAX_DEVICES; i++) {
+		dev = &ap->device[i];
+
+		if (ata_dev_present(dev))
+			scsi_scan_target(&ap->host->shost_gendev, 0, i, 0, 0);
+	}
+}
+
diff --git a/drivers/scsi/libata.h b/drivers/scsi/libata.h
index d608b3a..a18f2ac 100644
--- a/drivers/scsi/libata.h
+++ b/drivers/scsi/libata.h
@@ -39,6 +39,7 @@
 
 /* libata-core.c */
 extern int atapi_enabled;
+extern int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat);
 extern struct ata_queued_cmd *ata_qc_new_init(struct ata_port *ap,
 				      struct ata_device *dev);
 extern void ata_qc_free(struct ata_queued_cmd *qc);
@@ -51,6 +52,9 @@
 
 
 /* libata-scsi.c */
+extern void atapi_request_sense(struct ata_port *ap, struct ata_device *dev,
+			 struct scsi_cmnd *cmd);
+extern void ata_scsi_scan_host(struct ata_port *ap);
 extern void ata_to_sense_error(struct ata_queued_cmd *qc, u8 drv_stat);
 extern int ata_scsi_error(struct Scsi_Host *host);
 extern unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf,
@@ -76,18 +80,10 @@
 extern void ata_scsi_badcmd(struct scsi_cmnd *cmd,
 			    void (*done)(struct scsi_cmnd *),
 			    u8 asc, u8 ascq);
+extern void ata_scsi_set_sense(struct scsi_cmnd *cmd,
+			       u8 sk, u8 asc, u8 ascq);
 extern void ata_scsi_rbuf_fill(struct ata_scsi_args *args,
                         unsigned int (*actor) (struct ata_scsi_args *args,
                                            u8 *rbuf, unsigned int buflen));
 
-static inline void ata_bad_scsiop(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
-{
-	ata_scsi_badcmd(cmd, done, 0x20, 0x00);
-}
-
-static inline void ata_bad_cdb(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
-{
-	ata_scsi_badcmd(cmd, done, 0x24, 0x00);
-}
-
 #endif /* __LIBATA_H__ */
diff --git a/drivers/scsi/megaraid/megaraid_sas.c b/drivers/scsi/megaraid/megaraid_sas.c
index 1b3148e..c3f6373 100644
--- a/drivers/scsi/megaraid/megaraid_sas.c
+++ b/drivers/scsi/megaraid/megaraid_sas.c
@@ -34,6 +34,7 @@
 #include <linux/delay.h>
 #include <linux/uio.h>
 #include <asm/uaccess.h>
+#include <linux/fs.h>
 #include <linux/compat.h>
 
 #include <scsi/scsi.h>
diff --git a/drivers/scsi/sata_mv.c b/drivers/scsi/sata_mv.c
index ea76fe4..d457f56 100644
--- a/drivers/scsi/sata_mv.c
+++ b/drivers/scsi/sata_mv.c
@@ -35,7 +35,7 @@
 #include <asm/io.h>
 
 #define DRV_NAME	"sata_mv"
-#define DRV_VERSION	"0.12"
+#define DRV_VERSION	"0.24"
 
 enum {
 	/* BAR's are enumerated in terms of pci_resource_start() terms */
@@ -55,31 +55,61 @@
 	MV_SATAHC_ARBTR_REG_SZ	= MV_MINOR_REG_AREA_SZ,		/* arbiter */
 	MV_PORT_REG_SZ		= MV_MINOR_REG_AREA_SZ,
 
-	MV_Q_CT			= 32,
-	MV_CRQB_SZ		= 32,
-	MV_CRPB_SZ		= 8,
+	MV_USE_Q_DEPTH		= ATA_DEF_QUEUE,
 
-	MV_DMA_BOUNDARY		= 0xffffffffU,
-	SATAHC_MASK		= (~(MV_SATAHC_REG_SZ - 1)),
+	MV_MAX_Q_DEPTH		= 32,
+	MV_MAX_Q_DEPTH_MASK	= MV_MAX_Q_DEPTH - 1,
+
+	/* CRQB needs alignment on a 1KB boundary. Size == 1KB
+	 * CRPB needs alignment on a 256B boundary. Size == 256B
+	 * SG count of 176 leads to MV_PORT_PRIV_DMA_SZ == 4KB
+	 * ePRD (SG) entries need alignment on a 16B boundary. Size == 16B
+	 */
+	MV_CRQB_Q_SZ		= (32 * MV_MAX_Q_DEPTH),
+	MV_CRPB_Q_SZ		= (8 * MV_MAX_Q_DEPTH),
+	MV_MAX_SG_CT		= 176,
+	MV_SG_TBL_SZ		= (16 * MV_MAX_SG_CT),
+	MV_PORT_PRIV_DMA_SZ	= (MV_CRQB_Q_SZ + MV_CRPB_Q_SZ + MV_SG_TBL_SZ),
+
+	/* Our DMA boundary is determined by an ePRD being unable to handle
+	 * anything larger than 64KB
+	 */
+	MV_DMA_BOUNDARY		= 0xffffU,
 
 	MV_PORTS_PER_HC		= 4,
 	/* == (port / MV_PORTS_PER_HC) to determine HC from 0-7 port */
 	MV_PORT_HC_SHIFT	= 2,
-	/* == (port % MV_PORTS_PER_HC) to determine port from 0-7 port */
+	/* == (port % MV_PORTS_PER_HC) to determine hard port from 0-7 port */
 	MV_PORT_MASK		= 3,
 
 	/* Host Flags */
 	MV_FLAG_DUAL_HC		= (1 << 30),  /* two SATA Host Controllers */
 	MV_FLAG_IRQ_COALESCE	= (1 << 29),  /* IRQ coalescing capability */
-	MV_FLAG_BDMA		= (1 << 28),  /* Basic DMA */
+	MV_FLAG_GLBL_SFT_RST	= (1 << 28),  /* Global Soft Reset support */
+	MV_COMMON_FLAGS		= (ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+				   ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO),
+	MV_6XXX_FLAGS		= (MV_FLAG_IRQ_COALESCE | 
+				   MV_FLAG_GLBL_SFT_RST),
 
 	chip_504x		= 0,
 	chip_508x		= 1,
 	chip_604x		= 2,
 	chip_608x		= 3,
 
+	CRQB_FLAG_READ		= (1 << 0),
+	CRQB_TAG_SHIFT		= 1,
+	CRQB_CMD_ADDR_SHIFT	= 8,
+	CRQB_CMD_CS		= (0x2 << 11),
+	CRQB_CMD_LAST		= (1 << 15),
+
+	CRPB_FLAG_STATUS_SHIFT	= 8,
+
+	EPRD_FLAG_END_OF_TBL	= (1 << 31),
+
 	/* PCI interface registers */
 
+	PCI_COMMAND_OFS		= 0xc00,
+
 	PCI_MAIN_CMD_STS_OFS	= 0xd30,
 	STOP_PCI_MASTER		= (1 << 2),
 	PCI_MASTER_EMPTY	= (1 << 3),
@@ -111,20 +141,13 @@
 	HC_CFG_OFS		= 0,
 
 	HC_IRQ_CAUSE_OFS	= 0x14,
-	CRBP_DMA_DONE		= (1 << 0),	/* shift by port # */
+	CRPB_DMA_DONE		= (1 << 0),	/* shift by port # */
 	HC_IRQ_COAL		= (1 << 4),	/* IRQ coalescing */
 	DEV_IRQ			= (1 << 8),	/* shift by port # */
 
 	/* Shadow block registers */
-	SHD_PIO_DATA_OFS	= 0x100,
-	SHD_FEA_ERR_OFS		= 0x104,
-	SHD_SECT_CNT_OFS	= 0x108,
-	SHD_LBA_L_OFS		= 0x10C,
-	SHD_LBA_M_OFS		= 0x110,
-	SHD_LBA_H_OFS		= 0x114,
-	SHD_DEV_HD_OFS		= 0x118,
-	SHD_CMD_STA_OFS		= 0x11C,
-	SHD_CTL_AST_OFS		= 0x120,
+	SHD_BLK_OFS		= 0x100,
+	SHD_CTL_AST_OFS		= 0x20,		/* ofs from SHD_BLK_OFS */
 
 	/* SATA registers */
 	SATA_STATUS_OFS		= 0x300,  /* ctrl, err regs follow status */
@@ -132,6 +155,11 @@
 
 	/* Port registers */
 	EDMA_CFG_OFS		= 0,
+	EDMA_CFG_Q_DEPTH	= 0,			/* queueing disabled */
+	EDMA_CFG_NCQ		= (1 << 5),
+	EDMA_CFG_NCQ_GO_ON_ERR	= (1 << 14),		/* continue on error */
+	EDMA_CFG_RD_BRST_EXT	= (1 << 11),		/* read burst 512B */
+	EDMA_CFG_WR_BUFF_LEN	= (1 << 13),		/* write buffer 512B */
 
 	EDMA_ERR_IRQ_CAUSE_OFS	= 0x8,
 	EDMA_ERR_IRQ_MASK_OFS	= 0xc,
@@ -161,33 +189,85 @@
 				   EDMA_ERR_LNK_DATA_TX | 
 				   EDMA_ERR_TRANS_PROTO),
 
+	EDMA_REQ_Q_BASE_HI_OFS	= 0x10,
+	EDMA_REQ_Q_IN_PTR_OFS	= 0x14,		/* also contains BASE_LO */
+	EDMA_REQ_Q_BASE_LO_MASK	= 0xfffffc00U,
+
+	EDMA_REQ_Q_OUT_PTR_OFS	= 0x18,
+	EDMA_REQ_Q_PTR_SHIFT	= 5,
+
+	EDMA_RSP_Q_BASE_HI_OFS	= 0x1c,
+	EDMA_RSP_Q_IN_PTR_OFS	= 0x20,
+	EDMA_RSP_Q_OUT_PTR_OFS	= 0x24,		/* also contains BASE_LO */
+	EDMA_RSP_Q_BASE_LO_MASK	= 0xffffff00U,
+	EDMA_RSP_Q_PTR_SHIFT	= 3,
+
 	EDMA_CMD_OFS		= 0x28,
 	EDMA_EN			= (1 << 0),
 	EDMA_DS			= (1 << 1),
 	ATA_RST			= (1 << 2),
 
-	/* BDMA is 6xxx part only */
-	BDMA_CMD_OFS		= 0x224,
-	BDMA_START		= (1 << 0),
+	/* Host private flags (hp_flags) */
+	MV_HP_FLAG_MSI		= (1 << 0),
 
-	MV_UNDEF		= 0,
+	/* Port private flags (pp_flags) */
+	MV_PP_FLAG_EDMA_EN	= (1 << 0),
+	MV_PP_FLAG_EDMA_DS_ACT	= (1 << 1),
+};
+
+/* Command ReQuest Block: 32B */
+struct mv_crqb {
+	u32			sg_addr;
+	u32			sg_addr_hi;
+	u16			ctrl_flags;
+	u16			ata_cmd[11];
+};
+
+/* Command ResPonse Block: 8B */
+struct mv_crpb {
+	u16			id;
+	u16			flags;
+	u32			tmstmp;
+};
+
+/* EDMA Physical Region Descriptor (ePRD); A.K.A. SG */
+struct mv_sg {
+	u32			addr;
+	u32			flags_size;
+	u32			addr_hi;
+	u32			reserved;
 };
 
 struct mv_port_priv {
+	struct mv_crqb		*crqb;
+	dma_addr_t		crqb_dma;
+	struct mv_crpb		*crpb;
+	dma_addr_t		crpb_dma;
+	struct mv_sg		*sg_tbl;
+	dma_addr_t		sg_tbl_dma;
 
+	unsigned		req_producer;		/* cp of req_in_ptr */
+	unsigned		rsp_consumer;		/* cp of rsp_out_ptr */
+	u32			pp_flags;
 };
 
 struct mv_host_priv {
-
+	u32			hp_flags;
 };
 
 static void mv_irq_clear(struct ata_port *ap);
 static u32 mv_scr_read(struct ata_port *ap, unsigned int sc_reg_in);
 static void mv_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val);
+static u8 mv_check_err(struct ata_port *ap);
 static void mv_phy_reset(struct ata_port *ap);
-static int mv_master_reset(void __iomem *mmio_base);
+static void mv_host_stop(struct ata_host_set *host_set);
+static int mv_port_start(struct ata_port *ap);
+static void mv_port_stop(struct ata_port *ap);
+static void mv_qc_prep(struct ata_queued_cmd *qc);
+static int mv_qc_issue(struct ata_queued_cmd *qc);
 static irqreturn_t mv_interrupt(int irq, void *dev_instance,
 				struct pt_regs *regs);
+static void mv_eng_timeout(struct ata_port *ap);
 static int mv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
 
 static Scsi_Host_Template mv_sht = {
@@ -196,13 +276,13 @@
 	.ioctl			= ata_scsi_ioctl,
 	.queuecommand		= ata_scsi_queuecmd,
 	.eh_strategy_handler	= ata_scsi_error,
-	.can_queue		= ATA_DEF_QUEUE,
+	.can_queue		= MV_USE_Q_DEPTH,
 	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= MV_UNDEF,
+	.sg_tablesize		= MV_MAX_SG_CT,
 	.max_sectors		= ATA_MAX_SECTORS,
 	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
 	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= MV_UNDEF,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
 	.proc_name		= DRV_NAME,
 	.dma_boundary		= MV_DMA_BOUNDARY,
 	.slave_configure	= ata_scsi_slave_config,
@@ -216,15 +296,16 @@
 	.tf_load		= ata_tf_load,
 	.tf_read		= ata_tf_read,
 	.check_status		= ata_check_status,
+	.check_err		= mv_check_err,
 	.exec_command		= ata_exec_command,
 	.dev_select		= ata_std_dev_select,
 
 	.phy_reset		= mv_phy_reset,
 
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
+	.qc_prep		= mv_qc_prep,
+	.qc_issue		= mv_qc_issue,
 
-	.eng_timeout		= ata_eng_timeout,
+	.eng_timeout		= mv_eng_timeout,
 
 	.irq_handler		= mv_interrupt,
 	.irq_clear		= mv_irq_clear,
@@ -232,46 +313,39 @@
 	.scr_read		= mv_scr_read,
 	.scr_write		= mv_scr_write,
 
-	.port_start		= ata_port_start,
-	.port_stop		= ata_port_stop,
-	.host_stop		= ata_host_stop,
+	.port_start		= mv_port_start,
+	.port_stop		= mv_port_stop,
+	.host_stop		= mv_host_stop,
 };
 
 static struct ata_port_info mv_port_info[] = {
 	{  /* chip_504x */
 		.sht		= &mv_sht,
-		.host_flags	= (ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
-				   ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO),
-		.pio_mask	= 0x1f,	/* pio4-0 */
-		.udma_mask	= 0,	/* 0x7f (udma6-0 disabled for now) */
+		.host_flags	= MV_COMMON_FLAGS,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.udma_mask	= 0,	/* 0x7f (udma0-6 disabled for now) */
 		.port_ops	= &mv_ops,
 	},
 	{  /* chip_508x */
 		.sht		= &mv_sht,
-		.host_flags	= (ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
-				   ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO | 
-				   MV_FLAG_DUAL_HC),
-		.pio_mask	= 0x1f,	/* pio4-0 */
-		.udma_mask	= 0,	/* 0x7f (udma6-0 disabled for now) */
+		.host_flags	= (MV_COMMON_FLAGS | MV_FLAG_DUAL_HC),
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.udma_mask	= 0,	/* 0x7f (udma0-6 disabled for now) */
 		.port_ops	= &mv_ops,
 	},
 	{  /* chip_604x */
 		.sht		= &mv_sht,
-		.host_flags	= (ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
-				   ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO | 
-				   MV_FLAG_IRQ_COALESCE | MV_FLAG_BDMA),
-		.pio_mask	= 0x1f,	/* pio4-0 */
-		.udma_mask	= 0,	/* 0x7f (udma6-0 disabled for now) */
+		.host_flags	= (MV_COMMON_FLAGS | MV_6XXX_FLAGS),
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv_ops,
 	},
 	{  /* chip_608x */
 		.sht		= &mv_sht,
-		.host_flags	= (ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
-				   ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO |
-				   MV_FLAG_IRQ_COALESCE | MV_FLAG_DUAL_HC |
-				   MV_FLAG_BDMA),
-		.pio_mask	= 0x1f,	/* pio4-0 */
-		.udma_mask	= 0,	/* 0x7f (udma6-0 disabled for now) */
+		.host_flags	= (MV_COMMON_FLAGS | MV_6XXX_FLAGS | 
+				   MV_FLAG_DUAL_HC),
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv_ops,
 	},
 };
@@ -306,12 +380,6 @@
 	(void) readl(addr);	/* flush to avoid PCI posted write */
 }
 
-static inline void __iomem *mv_port_addr_to_hc_base(void __iomem *port_mmio)
-{
-	return ((void __iomem *)((unsigned long)port_mmio & 
-				 (unsigned long)SATAHC_MASK));
-}
-
 static inline void __iomem *mv_hc_base(void __iomem *base, unsigned int hc)
 {
 	return (base + MV_SATAHC0_REG_BASE + (hc * MV_SATAHC_REG_SZ));
@@ -329,26 +397,152 @@
 	return mv_port_base(ap->host_set->mmio_base, ap->port_no);
 }
 
-static inline int mv_get_hc_count(unsigned long flags)
+static inline int mv_get_hc_count(unsigned long hp_flags)
 {
-	return ((flags & MV_FLAG_DUAL_HC) ? 2 : 1);
-}
-
-static inline int mv_is_edma_active(struct ata_port *ap)
-{
-	void __iomem *port_mmio = mv_ap_base(ap);
-	return (EDMA_EN & readl(port_mmio + EDMA_CMD_OFS));
-}
-
-static inline int mv_port_bdma_capable(struct ata_port *ap)
-{
-	return (ap->flags & MV_FLAG_BDMA);
+	return ((hp_flags & MV_FLAG_DUAL_HC) ? 2 : 1);
 }
 
 static void mv_irq_clear(struct ata_port *ap)
 {
 }
 
+/**
+ *      mv_start_dma - Enable eDMA engine
+ *      @base: port base address
+ *      @pp: port private data
+ *
+ *      Verify the local cache of the eDMA state is accurate with an
+ *      assert.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_start_dma(void __iomem *base, struct mv_port_priv *pp)
+{
+	if (!(MV_PP_FLAG_EDMA_EN & pp->pp_flags)) {
+		writelfl(EDMA_EN, base + EDMA_CMD_OFS);
+		pp->pp_flags |= MV_PP_FLAG_EDMA_EN;
+	}
+	assert(EDMA_EN & readl(base + EDMA_CMD_OFS));
+}
+
+/**
+ *      mv_stop_dma - Disable eDMA engine
+ *      @ap: ATA channel to manipulate
+ *
+ *      Verify the local cache of the eDMA state is accurate with an
+ *      assert.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_stop_dma(struct ata_port *ap)
+{
+	void __iomem *port_mmio = mv_ap_base(ap);
+	struct mv_port_priv *pp	= ap->private_data;
+	u32 reg;
+	int i;
+
+	if (MV_PP_FLAG_EDMA_EN & pp->pp_flags) {
+		/* Disable EDMA if active.   The disable bit auto clears.
+		 */
+		writelfl(EDMA_DS, port_mmio + EDMA_CMD_OFS);
+		pp->pp_flags &= ~MV_PP_FLAG_EDMA_EN;
+	} else {
+		assert(!(EDMA_EN & readl(port_mmio + EDMA_CMD_OFS)));
+  	}
+	
+	/* now properly wait for the eDMA to stop */
+	for (i = 1000; i > 0; i--) {
+		reg = readl(port_mmio + EDMA_CMD_OFS);
+		if (!(EDMA_EN & reg)) {
+			break;
+		}
+		udelay(100);
+	}
+
+	if (EDMA_EN & reg) {
+		printk(KERN_ERR "ata%u: Unable to stop eDMA\n", ap->id);
+		/* FIXME: Consider doing a reset here to recover */
+	}
+}
+
+#ifdef ATA_DEBUG
+static void mv_dump_mem(void __iomem *start, unsigned bytes)
+{
+	int b, w;
+	for (b = 0; b < bytes; ) {
+		DPRINTK("%p: ", start + b);
+		for (w = 0; b < bytes && w < 4; w++) {
+			printk("%08x ",readl(start + b));
+			b += sizeof(u32);
+		}
+		printk("\n");
+	}
+}
+#endif
+
+static void mv_dump_pci_cfg(struct pci_dev *pdev, unsigned bytes)
+{
+#ifdef ATA_DEBUG
+	int b, w;
+	u32 dw;
+	for (b = 0; b < bytes; ) {
+		DPRINTK("%02x: ", b);
+		for (w = 0; b < bytes && w < 4; w++) {
+			(void) pci_read_config_dword(pdev,b,&dw);
+			printk("%08x ",dw);
+			b += sizeof(u32);
+		}
+		printk("\n");
+	}
+#endif
+}
+static void mv_dump_all_regs(void __iomem *mmio_base, int port,
+			     struct pci_dev *pdev)
+{
+#ifdef ATA_DEBUG
+	void __iomem *hc_base = mv_hc_base(mmio_base, 
+					   port >> MV_PORT_HC_SHIFT);
+	void __iomem *port_base;
+	int start_port, num_ports, p, start_hc, num_hcs, hc;
+
+	if (0 > port) {
+		start_hc = start_port = 0;
+		num_ports = 8;		/* shld be benign for 4 port devs */
+		num_hcs = 2;
+	} else {
+		start_hc = port >> MV_PORT_HC_SHIFT;
+		start_port = port;
+		num_ports = num_hcs = 1;
+	}
+	DPRINTK("All registers for port(s) %u-%u:\n", start_port, 
+		num_ports > 1 ? num_ports - 1 : start_port);
+
+	if (NULL != pdev) {
+		DPRINTK("PCI config space regs:\n");
+		mv_dump_pci_cfg(pdev, 0x68);
+	}
+	DPRINTK("PCI regs:\n");
+	mv_dump_mem(mmio_base+0xc00, 0x3c);
+	mv_dump_mem(mmio_base+0xd00, 0x34);
+	mv_dump_mem(mmio_base+0xf00, 0x4);
+	mv_dump_mem(mmio_base+0x1d00, 0x6c);
+	for (hc = start_hc; hc < start_hc + num_hcs; hc++) {
+		hc_base = mv_hc_base(mmio_base, port >> MV_PORT_HC_SHIFT);
+		DPRINTK("HC regs (HC %i):\n", hc);
+		mv_dump_mem(hc_base, 0x1c);
+	}
+	for (p = start_port; p < start_port + num_ports; p++) {
+		port_base = mv_port_base(mmio_base, p);
+		DPRINTK("EDMA regs (port %i):\n",p);
+		mv_dump_mem(port_base, 0x54);
+		DPRINTK("SATA regs (port %i):\n",p);
+		mv_dump_mem(port_base+0x300, 0x60);
+	}
+#endif
+}
+
 static unsigned int mv_scr_offset(unsigned int sc_reg_in)
 {
 	unsigned int ofs;
@@ -389,30 +583,37 @@
 	}
 }
 
-static int mv_master_reset(void __iomem *mmio_base)
+/**
+ *      mv_global_soft_reset - Perform the 6xxx global soft reset
+ *      @mmio_base: base address of the HBA
+ *
+ *      This routine only applies to 6xxx parts.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static int mv_global_soft_reset(void __iomem *mmio_base)
 {
 	void __iomem *reg = mmio_base + PCI_MAIN_CMD_STS_OFS;
 	int i, rc = 0;
 	u32 t;
 
-	VPRINTK("ENTER\n");
-
 	/* Following procedure defined in PCI "main command and status
 	 * register" table.
 	 */
 	t = readl(reg);
 	writel(t | STOP_PCI_MASTER, reg);
 
-	for (i = 0; i < 100; i++) {
-		msleep(10);
+	for (i = 0; i < 1000; i++) {
+		udelay(1);
 		t = readl(reg);
 		if (PCI_MASTER_EMPTY & t) {
 			break;
 		}
 	}
 	if (!(PCI_MASTER_EMPTY & t)) {
-		printk(KERN_ERR DRV_NAME "PCI master won't flush\n");
-		rc = 1;		/* broken HW? */
+		printk(KERN_ERR DRV_NAME ": PCI master won't flush\n");
+		rc = 1;
 		goto done;
 	}
 
@@ -425,48 +626,411 @@
 	} while (!(GLOB_SFT_RST & t) && (i-- > 0));
 
 	if (!(GLOB_SFT_RST & t)) {
-		printk(KERN_ERR DRV_NAME "can't set global reset\n");
-		rc = 1;		/* broken HW? */
+		printk(KERN_ERR DRV_NAME ": can't set global reset\n");
+		rc = 1;
 		goto done;
 	}
 
-	/* clear reset */
+	/* clear reset and *reenable the PCI master* (not mentioned in spec) */
 	i = 5;
 	do {
-		writel(t & ~GLOB_SFT_RST, reg);
+		writel(t & ~(GLOB_SFT_RST | STOP_PCI_MASTER), reg);
 		t = readl(reg);
 		udelay(1);
 	} while ((GLOB_SFT_RST & t) && (i-- > 0));
 
 	if (GLOB_SFT_RST & t) {
-		printk(KERN_ERR DRV_NAME "can't clear global reset\n");
-		rc = 1;		/* broken HW? */
+		printk(KERN_ERR DRV_NAME ": can't clear global reset\n");
+		rc = 1;
 	}
-
- done:
-	VPRINTK("EXIT, rc = %i\n", rc);
+done:
 	return rc;
 }
 
+/**
+ *      mv_host_stop - Host specific cleanup/stop routine.
+ *      @host_set: host data structure
+ *
+ *      Disable ints, cleanup host memory, call general purpose
+ *      host_stop.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_host_stop(struct ata_host_set *host_set)
+{
+	struct mv_host_priv *hpriv = host_set->private_data;
+	struct pci_dev *pdev = to_pci_dev(host_set->dev);
+
+	if (hpriv->hp_flags & MV_HP_FLAG_MSI) {
+		pci_disable_msi(pdev);
+	} else {
+		pci_intx(pdev, 0);
+	}
+	kfree(hpriv);
+	ata_host_stop(host_set);
+}
+
+/**
+ *      mv_port_start - Port specific init/start routine.
+ *      @ap: ATA channel to manipulate
+ *
+ *      Allocate and point to DMA memory, init port private memory,
+ *      zero indices.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static int mv_port_start(struct ata_port *ap)
+{
+	struct device *dev = ap->host_set->dev;
+	struct mv_port_priv *pp;
+	void __iomem *port_mmio = mv_ap_base(ap);
+	void *mem;
+	dma_addr_t mem_dma;
+
+	pp = kmalloc(sizeof(*pp), GFP_KERNEL);
+	if (!pp) {
+		return -ENOMEM;
+	}
+	memset(pp, 0, sizeof(*pp));
+
+	mem = dma_alloc_coherent(dev, MV_PORT_PRIV_DMA_SZ, &mem_dma, 
+				 GFP_KERNEL);
+	if (!mem) {
+		kfree(pp);
+		return -ENOMEM;
+	}
+	memset(mem, 0, MV_PORT_PRIV_DMA_SZ);
+
+	/* First item in chunk of DMA memory: 
+	 * 32-slot command request table (CRQB), 32 bytes each in size
+	 */
+	pp->crqb = mem;
+	pp->crqb_dma = mem_dma;
+	mem += MV_CRQB_Q_SZ;
+	mem_dma += MV_CRQB_Q_SZ;
+
+	/* Second item: 
+	 * 32-slot command response table (CRPB), 8 bytes each in size
+	 */
+	pp->crpb = mem;
+	pp->crpb_dma = mem_dma;
+	mem += MV_CRPB_Q_SZ;
+	mem_dma += MV_CRPB_Q_SZ;
+
+	/* Third item:
+	 * Table of scatter-gather descriptors (ePRD), 16 bytes each
+	 */
+	pp->sg_tbl = mem;
+	pp->sg_tbl_dma = mem_dma;
+
+	writelfl(EDMA_CFG_Q_DEPTH | EDMA_CFG_RD_BRST_EXT | 
+		 EDMA_CFG_WR_BUFF_LEN, port_mmio + EDMA_CFG_OFS);
+
+	writel((pp->crqb_dma >> 16) >> 16, port_mmio + EDMA_REQ_Q_BASE_HI_OFS);
+	writelfl(pp->crqb_dma & EDMA_REQ_Q_BASE_LO_MASK, 
+		 port_mmio + EDMA_REQ_Q_IN_PTR_OFS);
+
+	writelfl(0, port_mmio + EDMA_REQ_Q_OUT_PTR_OFS);
+	writelfl(0, port_mmio + EDMA_RSP_Q_IN_PTR_OFS);
+
+	writel((pp->crpb_dma >> 16) >> 16, port_mmio + EDMA_RSP_Q_BASE_HI_OFS);
+	writelfl(pp->crpb_dma & EDMA_RSP_Q_BASE_LO_MASK, 
+		 port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
+
+	pp->req_producer = pp->rsp_consumer = 0;
+
+	/* Don't turn on EDMA here...do it before DMA commands only.  Else
+	 * we'll be unable to send non-data, PIO, etc due to restricted access
+	 * to shadow regs.
+	 */
+	ap->private_data = pp;
+	return 0;
+}
+
+/**
+ *      mv_port_stop - Port specific cleanup/stop routine.
+ *      @ap: ATA channel to manipulate
+ *
+ *      Stop DMA, cleanup port memory.
+ *
+ *      LOCKING:
+ *      This routine uses the host_set lock to protect the DMA stop.
+ */
+static void mv_port_stop(struct ata_port *ap)
+{
+	struct device *dev = ap->host_set->dev;
+	struct mv_port_priv *pp = ap->private_data;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ap->host_set->lock, flags);
+	mv_stop_dma(ap);
+	spin_unlock_irqrestore(&ap->host_set->lock, flags);
+
+	ap->private_data = NULL;
+	dma_free_coherent(dev, MV_PORT_PRIV_DMA_SZ, pp->crpb, pp->crpb_dma);
+	kfree(pp);
+}
+
+/**
+ *      mv_fill_sg - Fill out the Marvell ePRD (scatter gather) entries
+ *      @qc: queued command whose SG list to source from
+ *
+ *      Populate the SG list and mark the last entry.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_fill_sg(struct ata_queued_cmd *qc)
+{
+	struct mv_port_priv *pp = qc->ap->private_data;
+	unsigned int i;
+
+	for (i = 0; i < qc->n_elem; i++) {
+		u32 sg_len;
+		dma_addr_t addr;
+
+		addr = sg_dma_address(&qc->sg[i]);
+		sg_len = sg_dma_len(&qc->sg[i]);
+
+		pp->sg_tbl[i].addr = cpu_to_le32(addr & 0xffffffff);
+		pp->sg_tbl[i].addr_hi = cpu_to_le32((addr >> 16) >> 16);
+		assert(0 == (sg_len & ~MV_DMA_BOUNDARY));
+		pp->sg_tbl[i].flags_size = cpu_to_le32(sg_len);
+	}
+	if (0 < qc->n_elem) {
+		pp->sg_tbl[qc->n_elem - 1].flags_size |= EPRD_FLAG_END_OF_TBL;
+	}
+}
+
+static inline unsigned mv_inc_q_index(unsigned *index)
+{
+	*index = (*index + 1) & MV_MAX_Q_DEPTH_MASK;
+	return *index;
+}
+
+static inline void mv_crqb_pack_cmd(u16 *cmdw, u8 data, u8 addr, unsigned last)
+{
+	*cmdw = data | (addr << CRQB_CMD_ADDR_SHIFT) | CRQB_CMD_CS |
+		(last ? CRQB_CMD_LAST : 0);
+}
+
+/**
+ *      mv_qc_prep - Host specific command preparation.
+ *      @qc: queued command to prepare
+ *
+ *      This routine simply redirects to the general purpose routine
+ *      if command is not DMA.  Else, it handles prep of the CRQB
+ *      (command request block), does some sanity checking, and calls
+ *      the SG load routine.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_qc_prep(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct mv_port_priv *pp = ap->private_data;
+	u16 *cw;
+	struct ata_taskfile *tf;
+	u16 flags = 0;
+
+ 	if (ATA_PROT_DMA != qc->tf.protocol) {
+		return;
+	}
+
+	/* the req producer index should be the same as we remember it */
+	assert(((readl(mv_ap_base(qc->ap) + EDMA_REQ_Q_IN_PTR_OFS) >> 
+		 EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) ==
+	       pp->req_producer);
+
+	/* Fill in command request block
+	 */
+	if (!(qc->tf.flags & ATA_TFLAG_WRITE)) {
+		flags |= CRQB_FLAG_READ;
+	}
+	assert(MV_MAX_Q_DEPTH > qc->tag);
+	flags |= qc->tag << CRQB_TAG_SHIFT;
+
+	pp->crqb[pp->req_producer].sg_addr = 
+		cpu_to_le32(pp->sg_tbl_dma & 0xffffffff);
+	pp->crqb[pp->req_producer].sg_addr_hi = 
+		cpu_to_le32((pp->sg_tbl_dma >> 16) >> 16);
+	pp->crqb[pp->req_producer].ctrl_flags = cpu_to_le16(flags);
+
+	cw = &pp->crqb[pp->req_producer].ata_cmd[0];
+	tf = &qc->tf;
+
+	/* Sadly, the CRQB cannot accomodate all registers--there are
+	 * only 11 bytes...so we must pick and choose required
+	 * registers based on the command.  So, we drop feature and
+	 * hob_feature for [RW] DMA commands, but they are needed for
+	 * NCQ.  NCQ will drop hob_nsect.
+	 */
+	switch (tf->command) {
+	case ATA_CMD_READ:
+	case ATA_CMD_READ_EXT:
+	case ATA_CMD_WRITE:
+	case ATA_CMD_WRITE_EXT:
+		mv_crqb_pack_cmd(cw++, tf->hob_nsect, ATA_REG_NSECT, 0);
+		break;
+#ifdef LIBATA_NCQ		/* FIXME: remove this line when NCQ added */
+	case ATA_CMD_FPDMA_READ:
+	case ATA_CMD_FPDMA_WRITE:
+		mv_crqb_pack_cmd(cw++, tf->hob_feature, ATA_REG_FEATURE, 0); 
+		mv_crqb_pack_cmd(cw++, tf->feature, ATA_REG_FEATURE, 0);
+		break;
+#endif				/* FIXME: remove this line when NCQ added */
+	default:
+		/* The only other commands EDMA supports in non-queued and
+		 * non-NCQ mode are: [RW] STREAM DMA and W DMA FUA EXT, none
+		 * of which are defined/used by Linux.  If we get here, this
+		 * driver needs work.
+		 *
+		 * FIXME: modify libata to give qc_prep a return value and
+		 * return error here.
+		 */
+		BUG_ON(tf->command);
+		break;
+	}
+	mv_crqb_pack_cmd(cw++, tf->nsect, ATA_REG_NSECT, 0);
+	mv_crqb_pack_cmd(cw++, tf->hob_lbal, ATA_REG_LBAL, 0);
+	mv_crqb_pack_cmd(cw++, tf->lbal, ATA_REG_LBAL, 0);
+	mv_crqb_pack_cmd(cw++, tf->hob_lbam, ATA_REG_LBAM, 0);
+	mv_crqb_pack_cmd(cw++, tf->lbam, ATA_REG_LBAM, 0);
+	mv_crqb_pack_cmd(cw++, tf->hob_lbah, ATA_REG_LBAH, 0);
+	mv_crqb_pack_cmd(cw++, tf->lbah, ATA_REG_LBAH, 0);
+	mv_crqb_pack_cmd(cw++, tf->device, ATA_REG_DEVICE, 0);
+	mv_crqb_pack_cmd(cw++, tf->command, ATA_REG_CMD, 1);	/* last */
+
+	if (!(qc->flags & ATA_QCFLAG_DMAMAP)) {
+		return;
+	}
+	mv_fill_sg(qc);
+}
+
+/**
+ *      mv_qc_issue - Initiate a command to the host
+ *      @qc: queued command to start
+ *
+ *      This routine simply redirects to the general purpose routine
+ *      if command is not DMA.  Else, it sanity checks our local
+ *      caches of the request producer/consumer indices then enables
+ *      DMA and bumps the request producer index.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static int mv_qc_issue(struct ata_queued_cmd *qc)
+{
+	void __iomem *port_mmio = mv_ap_base(qc->ap);
+	struct mv_port_priv *pp = qc->ap->private_data;
+	u32 in_ptr;
+
+	if (ATA_PROT_DMA != qc->tf.protocol) {
+		/* We're about to send a non-EDMA capable command to the
+		 * port.  Turn off EDMA so there won't be problems accessing
+		 * shadow block, etc registers.
+		 */
+		mv_stop_dma(qc->ap);
+		return ata_qc_issue_prot(qc);
+	}
+
+	in_ptr = readl(port_mmio + EDMA_REQ_Q_IN_PTR_OFS);
+
+	/* the req producer index should be the same as we remember it */
+	assert(((in_ptr >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) ==
+	       pp->req_producer);
+	/* until we do queuing, the queue should be empty at this point */
+	assert(((in_ptr >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) ==
+	       ((readl(port_mmio + EDMA_REQ_Q_OUT_PTR_OFS) >> 
+		 EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK));
+
+	mv_inc_q_index(&pp->req_producer);	/* now incr producer index */
+
+	mv_start_dma(port_mmio, pp);
+
+	/* and write the request in pointer to kick the EDMA to life */
+	in_ptr &= EDMA_REQ_Q_BASE_LO_MASK;
+	in_ptr |= pp->req_producer << EDMA_REQ_Q_PTR_SHIFT;
+	writelfl(in_ptr, port_mmio + EDMA_REQ_Q_IN_PTR_OFS);
+
+	return 0;
+}
+
+/**
+ *      mv_get_crpb_status - get status from most recently completed cmd
+ *      @ap: ATA channel to manipulate
+ *
+ *      This routine is for use when the port is in DMA mode, when it
+ *      will be using the CRPB (command response block) method of
+ *      returning command completion information.  We assert indices
+ *      are good, grab status, and bump the response consumer index to
+ *      prove that we're up to date.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static u8 mv_get_crpb_status(struct ata_port *ap)
+{
+	void __iomem *port_mmio = mv_ap_base(ap);
+	struct mv_port_priv *pp = ap->private_data;
+	u32 out_ptr;
+
+	out_ptr = readl(port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
+
+	/* the response consumer index should be the same as we remember it */
+	assert(((out_ptr >> EDMA_RSP_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) == 
+	       pp->rsp_consumer);
+
+	/* increment our consumer index... */
+	pp->rsp_consumer = mv_inc_q_index(&pp->rsp_consumer);
+	
+	/* and, until we do NCQ, there should only be 1 CRPB waiting */
+	assert(((readl(port_mmio + EDMA_RSP_Q_IN_PTR_OFS) >> 
+		 EDMA_RSP_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) == 
+	       pp->rsp_consumer);
+
+	/* write out our inc'd consumer index so EDMA knows we're caught up */
+	out_ptr &= EDMA_RSP_Q_BASE_LO_MASK;
+	out_ptr |= pp->rsp_consumer << EDMA_RSP_Q_PTR_SHIFT;
+	writelfl(out_ptr, port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
+
+	/* Return ATA status register for completed CRPB */
+	return (pp->crpb[pp->rsp_consumer].flags >> CRPB_FLAG_STATUS_SHIFT);
+}
+
+/**
+ *      mv_err_intr - Handle error interrupts on the port
+ *      @ap: ATA channel to manipulate
+ *
+ *      In most cases, just clear the interrupt and move on.  However,
+ *      some cases require an eDMA reset, which is done right before
+ *      the COMRESET in mv_phy_reset().  The SERR case requires a
+ *      clear of pending errors in the SATA SERROR register.  Finally,
+ *      if the port disabled DMA, update our cached copy to match.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
 static void mv_err_intr(struct ata_port *ap)
 {
-	void __iomem *port_mmio;
+	void __iomem *port_mmio = mv_ap_base(ap);
 	u32 edma_err_cause, serr = 0;
 
-	/* bug here b/c we got an err int on a port we don't know about,
-	 * so there's no way to clear it
-	 */
-	BUG_ON(NULL == ap);
-	port_mmio = mv_ap_base(ap);
-
 	edma_err_cause = readl(port_mmio + EDMA_ERR_IRQ_CAUSE_OFS);
 
 	if (EDMA_ERR_SERR & edma_err_cause) {
 		serr = scr_read(ap, SCR_ERROR);
 		scr_write_flush(ap, SCR_ERROR, serr);
 	}
-	DPRINTK("port %u error; EDMA err cause: 0x%08x SERR: 0x%08x\n", 
-		ap->port_no, edma_err_cause, serr);
+	if (EDMA_ERR_SELF_DIS & edma_err_cause) {
+		struct mv_port_priv *pp	= ap->private_data;
+		pp->pp_flags &= ~MV_PP_FLAG_EDMA_EN;
+	}
+	DPRINTK(KERN_ERR "ata%u: port error; EDMA err cause: 0x%08x "
+		"SERR: 0x%08x\n", ap->id, edma_err_cause, serr);
 
 	/* Clear EDMA now that SERR cleanup done */
 	writelfl(0, port_mmio + EDMA_ERR_IRQ_CAUSE_OFS);
@@ -477,7 +1041,21 @@
 	}
 }
 
-/* Handle any outstanding interrupts in a single SATAHC 
+/**
+ *      mv_host_intr - Handle all interrupts on the given host controller
+ *      @host_set: host specific structure
+ *      @relevant: port error bits relevant to this host controller
+ *      @hc: which host controller we're to look at
+ *
+ *      Read then write clear the HC interrupt status then walk each
+ *      port connected to the HC and see if it needs servicing.  Port
+ *      success ints are reported in the HC interrupt status reg, the
+ *      port error ints are reported in the higher level main
+ *      interrupt status register and thus are passed in via the
+ *      'relevant' argument.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
  */
 static void mv_host_intr(struct ata_host_set *host_set, u32 relevant,
 			 unsigned int hc)
@@ -487,8 +1065,8 @@
 	struct ata_port *ap;
 	struct ata_queued_cmd *qc;
 	u32 hc_irq_cause;
-	int shift, port, port0, hard_port;
-	u8 ata_status;
+	int shift, port, port0, hard_port, handled;
+	u8 ata_status = 0;
 
 	if (hc == 0) {
 		port0 = 0;
@@ -499,7 +1077,7 @@
 	/* we'll need the HC success int register in most cases */
 	hc_irq_cause = readl(hc_mmio + HC_IRQ_CAUSE_OFS);
 	if (hc_irq_cause) {
-		writelfl(0, hc_mmio + HC_IRQ_CAUSE_OFS);
+		writelfl(~hc_irq_cause, hc_mmio + HC_IRQ_CAUSE_OFS);
 	}
 
 	VPRINTK("ENTER, hc%u relevant=0x%08x HC IRQ cause=0x%08x\n",
@@ -508,35 +1086,38 @@
 	for (port = port0; port < port0 + MV_PORTS_PER_HC; port++) {
 		ap = host_set->ports[port];
 		hard_port = port & MV_PORT_MASK;	/* range 0-3 */
-		ata_status = 0xffU;
+		handled = 0;	/* ensure ata_status is set if handled++ */
 
-		if (((CRBP_DMA_DONE | DEV_IRQ) << hard_port) & hc_irq_cause) {
-			BUG_ON(NULL == ap);
-			/* rcv'd new resp, basic DMA complete, or ATA IRQ */
-			/* This is needed to clear the ATA INTRQ.
-			 * FIXME: don't read the status reg in EDMA mode!
+		if ((CRPB_DMA_DONE << hard_port) & hc_irq_cause) {
+			/* new CRPB on the queue; just one at a time until NCQ
+			 */
+			ata_status = mv_get_crpb_status(ap);
+			handled++;
+		} else if ((DEV_IRQ << hard_port) & hc_irq_cause) {
+			/* received ATA IRQ; read the status reg to clear INTRQ
 			 */
 			ata_status = readb((void __iomem *)
 					   ap->ioaddr.status_addr);
+			handled++;
 		}
 
-		shift = port * 2;
+		shift = port << 1;		/* (port * 2) */
 		if (port >= MV_PORTS_PER_HC) {
 			shift++;	/* skip bit 8 in the HC Main IRQ reg */
 		}
 		if ((PORT0_ERR << shift) & relevant) {
 			mv_err_intr(ap);
-			/* FIXME: smart to OR in ATA_ERR? */
+			/* OR in ATA_ERR to ensure libata knows we took one */
 			ata_status = readb((void __iomem *)
 					   ap->ioaddr.status_addr) | ATA_ERR;
+			handled++;
 		}
 		
-		if (ap) {
+		if (handled && ap) {
 			qc = ata_qc_from_tag(ap, ap->active_tag);
 			if (NULL != qc) {
 				VPRINTK("port %u IRQ found for qc, "
 					"ata_status 0x%x\n", port,ata_status);
-				BUG_ON(0xffU == ata_status);
 				/* mark qc status appropriately */
 				ata_qc_complete(qc, ata_status);
 			}
@@ -545,17 +1126,30 @@
 	VPRINTK("EXIT\n");
 }
 
+/**
+ *      mv_interrupt - 
+ *      @irq: unused
+ *      @dev_instance: private data; in this case the host structure
+ *      @regs: unused
+ *
+ *      Read the read only register to determine if any host
+ *      controllers have pending interrupts.  If so, call lower level
+ *      routine to handle.  Also check for PCI errors which are only
+ *      reported here.
+ *
+ *      LOCKING: 
+ *      This routine holds the host_set lock while processing pending
+ *      interrupts.
+ */
 static irqreturn_t mv_interrupt(int irq, void *dev_instance,
 				struct pt_regs *regs)
 {
 	struct ata_host_set *host_set = dev_instance;
 	unsigned int hc, handled = 0, n_hcs;
-	void __iomem *mmio;
+	void __iomem *mmio = host_set->mmio_base;
 	u32 irq_stat;
 
-	mmio = host_set->mmio_base;
 	irq_stat = readl(mmio + HC_MAIN_IRQ_CAUSE_OFS);
-	n_hcs = mv_get_hc_count(host_set->ports[0]->flags);
 
 	/* check the cases where we either have nothing pending or have read
 	 * a bogus register value which can indicate HW removal or PCI fault
@@ -564,64 +1158,105 @@
 		return IRQ_NONE;
 	}
 
+	n_hcs = mv_get_hc_count(host_set->ports[0]->flags);
 	spin_lock(&host_set->lock);
 
 	for (hc = 0; hc < n_hcs; hc++) {
 		u32 relevant = irq_stat & (HC0_IRQ_PEND << (hc * HC_SHIFT));
 		if (relevant) {
 			mv_host_intr(host_set, relevant, hc);
-			handled = 1;
+			handled++;
 		}
 	}
 	if (PCI_ERR & irq_stat) {
-		/* FIXME: these are all masked by default, but still need
-		 * to recover from them properly.
-		 */
-	}
+		printk(KERN_ERR DRV_NAME ": PCI ERROR; PCI IRQ cause=0x%08x\n",
+		       readl(mmio + PCI_IRQ_CAUSE_OFS));
 
+		DPRINTK("All regs @ PCI error\n");
+		mv_dump_all_regs(mmio, -1, to_pci_dev(host_set->dev));
+
+		writelfl(0, mmio + PCI_IRQ_CAUSE_OFS);
+		handled++;
+	}
 	spin_unlock(&host_set->lock);
 
 	return IRQ_RETVAL(handled);
 }
 
+/**
+ *      mv_check_err - Return the error shadow register to caller.
+ *      @ap: ATA channel to manipulate
+ *
+ *      Marvell requires DMA to be stopped before accessing shadow
+ *      registers.  So we do that, then return the needed register.
+ *
+ *      LOCKING:
+ *      Inherited from caller.  FIXME: protect mv_stop_dma with lock?
+ */
+static u8 mv_check_err(struct ata_port *ap)
+{
+	mv_stop_dma(ap);		/* can't read shadow regs if DMA on */
+	return readb((void __iomem *) ap->ioaddr.error_addr);
+}
+
+/**
+ *      mv_phy_reset - Perform eDMA reset followed by COMRESET
+ *      @ap: ATA channel to manipulate
+ *
+ *      Part of this is taken from __sata_phy_reset and modified to
+ *      not sleep since this routine gets called from interrupt level.
+ *
+ *      LOCKING:
+ *      Inherited from caller.  This is coded to safe to call at
+ *      interrupt level, i.e. it does not sleep.
+ */
 static void mv_phy_reset(struct ata_port *ap)
 {
 	void __iomem *port_mmio = mv_ap_base(ap);
 	struct ata_taskfile tf;
 	struct ata_device *dev = &ap->device[0];
-	u32 edma = 0, bdma;
+	unsigned long timeout;
 
 	VPRINTK("ENTER, port %u, mmio 0x%p\n", ap->port_no, port_mmio);
 
-	edma = readl(port_mmio + EDMA_CMD_OFS);
-	if (EDMA_EN & edma) {
-		/* disable EDMA if active */
-		edma &= ~EDMA_EN;
-		writelfl(edma | EDMA_DS, port_mmio + EDMA_CMD_OFS);
-		udelay(1);
-	} else if (mv_port_bdma_capable(ap) &&
-		   (bdma = readl(port_mmio + BDMA_CMD_OFS)) & BDMA_START) {
-		/* disable BDMA if active */
-		writelfl(bdma & ~BDMA_START, port_mmio + BDMA_CMD_OFS);
-	}
+	mv_stop_dma(ap);
 
-	writelfl(edma | ATA_RST, port_mmio + EDMA_CMD_OFS);
+	writelfl(ATA_RST, port_mmio + EDMA_CMD_OFS);
 	udelay(25);		/* allow reset propagation */
 
 	/* Spec never mentions clearing the bit.  Marvell's driver does
 	 * clear the bit, however.
 	 */
-	writelfl(edma & ~ATA_RST, port_mmio + EDMA_CMD_OFS);
+	writelfl(0, port_mmio + EDMA_CMD_OFS);
 
-	VPRINTK("Done.  Now calling __sata_phy_reset()\n");
+	VPRINTK("S-regs after ATA_RST: SStat 0x%08x SErr 0x%08x "
+		"SCtrl 0x%08x\n", mv_scr_read(ap, SCR_STATUS),
+		mv_scr_read(ap, SCR_ERROR), mv_scr_read(ap, SCR_CONTROL));
 
 	/* proceed to init communications via the scr_control reg */
-	__sata_phy_reset(ap);
+	scr_write_flush(ap, SCR_CONTROL, 0x301);
+	mdelay(1);
+	scr_write_flush(ap, SCR_CONTROL, 0x300);
+	timeout = jiffies + (HZ * 1);
+	do {
+		mdelay(10);
+		if ((scr_read(ap, SCR_STATUS) & 0xf) != 1)
+			break;
+	} while (time_before(jiffies, timeout));
 
-	if (ap->flags & ATA_FLAG_PORT_DISABLED) {
-		VPRINTK("Port disabled pre-sig.  Exiting.\n");
+	VPRINTK("S-regs after PHY wake: SStat 0x%08x SErr 0x%08x "
+		"SCtrl 0x%08x\n", mv_scr_read(ap, SCR_STATUS),
+		mv_scr_read(ap, SCR_ERROR), mv_scr_read(ap, SCR_CONTROL));
+
+	if (sata_dev_present(ap)) {
+		ata_port_probe(ap);
+	} else {
+		printk(KERN_INFO "ata%u: no device found (phy stat %08x)\n",
+		       ap->id, scr_read(ap, SCR_STATUS));
+		ata_port_disable(ap);
 		return;
 	}
+	ap->cbl = ATA_CBL_SATA;
 
 	tf.lbah = readb((void __iomem *) ap->ioaddr.lbah_addr);
 	tf.lbam = readb((void __iomem *) ap->ioaddr.lbam_addr);
@@ -636,37 +1271,118 @@
 	VPRINTK("EXIT\n");
 }
 
-static void mv_port_init(struct ata_ioports *port, unsigned long base)
+/**
+ *      mv_eng_timeout - Routine called by libata when SCSI times out I/O
+ *      @ap: ATA channel to manipulate
+ *
+ *      Intent is to clear all pending error conditions, reset the
+ *      chip/bus, fail the command, and move on.
+ *
+ *      LOCKING:
+ *      This routine holds the host_set lock while failing the command.
+ */
+static void mv_eng_timeout(struct ata_port *ap)
 {
-	/* PIO related setup */
-	port->data_addr = base + SHD_PIO_DATA_OFS;
-	port->error_addr = port->feature_addr = base + SHD_FEA_ERR_OFS;
-	port->nsect_addr = base + SHD_SECT_CNT_OFS;
-	port->lbal_addr = base + SHD_LBA_L_OFS;
-	port->lbam_addr = base + SHD_LBA_M_OFS;
-	port->lbah_addr = base + SHD_LBA_H_OFS;
-	port->device_addr = base + SHD_DEV_HD_OFS;
-	port->status_addr = port->command_addr = base + SHD_CMD_STA_OFS;
-	port->altstatus_addr = port->ctl_addr = base + SHD_CTL_AST_OFS;
-	/* unused */
-	port->cmd_addr = port->bmdma_addr = port->scr_addr = 0;
+	struct ata_queued_cmd *qc;
+	unsigned long flags;
 
-	/* unmask all EDMA error interrupts */
-	writel(~0, (void __iomem *)base + EDMA_ERR_IRQ_MASK_OFS);
+	printk(KERN_ERR "ata%u: Entering mv_eng_timeout\n",ap->id);
+	DPRINTK("All regs @ start of eng_timeout\n");
+	mv_dump_all_regs(ap->host_set->mmio_base, ap->port_no, 
+			 to_pci_dev(ap->host_set->dev));
 
-	VPRINTK("EDMA cfg=0x%08x EDMA IRQ err cause/mask=0x%08x/0x%08x\n", 
-		readl((void __iomem *)base + EDMA_CFG_OFS),
-		readl((void __iomem *)base + EDMA_ERR_IRQ_CAUSE_OFS),
-		readl((void __iomem *)base + EDMA_ERR_IRQ_MASK_OFS));
+	qc = ata_qc_from_tag(ap, ap->active_tag);
+        printk(KERN_ERR "mmio_base %p ap %p qc %p scsi_cmnd %p &cmnd %p\n",
+	       ap->host_set->mmio_base, ap, qc, qc->scsicmd, 
+	       &qc->scsicmd->cmnd);
+
+	mv_err_intr(ap);
+	mv_phy_reset(ap);
+
+	if (!qc) {
+		printk(KERN_ERR "ata%u: BUG: timeout without command\n",
+		       ap->id);
+	} else {
+		/* hack alert!  We cannot use the supplied completion
+	 	 * function from inside the ->eh_strategy_handler() thread.
+	 	 * libata is the only user of ->eh_strategy_handler() in
+	 	 * any kernel, so the default scsi_done() assumes it is
+	 	 * not being called from the SCSI EH.
+	 	 */
+		spin_lock_irqsave(&ap->host_set->lock, flags);
+		qc->scsidone = scsi_finish_command;
+		ata_qc_complete(qc, ATA_ERR);
+		spin_unlock_irqrestore(&ap->host_set->lock, flags);
+	}
 }
 
+/**
+ *      mv_port_init - Perform some early initialization on a single port.
+ *      @port: libata data structure storing shadow register addresses
+ *      @port_mmio: base address of the port
+ *
+ *      Initialize shadow register mmio addresses, clear outstanding
+ *      interrupts on the port, and unmask interrupts for the future
+ *      start of the port.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_port_init(struct ata_ioports *port,  void __iomem *port_mmio)
+{
+	unsigned long shd_base = (unsigned long) port_mmio + SHD_BLK_OFS;
+	unsigned serr_ofs;
+
+	/* PIO related setup 
+	 */
+	port->data_addr = shd_base + (sizeof(u32) * ATA_REG_DATA);
+	port->error_addr = 
+		port->feature_addr = shd_base + (sizeof(u32) * ATA_REG_ERR);
+	port->nsect_addr = shd_base + (sizeof(u32) * ATA_REG_NSECT);
+	port->lbal_addr = shd_base + (sizeof(u32) * ATA_REG_LBAL);
+	port->lbam_addr = shd_base + (sizeof(u32) * ATA_REG_LBAM);
+	port->lbah_addr = shd_base + (sizeof(u32) * ATA_REG_LBAH);
+	port->device_addr = shd_base + (sizeof(u32) * ATA_REG_DEVICE);
+	port->status_addr = 
+		port->command_addr = shd_base + (sizeof(u32) * ATA_REG_STATUS);
+	/* special case: control/altstatus doesn't have ATA_REG_ address */
+	port->altstatus_addr = port->ctl_addr = shd_base + SHD_CTL_AST_OFS;
+
+	/* unused: */
+	port->cmd_addr = port->bmdma_addr = port->scr_addr = 0;
+
+	/* Clear any currently outstanding port interrupt conditions */
+	serr_ofs = mv_scr_offset(SCR_ERROR);
+	writelfl(readl(port_mmio + serr_ofs), port_mmio + serr_ofs);
+	writelfl(0, port_mmio + EDMA_ERR_IRQ_CAUSE_OFS);
+
+	/* unmask all EDMA error interrupts */
+	writelfl(~0, port_mmio + EDMA_ERR_IRQ_MASK_OFS);
+
+	VPRINTK("EDMA cfg=0x%08x EDMA IRQ err cause/mask=0x%08x/0x%08x\n", 
+		readl(port_mmio + EDMA_CFG_OFS),
+		readl(port_mmio + EDMA_ERR_IRQ_CAUSE_OFS),
+		readl(port_mmio + EDMA_ERR_IRQ_MASK_OFS));
+}
+
+/**
+ *      mv_host_init - Perform some early initialization of the host.
+ *      @probe_ent: early data struct representing the host
+ *
+ *      If possible, do an early global reset of the host.  Then do
+ *      our port init and clear/unmask all/relevant host interrupts.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
 static int mv_host_init(struct ata_probe_ent *probe_ent)
 {
 	int rc = 0, n_hc, port, hc;
 	void __iomem *mmio = probe_ent->mmio_base;
 	void __iomem *port_mmio;
 
-	if (mv_master_reset(probe_ent->mmio_base)) {
+	if ((MV_FLAG_GLBL_SFT_RST & probe_ent->host_flags) && 
+	    mv_global_soft_reset(probe_ent->mmio_base)) {
 		rc = 1;
 		goto done;
 	}
@@ -676,17 +1392,27 @@
 
 	for (port = 0; port < probe_ent->n_ports; port++) {
 		port_mmio = mv_port_base(mmio, port);
-		mv_port_init(&probe_ent->port[port], (unsigned long)port_mmio);
+		mv_port_init(&probe_ent->port[port], port_mmio);
 	}
 
 	for (hc = 0; hc < n_hc; hc++) {
-		VPRINTK("HC%i: HC config=0x%08x HC IRQ cause=0x%08x\n", hc,
-			readl(mv_hc_base(mmio, hc) + HC_CFG_OFS),
-			readl(mv_hc_base(mmio, hc) + HC_IRQ_CAUSE_OFS));
+		void __iomem *hc_mmio = mv_hc_base(mmio, hc);
+
+		VPRINTK("HC%i: HC config=0x%08x HC IRQ cause "
+			"(before clear)=0x%08x\n", hc,
+			readl(hc_mmio + HC_CFG_OFS),
+			readl(hc_mmio + HC_IRQ_CAUSE_OFS));
+
+		/* Clear any currently outstanding hc interrupt conditions */
+		writelfl(0, hc_mmio + HC_IRQ_CAUSE_OFS);
 	}
 
-	writel(~HC_MAIN_MASKED_IRQS, mmio + HC_MAIN_IRQ_MASK_OFS);
-	writel(PCI_UNMASK_ALL_IRQS, mmio + PCI_IRQ_MASK_OFS);
+	/* Clear any currently outstanding host interrupt conditions */
+	writelfl(0, mmio + PCI_IRQ_CAUSE_OFS);
+
+	/* and unmask interrupt generation for host regs */
+	writelfl(PCI_UNMASK_ALL_IRQS, mmio + PCI_IRQ_MASK_OFS);
+	writelfl(~HC_MAIN_MASKED_IRQS, mmio + HC_MAIN_IRQ_MASK_OFS);
 
 	VPRINTK("HC MAIN IRQ cause/mask=0x%08x/0x%08x "
 		"PCI int cause/mask=0x%08x/0x%08x\n", 
@@ -694,11 +1420,53 @@
 		readl(mmio + HC_MAIN_IRQ_MASK_OFS),
 		readl(mmio + PCI_IRQ_CAUSE_OFS),
 		readl(mmio + PCI_IRQ_MASK_OFS));
-
- done:
+done:
 	return rc;
 }
 
+/**
+ *      mv_print_info - Dump key info to kernel log for perusal.
+ *      @probe_ent: early data struct representing the host
+ *
+ *      FIXME: complete this.
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
+static void mv_print_info(struct ata_probe_ent *probe_ent)
+{
+	struct pci_dev *pdev = to_pci_dev(probe_ent->dev);
+	struct mv_host_priv *hpriv = probe_ent->private_data;
+	u8 rev_id, scc;
+	const char *scc_s;
+
+	/* Use this to determine the HW stepping of the chip so we know
+	 * what errata to workaround
+	 */
+	pci_read_config_byte(pdev, PCI_REVISION_ID, &rev_id);
+
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &scc);
+	if (scc == 0)
+		scc_s = "SCSI";
+	else if (scc == 0x01)
+		scc_s = "RAID";
+	else
+		scc_s = "unknown";
+
+	printk(KERN_INFO DRV_NAME 
+	       "(%s) %u slots %u ports %s mode IRQ via %s\n",
+	       pci_name(pdev), (unsigned)MV_MAX_Q_DEPTH, probe_ent->n_ports, 
+	       scc_s, (MV_HP_FLAG_MSI & hpriv->hp_flags) ? "MSI" : "INTx");
+}
+
+/**
+ *      mv_init_one - handle a positive probe of a Marvell host
+ *      @pdev: PCI device found
+ *      @ent: PCI device ID entry for the matched host
+ *
+ *      LOCKING:
+ *      Inherited from caller.
+ */
 static int mv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	static int printed_version = 0;
@@ -706,16 +1474,12 @@
 	struct mv_host_priv *hpriv;
 	unsigned int board_idx = (unsigned int)ent->driver_data;
 	void __iomem *mmio_base;
-	int pci_dev_busy = 0;
-	int rc;
+	int pci_dev_busy = 0, rc;
 
 	if (!printed_version++) {
-		printk(KERN_DEBUG DRV_NAME " version " DRV_VERSION "\n");
+		printk(KERN_INFO DRV_NAME " version " DRV_VERSION "\n");
 	}
 
-	VPRINTK("ENTER for PCI Bus:Slot.Func=%u:%u.%u\n", pdev->bus->number,
-		PCI_SLOT(pdev->devfn), PCI_FUNC(pdev->devfn));
-
 	rc = pci_enable_device(pdev);
 	if (rc) {
 		return rc;
@@ -727,8 +1491,6 @@
 		goto err_out;
 	}
 
-	pci_intx(pdev, 1);
-
 	probe_ent = kmalloc(sizeof(*probe_ent), GFP_KERNEL);
 	if (probe_ent == NULL) {
 		rc = -ENOMEM;
@@ -739,8 +1501,7 @@
 	probe_ent->dev = pci_dev_to_dev(pdev);
 	INIT_LIST_HEAD(&probe_ent->node);
 
-	mmio_base = ioremap_nocache(pci_resource_start(pdev, MV_PRIMARY_BAR),
-				    pci_resource_len(pdev, MV_PRIMARY_BAR));
+	mmio_base = pci_iomap(pdev, MV_PRIMARY_BAR, 0);
 	if (mmio_base == NULL) {
 		rc = -ENOMEM;
 		goto err_out_free_ent;
@@ -769,37 +1530,40 @@
 	if (rc) {
 		goto err_out_hpriv;
 	}
-/* 	mv_print_info(probe_ent); */
 
-	{
-		int b, w;
-		u32 dw[4];	/* hold a line of 16b */
-		VPRINTK("PCI config space:\n");
-		for (b = 0; b < 0x40; ) {
-			for (w = 0; w < 4; w++) {
-				(void) pci_read_config_dword(pdev,b,&dw[w]);
-				b += sizeof(*dw);
-			}
-			VPRINTK("%08x %08x %08x %08x\n",
-				dw[0],dw[1],dw[2],dw[3]);
-		}
+	/* Enable interrupts */
+	if (pci_enable_msi(pdev) == 0) {
+		hpriv->hp_flags |= MV_HP_FLAG_MSI;
+	} else {
+		pci_intx(pdev, 1);
 	}
 
-	/* FIXME: check ata_device_add return value */
-	ata_device_add(probe_ent);
-	kfree(probe_ent);
+	mv_dump_pci_cfg(pdev, 0x68);
+	mv_print_info(probe_ent);
 
+	if (ata_device_add(probe_ent) == 0) {
+		rc = -ENODEV;		/* No devices discovered */
+		goto err_out_dev_add;
+	}
+
+	kfree(probe_ent);
 	return 0;
 
- err_out_hpriv:
+err_out_dev_add:
+	if (MV_HP_FLAG_MSI & hpriv->hp_flags) {
+		pci_disable_msi(pdev);
+	} else {
+		pci_intx(pdev, 0);
+	}
+err_out_hpriv:
 	kfree(hpriv);
- err_out_iounmap:
-	iounmap(mmio_base);
- err_out_free_ent:
+err_out_iounmap:
+	pci_iounmap(pdev, mmio_base);
+err_out_free_ent:
 	kfree(probe_ent);
- err_out_regions:
+err_out_regions:
 	pci_release_regions(pdev);
- err_out:
+err_out:
 	if (!pci_dev_busy) {
 		pci_disable_device(pdev);
 	}
diff --git a/drivers/scsi/sata_nv.c b/drivers/scsi/sata_nv.c
index c05653c..9fa2535 100644
--- a/drivers/scsi/sata_nv.c
+++ b/drivers/scsi/sata_nv.c
@@ -29,6 +29,8 @@
  *  NV-specific details such as register offsets, SATA phy location,
  *  hotplug info, etc.
  *
+ *  0.09
+ *     - Fixed bug introduced by 0.08's MCP51 and MCP55 support.
  *
  *  0.08
  *     - Added support for MCP51 and MCP55.
@@ -132,9 +134,7 @@
 	GENERIC,
 	NFORCE2,
 	NFORCE3,
-	CK804,
-	MCP51,
-	MCP55
+	CK804
 };
 
 static struct pci_device_id nv_pci_tbl[] = {
@@ -153,13 +153,13 @@
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_SATA2,
 		PCI_ANY_ID, PCI_ANY_ID, 0, 0, CK804 },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA,
-		PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP51 },
+		PCI_ANY_ID, PCI_ANY_ID, 0, 0, GENERIC },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA2,
-		PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP51 },
+		PCI_ANY_ID, PCI_ANY_ID, 0, 0, GENERIC },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA,
-		PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP55 },
+		PCI_ANY_ID, PCI_ANY_ID, 0, 0, GENERIC },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA2,
-		PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP55 },
+		PCI_ANY_ID, PCI_ANY_ID, 0, 0, GENERIC },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID,
 		PCI_ANY_ID, PCI_ANY_ID,
 		PCI_CLASS_STORAGE_IDE<<8, 0xffff00, GENERIC },
@@ -405,7 +405,7 @@
 	rc = -ENOMEM;
 
 	ppi = &nv_port_info;
-	probe_ent = ata_pci_init_native_mode(pdev, &ppi);
+	probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
 	if (!probe_ent)
 		goto err_out_regions;
 
diff --git a/drivers/scsi/sata_promise.c b/drivers/scsi/sata_promise.c
index 538ad72..def7e0d 100644
--- a/drivers/scsi/sata_promise.c
+++ b/drivers/scsi/sata_promise.c
@@ -438,11 +438,11 @@
 		break;
 
         default:
-                ap->stats.idle_irq++;
-                break;
+		ap->stats.idle_irq++;
+		break;
         }
 
-        return handled;
+	return handled;
 }
 
 static void pdc_irq_clear(struct ata_port *ap)
diff --git a/drivers/scsi/sata_sil24.c b/drivers/scsi/sata_sil24.c
new file mode 100644
index 0000000..1985781
--- /dev/null
+++ b/drivers/scsi/sata_sil24.c
@@ -0,0 +1,875 @@
+/*
+ * sata_sil24.c - Driver for Silicon Image 3124/3132 SATA-2 controllers
+ *
+ * Copyright 2005  Tejun Heo
+ *
+ * Based on preview driver from Silicon Image.
+ *
+ * NOTE: No NCQ/ATAPI support yet.  The preview driver didn't support
+ * NCQ nor ATAPI, and, unfortunately, I couldn't find out how to make
+ * those work.  Enabling those shouldn't be difficult.  Basic
+ * structure is all there (in libata-dev tree).  If you have any
+ * information about this hardware, please contact me or linux-ide.
+ * Info is needed on...
+ *
+ * - How to issue tagged commands and turn on sactive on issue accordingly.
+ * - Where to put an ATAPI command and how to tell the device to send it.
+ * - How to enable/use 64bit.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/dma-mapping.h>
+#include <scsi/scsi_host.h>
+#include "scsi.h"
+#include <linux/libata.h>
+#include <asm/io.h>
+
+#define DRV_NAME	"sata_sil24"
+#define DRV_VERSION	"0.22"	/* Silicon Image's preview driver was 0.10 */
+
+/*
+ * Port request block (PRB) 32 bytes
+ */
+struct sil24_prb {
+	u16	ctrl;
+	u16	prot;
+	u32	rx_cnt;
+	u8	fis[6 * 4];
+};
+
+/*
+ * Scatter gather entry (SGE) 16 bytes
+ */
+struct sil24_sge {
+	u64	addr;
+	u32	cnt;
+	u32	flags;
+};
+
+/*
+ * Port multiplier
+ */
+struct sil24_port_multiplier {
+	u32	diag;
+	u32	sactive;
+};
+
+enum {
+	/*
+	 * Global controller registers (128 bytes @ BAR0)
+	 */
+		/* 32 bit regs */
+	HOST_SLOT_STAT		= 0x00, /* 32 bit slot stat * 4 */
+	HOST_CTRL		= 0x40,
+	HOST_IRQ_STAT		= 0x44,
+	HOST_PHY_CFG		= 0x48,
+	HOST_BIST_CTRL		= 0x50,
+	HOST_BIST_PTRN		= 0x54,
+	HOST_BIST_STAT		= 0x58,
+	HOST_MEM_BIST_STAT	= 0x5c,
+	HOST_FLASH_CMD		= 0x70,
+		/* 8 bit regs */
+	HOST_FLASH_DATA		= 0x74,
+	HOST_TRANSITION_DETECT	= 0x75,
+	HOST_GPIO_CTRL		= 0x76,
+	HOST_I2C_ADDR		= 0x78, /* 32 bit */
+	HOST_I2C_DATA		= 0x7c,
+	HOST_I2C_XFER_CNT	= 0x7e,
+	HOST_I2C_CTRL		= 0x7f,
+
+	/* HOST_SLOT_STAT bits */
+	HOST_SSTAT_ATTN		= (1 << 31),
+
+	/*
+	 * Port registers
+	 * (8192 bytes @ +0x0000, +0x2000, +0x4000 and +0x6000 @ BAR2)
+	 */
+	PORT_REGS_SIZE		= 0x2000,
+	PORT_PRB		= 0x0000, /* (32 bytes PRB + 16 bytes SGEs * 6) * 31 (3968 bytes) */
+
+	PORT_PM			= 0x0f80, /* 8 bytes PM * 16 (128 bytes) */
+		/* 32 bit regs */
+	PORT_CTRL_STAT		= 0x1000, /* write: ctrl-set, read: stat */
+	PORT_CTRL_CLR		= 0x1004, /* write: ctrl-clear */
+	PORT_IRQ_STAT		= 0x1008, /* high: status, low: interrupt */
+	PORT_IRQ_ENABLE_SET	= 0x1010, /* write: enable-set */
+	PORT_IRQ_ENABLE_CLR	= 0x1014, /* write: enable-clear */
+	PORT_ACTIVATE_UPPER_ADDR= 0x101c,
+	PORT_EXEC_FIFO		= 0x1020, /* command execution fifo */
+	PORT_CMD_ERR		= 0x1024, /* command error number */
+	PORT_FIS_CFG		= 0x1028,
+	PORT_FIFO_THRES		= 0x102c,
+		/* 16 bit regs */
+	PORT_DECODE_ERR_CNT	= 0x1040,
+	PORT_DECODE_ERR_THRESH	= 0x1042,
+	PORT_CRC_ERR_CNT	= 0x1044,
+	PORT_CRC_ERR_THRESH	= 0x1046,
+	PORT_HSHK_ERR_CNT	= 0x1048,
+	PORT_HSHK_ERR_THRESH	= 0x104a,
+		/* 32 bit regs */
+	PORT_PHY_CFG		= 0x1050,
+	PORT_SLOT_STAT		= 0x1800,
+	PORT_CMD_ACTIVATE	= 0x1c00, /* 64 bit cmd activate * 31 (248 bytes) */
+	PORT_EXEC_DIAG		= 0x1e00, /* 32bit exec diag * 16 (64 bytes, 0-10 used on 3124) */
+	PORT_PSD_DIAG		= 0x1e40, /* 32bit psd diag * 16 (64 bytes, 0-8 used on 3124) */
+	PORT_SCONTROL		= 0x1f00,
+	PORT_SSTATUS		= 0x1f04,
+	PORT_SERROR		= 0x1f08,
+	PORT_SACTIVE		= 0x1f0c,
+
+	/* PORT_CTRL_STAT bits */
+	PORT_CS_PORT_RST	= (1 << 0), /* port reset */
+	PORT_CS_DEV_RST		= (1 << 1), /* device reset */
+	PORT_CS_INIT		= (1 << 2), /* port initialize */
+	PORT_CS_IRQ_WOC		= (1 << 3), /* interrupt write one to clear */
+	PORT_CS_RESUME		= (1 << 6), /* port resume */
+	PORT_CS_32BIT_ACTV	= (1 << 10), /* 32-bit activation */
+	PORT_CS_PM_EN		= (1 << 13), /* port multiplier enable */
+	PORT_CS_RDY		= (1 << 31), /* port ready to accept commands */
+
+	/* PORT_IRQ_STAT/ENABLE_SET/CLR */
+	/* bits[11:0] are masked */
+	PORT_IRQ_COMPLETE	= (1 << 0), /* command(s) completed */
+	PORT_IRQ_ERROR		= (1 << 1), /* command execution error */
+	PORT_IRQ_PORTRDY_CHG	= (1 << 2), /* port ready change */
+	PORT_IRQ_PWR_CHG	= (1 << 3), /* power management change */
+	PORT_IRQ_PHYRDY_CHG	= (1 << 4), /* PHY ready change */
+	PORT_IRQ_COMWAKE	= (1 << 5), /* COMWAKE received */
+	PORT_IRQ_UNK_FIS	= (1 << 6), /* Unknown FIS received */
+	PORT_IRQ_SDB_FIS	= (1 << 11), /* SDB FIS received */
+
+	/* bits[27:16] are unmasked (raw) */
+	PORT_IRQ_RAW_SHIFT	= 16,
+	PORT_IRQ_MASKED_MASK	= 0x7ff,
+	PORT_IRQ_RAW_MASK	= (0x7ff << PORT_IRQ_RAW_SHIFT),
+
+	/* ENABLE_SET/CLR specific, intr steering - 2 bit field */
+	PORT_IRQ_STEER_SHIFT	= 30,
+	PORT_IRQ_STEER_MASK	= (3 << PORT_IRQ_STEER_SHIFT),
+
+	/* PORT_CMD_ERR constants */
+	PORT_CERR_DEV		= 1, /* Error bit in D2H Register FIS */
+	PORT_CERR_SDB		= 2, /* Error bit in SDB FIS */
+	PORT_CERR_DATA		= 3, /* Error in data FIS not detected by dev */
+	PORT_CERR_SEND		= 4, /* Initial cmd FIS transmission failure */
+	PORT_CERR_INCONSISTENT	= 5, /* Protocol mismatch */
+	PORT_CERR_DIRECTION	= 6, /* Data direction mismatch */
+	PORT_CERR_UNDERRUN	= 7, /* Ran out of SGEs while writing */
+	PORT_CERR_OVERRUN	= 8, /* Ran out of SGEs while reading */
+	PORT_CERR_PKT_PROT	= 11, /* DIR invalid in 1st PIO setup of ATAPI */
+	PORT_CERR_SGT_BOUNDARY	= 16, /* PLD ecode 00 - SGT not on qword boundary */
+	PORT_CERR_SGT_TGTABRT	= 17, /* PLD ecode 01 - target abort */
+	PORT_CERR_SGT_MSTABRT	= 18, /* PLD ecode 10 - master abort */
+	PORT_CERR_SGT_PCIPERR	= 19, /* PLD ecode 11 - PCI parity err while fetching SGT */
+	PORT_CERR_CMD_BOUNDARY	= 24, /* ctrl[15:13] 001 - PRB not on qword boundary */
+	PORT_CERR_CMD_TGTABRT	= 25, /* ctrl[15:13] 010 - target abort */
+	PORT_CERR_CMD_MSTABRT	= 26, /* ctrl[15:13] 100 - master abort */
+	PORT_CERR_CMD_PCIPERR	= 27, /* ctrl[15:13] 110 - PCI parity err while fetching PRB */
+	PORT_CERR_XFR_UNDEF	= 32, /* PSD ecode 00 - undefined */
+	PORT_CERR_XFR_TGTABRT	= 33, /* PSD ecode 01 - target abort */
+	PORT_CERR_XFR_MSGABRT	= 34, /* PSD ecode 10 - master abort */
+	PORT_CERR_XFR_PCIPERR	= 35, /* PSD ecode 11 - PCI prity err during transfer */
+	PORT_CERR_SENDSERVICE	= 36, /* FIS received while sending service */
+
+	/*
+	 * Other constants
+	 */
+	SGE_TRM			= (1 << 31), /* Last SGE in chain */
+	PRB_SOFT_RST		= (1 << 7),  /* Soft reset request (ign BSY?) */
+
+	/* board id */
+	BID_SIL3124		= 0,
+	BID_SIL3132		= 1,
+	BID_SIL3131		= 2,
+
+	IRQ_STAT_4PORTS		= 0xf,
+};
+
+struct sil24_cmd_block {
+	struct sil24_prb prb;
+	struct sil24_sge sge[LIBATA_MAX_PRD];
+};
+
+/*
+ * ap->private_data
+ *
+ * The preview driver always returned 0 for status.  We emulate it
+ * here from the previous interrupt.
+ */
+struct sil24_port_priv {
+	struct sil24_cmd_block *cmd_block;	/* 32 cmd blocks */
+	dma_addr_t cmd_block_dma;		/* DMA base addr for them */
+	struct ata_taskfile tf;			/* Cached taskfile registers */
+};
+
+/* ap->host_set->private_data */
+struct sil24_host_priv {
+	void *host_base;	/* global controller control (128 bytes @BAR0) */
+	void *port_base;	/* port registers (4 * 8192 bytes @BAR2) */
+};
+
+static u8 sil24_check_status(struct ata_port *ap);
+static u8 sil24_check_err(struct ata_port *ap);
+static u32 sil24_scr_read(struct ata_port *ap, unsigned sc_reg);
+static void sil24_scr_write(struct ata_port *ap, unsigned sc_reg, u32 val);
+static void sil24_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
+static void sil24_phy_reset(struct ata_port *ap);
+static void sil24_qc_prep(struct ata_queued_cmd *qc);
+static int sil24_qc_issue(struct ata_queued_cmd *qc);
+static void sil24_irq_clear(struct ata_port *ap);
+static void sil24_eng_timeout(struct ata_port *ap);
+static irqreturn_t sil24_interrupt(int irq, void *dev_instance, struct pt_regs *regs);
+static int sil24_port_start(struct ata_port *ap);
+static void sil24_port_stop(struct ata_port *ap);
+static void sil24_host_stop(struct ata_host_set *host_set);
+static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
+
+static struct pci_device_id sil24_pci_tbl[] = {
+	{ 0x1095, 0x3124, PCI_ANY_ID, PCI_ANY_ID, 0, 0, BID_SIL3124 },
+	{ 0x1095, 0x3132, PCI_ANY_ID, PCI_ANY_ID, 0, 0, BID_SIL3132 },
+	{ 0x1095, 0x3131, PCI_ANY_ID, PCI_ANY_ID, 0, 0, BID_SIL3131 },
+	{ 0x1095, 0x3531, PCI_ANY_ID, PCI_ANY_ID, 0, 0, BID_SIL3131 },
+	{ } /* terminate list */
+};
+
+static struct pci_driver sil24_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= sil24_pci_tbl,
+	.probe			= sil24_init_one,
+	.remove			= ata_pci_remove_one, /* safe? */
+};
+
+static Scsi_Host_Template sil24_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.eh_strategy_handler	= ata_scsi_error,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+	.ordered_flush		= 1, /* NCQ not supported yet */
+};
+
+static struct ata_port_operations sil24_ops = {
+	.port_disable		= ata_port_disable,
+
+	.check_status		= sil24_check_status,
+	.check_altstatus	= sil24_check_status,
+	.check_err		= sil24_check_err,
+	.dev_select		= ata_noop_dev_select,
+
+	.tf_read		= sil24_tf_read,
+
+	.phy_reset		= sil24_phy_reset,
+
+	.qc_prep		= sil24_qc_prep,
+	.qc_issue		= sil24_qc_issue,
+
+	.eng_timeout		= sil24_eng_timeout,
+
+	.irq_handler		= sil24_interrupt,
+	.irq_clear		= sil24_irq_clear,
+
+	.scr_read		= sil24_scr_read,
+	.scr_write		= sil24_scr_write,
+
+	.port_start		= sil24_port_start,
+	.port_stop		= sil24_port_stop,
+	.host_stop		= sil24_host_stop,
+};
+
+/*
+ * Use bits 30-31 of host_flags to encode available port numbers.
+ * Current maxium is 4.
+ */
+#define SIL24_NPORTS2FLAG(nports)	((((unsigned)(nports) - 1) & 0x3) << 30)
+#define SIL24_FLAG2NPORTS(flag)		((((flag) >> 30) & 0x3) + 1)
+
+static struct ata_port_info sil24_port_info[] = {
+	/* sil_3124 */
+	{
+		.sht		= &sil24_sht,
+		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+				  ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO |
+				  ATA_FLAG_PIO_DMA | SIL24_NPORTS2FLAG(4),
+		.pio_mask	= 0x1f,			/* pio0-4 */
+		.mwdma_mask	= 0x07,			/* mwdma0-2 */
+		.udma_mask	= 0x3f,			/* udma0-5 */
+		.port_ops	= &sil24_ops,
+	},
+	/* sil_3132 */ 
+	{
+		.sht		= &sil24_sht,
+		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+				  ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO |
+				  ATA_FLAG_PIO_DMA | SIL24_NPORTS2FLAG(2),
+		.pio_mask	= 0x1f,			/* pio0-4 */
+		.mwdma_mask	= 0x07,			/* mwdma0-2 */
+		.udma_mask	= 0x3f,			/* udma0-5 */
+		.port_ops	= &sil24_ops,
+	},
+	/* sil_3131/sil_3531 */
+	{
+		.sht		= &sil24_sht,
+		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+				  ATA_FLAG_SATA_RESET | ATA_FLAG_MMIO |
+				  ATA_FLAG_PIO_DMA | SIL24_NPORTS2FLAG(1),
+		.pio_mask	= 0x1f,			/* pio0-4 */
+		.mwdma_mask	= 0x07,			/* mwdma0-2 */
+		.udma_mask	= 0x3f,			/* udma0-5 */
+		.port_ops	= &sil24_ops,
+	},
+};
+
+static inline void sil24_update_tf(struct ata_port *ap)
+{
+	struct sil24_port_priv *pp = ap->private_data;
+	void *port = (void *)ap->ioaddr.cmd_addr;
+	struct sil24_prb *prb = port;
+
+	ata_tf_from_fis(prb->fis, &pp->tf);
+}
+
+static u8 sil24_check_status(struct ata_port *ap)
+{
+	struct sil24_port_priv *pp = ap->private_data;
+	return pp->tf.command;
+}
+
+static u8 sil24_check_err(struct ata_port *ap)
+{
+	struct sil24_port_priv *pp = ap->private_data;
+	return pp->tf.feature;
+}
+
+static int sil24_scr_map[] = {
+	[SCR_CONTROL]	= 0,
+	[SCR_STATUS]	= 1,
+	[SCR_ERROR]	= 2,
+	[SCR_ACTIVE]	= 3,
+};
+
+static u32 sil24_scr_read(struct ata_port *ap, unsigned sc_reg)
+{
+	void *scr_addr = (void *)ap->ioaddr.scr_addr;
+	if (sc_reg < ARRAY_SIZE(sil24_scr_map)) {
+		void *addr;
+		addr = scr_addr + sil24_scr_map[sc_reg] * 4;
+		return readl(scr_addr + sil24_scr_map[sc_reg] * 4);
+	}
+	return 0xffffffffU;
+}
+
+static void sil24_scr_write(struct ata_port *ap, unsigned sc_reg, u32 val)
+{
+	void *scr_addr = (void *)ap->ioaddr.scr_addr;
+	if (sc_reg < ARRAY_SIZE(sil24_scr_map)) {
+		void *addr;
+		addr = scr_addr + sil24_scr_map[sc_reg] * 4;
+		writel(val, scr_addr + sil24_scr_map[sc_reg] * 4);
+	}
+}
+
+static void sil24_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
+{
+	struct sil24_port_priv *pp = ap->private_data;
+	*tf = pp->tf;
+}
+
+static void sil24_phy_reset(struct ata_port *ap)
+{
+	__sata_phy_reset(ap);
+	/*
+	 * No ATAPI yet.  Just unconditionally indicate ATA device.
+	 * If ATAPI device is attached, it will fail ATA_CMD_ID_ATA
+	 * and libata core will ignore the device.
+	 */
+	if (!(ap->flags & ATA_FLAG_PORT_DISABLED))
+		ap->device[0].class = ATA_DEV_ATA;
+}
+
+static inline void sil24_fill_sg(struct ata_queued_cmd *qc,
+				 struct sil24_cmd_block *cb)
+{
+	struct scatterlist *sg = qc->sg;
+	struct sil24_sge *sge = cb->sge;
+	unsigned i;
+
+	for (i = 0; i < qc->n_elem; i++, sg++, sge++) {
+		sge->addr = cpu_to_le64(sg_dma_address(sg));
+		sge->cnt = cpu_to_le32(sg_dma_len(sg));
+		sge->flags = 0;
+		sge->flags = i < qc->n_elem - 1 ? 0 : cpu_to_le32(SGE_TRM);
+	}
+}
+
+static void sil24_qc_prep(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct sil24_port_priv *pp = ap->private_data;
+	struct sil24_cmd_block *cb = pp->cmd_block + qc->tag;
+	struct sil24_prb *prb = &cb->prb;
+
+	switch (qc->tf.protocol) {
+	case ATA_PROT_PIO:
+	case ATA_PROT_DMA:
+	case ATA_PROT_NODATA:
+		break;
+	default:
+		/* ATAPI isn't supported yet */
+		BUG();
+	}
+
+	ata_tf_to_fis(&qc->tf, prb->fis, 0);
+
+	if (qc->flags & ATA_QCFLAG_DMAMAP)
+		sil24_fill_sg(qc, cb);
+}
+
+static int sil24_qc_issue(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	void *port = (void *)ap->ioaddr.cmd_addr;
+	struct sil24_port_priv *pp = ap->private_data;
+	dma_addr_t paddr = pp->cmd_block_dma + qc->tag * sizeof(*pp->cmd_block);
+
+	writel((u32)paddr, port + PORT_CMD_ACTIVATE);
+	return 0;
+}
+
+static void sil24_irq_clear(struct ata_port *ap)
+{
+	/* unused */
+}
+
+static int __sil24_reset_controller(void *port)
+{
+	int cnt;
+	u32 tmp;
+
+	/* Reset controller state.  Is this correct? */
+	writel(PORT_CS_DEV_RST, port + PORT_CTRL_STAT);
+	readl(port + PORT_CTRL_STAT);	/* sync */
+
+	/* Max ~100ms */
+	for (cnt = 0; cnt < 1000; cnt++) {
+		udelay(100);
+		tmp = readl(port + PORT_CTRL_STAT);
+		if (!(tmp & PORT_CS_DEV_RST))
+			break;
+	}
+
+	if (tmp & PORT_CS_DEV_RST)
+		return -1;
+	return 0;
+}
+
+static void sil24_reset_controller(struct ata_port *ap)
+{
+	printk(KERN_NOTICE DRV_NAME
+	       " ata%u: resetting controller...\n", ap->id);
+	if (__sil24_reset_controller((void *)ap->ioaddr.cmd_addr))
+                printk(KERN_ERR DRV_NAME
+                       " ata%u: failed to reset controller\n", ap->id);
+}
+
+static void sil24_eng_timeout(struct ata_port *ap)
+{
+	struct ata_queued_cmd *qc;
+
+	qc = ata_qc_from_tag(ap, ap->active_tag);
+	if (!qc) {
+		printk(KERN_ERR "ata%u: BUG: tiemout without command\n",
+		       ap->id);
+		return;
+	}
+
+	/*
+	 * hack alert!  We cannot use the supplied completion
+	 * function from inside the ->eh_strategy_handler() thread.
+	 * libata is the only user of ->eh_strategy_handler() in
+	 * any kernel, so the default scsi_done() assumes it is
+	 * not being called from the SCSI EH.
+	 */
+	printk(KERN_ERR "ata%u: command timeout\n", ap->id);
+	qc->scsidone = scsi_finish_command;
+	ata_qc_complete(qc, ATA_ERR);
+
+	sil24_reset_controller(ap);
+}
+
+static void sil24_error_intr(struct ata_port *ap, u32 slot_stat)
+{
+	struct ata_queued_cmd *qc = ata_qc_from_tag(ap, ap->active_tag);
+	struct sil24_port_priv *pp = ap->private_data;
+	void *port = (void *)ap->ioaddr.cmd_addr;
+	u32 irq_stat, cmd_err, sstatus, serror;
+
+	irq_stat = readl(port + PORT_IRQ_STAT);
+	writel(irq_stat, port + PORT_IRQ_STAT);		/* clear irq */
+
+	if (!(irq_stat & PORT_IRQ_ERROR)) {
+		/* ignore non-completion, non-error irqs for now */
+		printk(KERN_WARNING DRV_NAME
+		       "ata%u: non-error exception irq (irq_stat %x)\n",
+		       ap->id, irq_stat);
+		return;
+	}
+
+	cmd_err = readl(port + PORT_CMD_ERR);
+	sstatus = readl(port + PORT_SSTATUS);
+	serror = readl(port + PORT_SERROR);
+	if (serror)
+		writel(serror, port + PORT_SERROR);
+
+	printk(KERN_ERR DRV_NAME " ata%u: error interrupt on port%d\n"
+	       "  stat=0x%x irq=0x%x cmd_err=%d sstatus=0x%x serror=0x%x\n",
+	       ap->id, ap->port_no, slot_stat, irq_stat, cmd_err, sstatus, serror);
+
+	if (cmd_err == PORT_CERR_DEV || cmd_err == PORT_CERR_SDB) {
+		/*
+		 * Device is reporting error, tf registers are valid.
+		 */
+		sil24_update_tf(ap);
+	} else {
+		/*
+		 * Other errors.  libata currently doesn't have any
+		 * mechanism to report these errors.  Just turn on
+		 * ATA_ERR.
+		 */
+		pp->tf.command = ATA_ERR;
+	}
+
+	if (qc)
+		ata_qc_complete(qc, pp->tf.command);
+
+	sil24_reset_controller(ap);
+}
+
+static inline void sil24_host_intr(struct ata_port *ap)
+{
+	struct ata_queued_cmd *qc = ata_qc_from_tag(ap, ap->active_tag);
+	void *port = (void *)ap->ioaddr.cmd_addr;
+	u32 slot_stat;
+
+	slot_stat = readl(port + PORT_SLOT_STAT);
+	if (!(slot_stat & HOST_SSTAT_ATTN)) {
+		struct sil24_port_priv *pp = ap->private_data;
+		/*
+		 * !HOST_SSAT_ATTN guarantees successful completion,
+		 * so reading back tf registers is unnecessary for
+		 * most commands.  TODO: read tf registers for
+		 * commands which require these values on successful
+		 * completion (EXECUTE DEVICE DIAGNOSTIC, CHECK POWER,
+		 * DEVICE RESET and READ PORT MULTIPLIER (any more?).
+		 */
+		sil24_update_tf(ap);
+
+		if (qc)
+			ata_qc_complete(qc, pp->tf.command);
+	} else
+		sil24_error_intr(ap, slot_stat);
+}
+
+static irqreturn_t sil24_interrupt(int irq, void *dev_instance, struct pt_regs *regs)
+{
+	struct ata_host_set *host_set = dev_instance;
+	struct sil24_host_priv *hpriv = host_set->private_data;
+	unsigned handled = 0;
+	u32 status;
+	int i;
+
+	status = readl(hpriv->host_base + HOST_IRQ_STAT);
+
+	if (status == 0xffffffff) {
+		printk(KERN_ERR DRV_NAME ": IRQ status == 0xffffffff, "
+		       "PCI fault or device removal?\n");
+		goto out;
+	}
+
+	if (!(status & IRQ_STAT_4PORTS))
+		goto out;
+
+	spin_lock(&host_set->lock);
+
+	for (i = 0; i < host_set->n_ports; i++)
+		if (status & (1 << i)) {
+			struct ata_port *ap = host_set->ports[i];
+			if (ap && !(ap->flags & ATA_FLAG_PORT_DISABLED)) {
+				sil24_host_intr(host_set->ports[i]);
+				handled++;
+			} else
+				printk(KERN_ERR DRV_NAME
+				       ": interrupt from disabled port %d\n", i);
+		}
+
+	spin_unlock(&host_set->lock);
+ out:
+	return IRQ_RETVAL(handled);
+}
+
+static int sil24_port_start(struct ata_port *ap)
+{
+	struct device *dev = ap->host_set->dev;
+	struct sil24_port_priv *pp;
+	struct sil24_cmd_block *cb;
+	size_t cb_size = sizeof(*cb);
+	dma_addr_t cb_dma;
+
+	pp = kmalloc(sizeof(*pp), GFP_KERNEL);
+	if (!pp)
+		return -ENOMEM;
+	memset(pp, 0, sizeof(*pp));
+
+	pp->tf.command = ATA_DRDY;
+
+	cb = dma_alloc_coherent(dev, cb_size, &cb_dma, GFP_KERNEL);
+	if (!cb) {
+		kfree(pp);
+		return -ENOMEM;
+	}
+	memset(cb, 0, cb_size);
+
+	pp->cmd_block = cb;
+	pp->cmd_block_dma = cb_dma;
+
+	ap->private_data = pp;
+
+	return 0;
+}
+
+static void sil24_port_stop(struct ata_port *ap)
+{
+	struct device *dev = ap->host_set->dev;
+	struct sil24_port_priv *pp = ap->private_data;
+	size_t cb_size = sizeof(*pp->cmd_block);
+
+	dma_free_coherent(dev, cb_size, pp->cmd_block, pp->cmd_block_dma);
+	kfree(pp);
+}
+
+static void sil24_host_stop(struct ata_host_set *host_set)
+{
+	struct sil24_host_priv *hpriv = host_set->private_data;
+
+	iounmap(hpriv->host_base);
+	iounmap(hpriv->port_base);
+	kfree(hpriv);
+}
+
+static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version = 0;
+	unsigned int board_id = (unsigned int)ent->driver_data;
+	struct ata_port_info *pinfo = &sil24_port_info[board_id];
+	struct ata_probe_ent *probe_ent = NULL;
+	struct sil24_host_priv *hpriv = NULL;
+	void *host_base = NULL, *port_base = NULL;
+	int i, rc;
+
+	if (!printed_version++)
+		printk(KERN_DEBUG DRV_NAME " version " DRV_VERSION "\n");
+
+	rc = pci_enable_device(pdev);
+	if (rc)
+		return rc;
+
+	rc = pci_request_regions(pdev, DRV_NAME);
+	if (rc)
+		goto out_disable;
+
+	rc = -ENOMEM;
+	/* ioremap mmio registers */
+	host_base = ioremap(pci_resource_start(pdev, 0),
+			    pci_resource_len(pdev, 0));
+	if (!host_base)
+		goto out_free;
+	port_base = ioremap(pci_resource_start(pdev, 2),
+			    pci_resource_len(pdev, 2));
+	if (!port_base)
+		goto out_free;
+
+	/* allocate & init probe_ent and hpriv */
+	probe_ent = kmalloc(sizeof(*probe_ent), GFP_KERNEL);
+	if (!probe_ent)
+		goto out_free;
+
+	hpriv = kmalloc(sizeof(*hpriv), GFP_KERNEL);
+	if (!hpriv)
+		goto out_free;
+
+	memset(probe_ent, 0, sizeof(*probe_ent));
+	probe_ent->dev = pci_dev_to_dev(pdev);
+	INIT_LIST_HEAD(&probe_ent->node);
+
+	probe_ent->sht		= pinfo->sht;
+	probe_ent->host_flags	= pinfo->host_flags;
+	probe_ent->pio_mask	= pinfo->pio_mask;
+	probe_ent->udma_mask	= pinfo->udma_mask;
+	probe_ent->port_ops	= pinfo->port_ops;
+	probe_ent->n_ports	= SIL24_FLAG2NPORTS(pinfo->host_flags);
+
+	probe_ent->irq = pdev->irq;
+	probe_ent->irq_flags = SA_SHIRQ;
+	probe_ent->mmio_base = port_base;
+	probe_ent->private_data = hpriv;
+
+	memset(hpriv, 0, sizeof(*hpriv));
+	hpriv->host_base = host_base;
+	hpriv->port_base = port_base;
+
+	/*
+	 * Configure the device
+	 */
+	/*
+	 * FIXME: This device is certainly 64-bit capable.  We just
+	 * don't know how to use it.  After fixing 32bit activation in
+	 * this function, enable 64bit masks here.
+	 */
+	rc = pci_set_dma_mask(pdev, DMA_32BIT_MASK);
+	if (rc) {
+		printk(KERN_ERR DRV_NAME "(%s): 32-bit DMA enable failed\n",
+		       pci_name(pdev));
+		goto out_free;
+	}
+	rc = pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK);
+	if (rc) {
+		printk(KERN_ERR DRV_NAME "(%s): 32-bit consistent DMA enable failed\n",
+		       pci_name(pdev));
+		goto out_free;
+	}
+
+	/* GPIO off */
+	writel(0, host_base + HOST_FLASH_CMD);
+
+	/* Mask interrupts during initialization */
+	writel(0, host_base + HOST_CTRL);
+
+	for (i = 0; i < probe_ent->n_ports; i++) {
+		void *port = port_base + i * PORT_REGS_SIZE;
+		unsigned long portu = (unsigned long)port;
+		u32 tmp;
+		int cnt;
+
+		probe_ent->port[i].cmd_addr = portu + PORT_PRB;
+		probe_ent->port[i].scr_addr = portu + PORT_SCONTROL;
+
+		ata_std_ports(&probe_ent->port[i]);
+
+		/* Initial PHY setting */
+		writel(0x20c, port + PORT_PHY_CFG);
+
+		/* Clear port RST */
+		tmp = readl(port + PORT_CTRL_STAT);
+		if (tmp & PORT_CS_PORT_RST) {
+			writel(PORT_CS_PORT_RST, port + PORT_CTRL_CLR);
+			readl(port + PORT_CTRL_STAT);	/* sync */
+			for (cnt = 0; cnt < 10; cnt++) {
+				msleep(10);
+				tmp = readl(port + PORT_CTRL_STAT);
+				if (!(tmp & PORT_CS_PORT_RST))
+					break;
+			}
+			if (tmp & PORT_CS_PORT_RST)
+				printk(KERN_ERR DRV_NAME
+				       "(%s): failed to clear port RST\n",
+				       pci_name(pdev));
+		}
+
+		/* Zero error counters. */
+		writel(0x8000, port + PORT_DECODE_ERR_THRESH);
+		writel(0x8000, port + PORT_CRC_ERR_THRESH);
+		writel(0x8000, port + PORT_HSHK_ERR_THRESH);
+		writel(0x0000, port + PORT_DECODE_ERR_CNT);
+		writel(0x0000, port + PORT_CRC_ERR_CNT);
+		writel(0x0000, port + PORT_HSHK_ERR_CNT);
+
+		/* FIXME: 32bit activation? */
+		writel(0, port + PORT_ACTIVATE_UPPER_ADDR);
+		writel(PORT_CS_32BIT_ACTV, port + PORT_CTRL_STAT);
+
+		/* Configure interrupts */
+		writel(0xffff, port + PORT_IRQ_ENABLE_CLR);
+		writel(PORT_IRQ_COMPLETE | PORT_IRQ_ERROR | PORT_IRQ_SDB_FIS,
+		       port + PORT_IRQ_ENABLE_SET);
+
+		/* Clear interrupts */
+		writel(0x0fff0fff, port + PORT_IRQ_STAT);
+		writel(PORT_CS_IRQ_WOC, port + PORT_CTRL_CLR);
+
+		/* Clear port multiplier enable and resume bits */
+		writel(PORT_CS_PM_EN | PORT_CS_RESUME, port + PORT_CTRL_CLR);
+
+		/* Reset itself */
+		if (__sil24_reset_controller(port))
+			printk(KERN_ERR DRV_NAME
+			       "(%s): failed to reset controller\n",
+			       pci_name(pdev));
+	}
+
+	/* Turn on interrupts */
+	writel(IRQ_STAT_4PORTS, host_base + HOST_CTRL);
+
+	pci_set_master(pdev);
+
+	/* FIXME: check ata_device_add return value */
+	ata_device_add(probe_ent);
+
+	kfree(probe_ent);
+	return 0;
+
+ out_free:
+	if (host_base)
+		iounmap(host_base);
+	if (port_base)
+		iounmap(port_base);
+	kfree(probe_ent);
+	kfree(hpriv);
+	pci_release_regions(pdev);
+ out_disable:
+	pci_disable_device(pdev);
+	return rc;
+}
+
+static int __init sil24_init(void)
+{
+	return pci_module_init(&sil24_pci_driver);
+}
+
+static void __exit sil24_exit(void)
+{
+	pci_unregister_driver(&sil24_pci_driver);
+}
+
+MODULE_AUTHOR("Tejun Heo");
+MODULE_DESCRIPTION("Silicon Image 3124/3132 SATA low-level driver");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, sil24_pci_tbl);
+
+module_init(sil24_init);
+module_exit(sil24_exit);
diff --git a/drivers/scsi/sata_sis.c b/drivers/scsi/sata_sis.c
index b227e51..0761a32 100644
--- a/drivers/scsi/sata_sis.c
+++ b/drivers/scsi/sata_sis.c
@@ -263,7 +263,7 @@
 		goto err_out_regions;
 
 	ppi = &sis_port_info;
-	probe_ent = ata_pci_init_native_mode(pdev, &ppi);
+	probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
 	if (!probe_ent) {
 		rc = -ENOMEM;
 		goto err_out_regions;
diff --git a/drivers/scsi/sata_uli.c b/drivers/scsi/sata_uli.c
index 4c9fb8b..9c06f2a 100644
--- a/drivers/scsi/sata_uli.c
+++ b/drivers/scsi/sata_uli.c
@@ -202,7 +202,7 @@
 		goto err_out_regions;
 
 	ppi = &uli_port_info;
-	probe_ent = ata_pci_init_native_mode(pdev, &ppi);
+	probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
 	if (!probe_ent) {
 		rc = -ENOMEM;
 		goto err_out_regions;
diff --git a/drivers/scsi/sata_via.c b/drivers/scsi/sata_via.c
index 128b996..5658724 100644
--- a/drivers/scsi/sata_via.c
+++ b/drivers/scsi/sata_via.c
@@ -212,7 +212,7 @@
 	struct ata_probe_ent *probe_ent;
 	struct ata_port_info *ppi = &svia_port_info;
 
-	probe_ent = ata_pci_init_native_mode(pdev, &ppi);
+	probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
 	if (!probe_ent)
 		return NULL;
 
diff --git a/drivers/serial/sunsu.c b/drivers/serial/sunsu.c
index 5959e67..656c0e8 100644
--- a/drivers/serial/sunsu.c
+++ b/drivers/serial/sunsu.c
@@ -518,11 +518,7 @@
 
 	quot = up->port.uartclk / (16 * new_baud);
 
-	spin_unlock(&up->port.lock);
-
 	sunsu_change_speed(&up->port, up->cflag, 0, quot);
-
-	spin_lock(&up->port.lock);
 }
 
 static void receive_kbd_ms_chars(struct uart_sunsu_port *up, struct pt_regs *regs, int is_break)
diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c
index e240c33..5af928f 100644
--- a/fs/bfs/dir.c
+++ b/fs/bfs/dir.c
@@ -108,7 +108,7 @@
 	inode->i_mapping->a_ops = &bfs_aops;
 	inode->i_mode = mode;
 	inode->i_ino = ino;
-	BFS_I(inode)->i_dsk_ino = cpu_to_le16(ino);
+	BFS_I(inode)->i_dsk_ino = ino;
 	BFS_I(inode)->i_sblock = 0;
 	BFS_I(inode)->i_eblock = 0;
 	insert_inode_hash(inode);
diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c
index c7b39aa2..3af6c73 100644
--- a/fs/bfs/inode.c
+++ b/fs/bfs/inode.c
@@ -357,28 +357,46 @@
 	}
 
 	info->si_blocks = (le32_to_cpu(bfs_sb->s_end) + 1)>>BFS_BSIZE_BITS; /* for statfs(2) */
-	info->si_freeb = (le32_to_cpu(bfs_sb->s_end) + 1 -  cpu_to_le32(bfs_sb->s_start))>>BFS_BSIZE_BITS;
+	info->si_freeb = (le32_to_cpu(bfs_sb->s_end) + 1 -  le32_to_cpu(bfs_sb->s_start))>>BFS_BSIZE_BITS;
 	info->si_freei = 0;
 	info->si_lf_eblk = 0;
 	info->si_lf_sblk = 0;
 	info->si_lf_ioff = 0;
+	bh = NULL;
 	for (i=BFS_ROOT_INO; i<=info->si_lasti; i++) {
-		inode = iget(s,i);
-		if (BFS_I(inode)->i_dsk_ino == 0)
-			info->si_freei++;
-		else {
-			set_bit(i, info->si_imap);
-			info->si_freeb -= inode->i_blocks;
-			if (BFS_I(inode)->i_eblock > info->si_lf_eblk) {
-				info->si_lf_eblk = BFS_I(inode)->i_eblock;
-				info->si_lf_sblk = BFS_I(inode)->i_sblock;
-				info->si_lf_ioff = BFS_INO2OFF(i);
-			}
+		struct bfs_inode *di;
+		int block = (i - BFS_ROOT_INO)/BFS_INODES_PER_BLOCK + 1;
+		int off = (i - BFS_ROOT_INO) % BFS_INODES_PER_BLOCK;
+		unsigned long sblock, eblock;
+
+		if (!off) {
+			brelse(bh);
+			bh = sb_bread(s, block);
 		}
-		iput(inode);
+
+		if (!bh)
+			continue;
+
+		di = (struct bfs_inode *)bh->b_data + off;
+
+		if (!di->i_ino) {
+			info->si_freei++;
+			continue;
+		}
+		set_bit(i, info->si_imap);
+		info->si_freeb -= BFS_FILEBLOCKS(di);
+
+		sblock =  le32_to_cpu(di->i_sblock);
+		eblock =  le32_to_cpu(di->i_eblock);
+		if (eblock > info->si_lf_eblk) {
+			info->si_lf_eblk = eblock;
+			info->si_lf_sblk = sblock;
+			info->si_lf_ioff = BFS_INO2OFF(i);
+		}
 	}
+	brelse(bh);
 	if (!(s->s_flags & MS_RDONLY)) {
-		mark_buffer_dirty(bh);
+		mark_buffer_dirty(info->si_sbh);
 		s->s_dirt = 1;
 	} 
 	dump_imap("read_super", s);
diff --git a/fs/namei.c b/fs/namei.c
index 043d587..aa62dbd 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1551,19 +1551,19 @@
 	if (nd->last_type != LAST_NORM)
 		goto exit;
 	if (nd->last.name[nd->last.len]) {
-		putname(nd->last.name);
+		__putname(nd->last.name);
 		goto exit;
 	}
 	error = -ELOOP;
 	if (count++==32) {
-		putname(nd->last.name);
+		__putname(nd->last.name);
 		goto exit;
 	}
 	dir = nd->dentry;
 	down(&dir->d_inode->i_sem);
 	path.dentry = __lookup_hash(&nd->last, nd->dentry, nd);
 	path.mnt = nd->mnt;
-	putname(nd->last.name);
+	__putname(nd->last.name);
 	goto do_last;
 }
 
diff --git a/fs/ntfs/ChangeLog b/fs/ntfs/ChangeLog
index 83f3322..de58579 100644
--- a/fs/ntfs/ChangeLog
+++ b/fs/ntfs/ChangeLog
@@ -102,6 +102,9 @@
 	  inode instead of a vfs inode as parameter.
 	- Fix the definition of the CHKD ntfs record magic.  It had an off by
 	  two error causing it to be CHKB instead of CHKD.
+	- Fix a stupid bug in __ntfs_bitmap_set_bits_in_run() which caused the
+	  count to become negative and hence we had a wild memset() scribbling
+	  all over the system's ram.
 
 2.1.23 - Implement extension of resident files and make writing safe as well as
 	 many bug fixes, cleanups, and enhancements...
diff --git a/fs/ntfs/bitmap.c b/fs/ntfs/bitmap.c
index 12cf2e3..7a190cd 100644
--- a/fs/ntfs/bitmap.c
+++ b/fs/ntfs/bitmap.c
@@ -1,7 +1,7 @@
 /*
  * bitmap.c - NTFS kernel bitmap handling.  Part of the Linux-NTFS project.
  *
- * Copyright (c) 2004 Anton Altaparmakov
+ * Copyright (c) 2004-2005 Anton Altaparmakov
  *
  * This program/include file is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License as published
@@ -90,7 +90,8 @@
 	/* If the first byte is partial, modify the appropriate bits in it. */
 	if (bit) {
 		u8 *byte = kaddr + pos;
-		while ((bit & 7) && cnt--) {
+		while ((bit & 7) && cnt) {
+			cnt--;
 			if (value)
 				*byte |= 1 << bit++;
 			else
diff --git a/fs/ntfs/layout.h b/fs/ntfs/layout.h
index 01f2dfa..5c248d4 100644
--- a/fs/ntfs/layout.h
+++ b/fs/ntfs/layout.h
@@ -309,7 +309,7 @@
  * Note: The _LE versions will return a CPU endian formatted value!
  */
 #define MFT_REF_MASK_CPU 0x0000ffffffffffffULL
-#define MFT_REF_MASK_LE const_cpu_to_le64(0x0000ffffffffffffULL)
+#define MFT_REF_MASK_LE const_cpu_to_le64(MFT_REF_MASK_CPU)
 
 typedef u64 MFT_REF;
 typedef le64 leMFT_REF;
diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c
index 247586d..b011369 100644
--- a/fs/ntfs/mft.c
+++ b/fs/ntfs/mft.c
@@ -58,7 +58,8 @@
 	 * overflowing the unsigned long, but I don't think we would ever get
 	 * here if the volume was that big...
 	 */
-	index = ni->mft_no << vol->mft_record_size_bits >> PAGE_CACHE_SHIFT;
+	index = (u64)ni->mft_no << vol->mft_record_size_bits >>
+			PAGE_CACHE_SHIFT;
 	ofs = (ni->mft_no << vol->mft_record_size_bits) & ~PAGE_CACHE_MASK;
 
 	i_size = i_size_read(mft_vi);
diff --git a/fs/ntfs/unistr.c b/fs/ntfs/unistr.c
index a389a5a..0ea887f 100644
--- a/fs/ntfs/unistr.c
+++ b/fs/ntfs/unistr.c
@@ -1,7 +1,7 @@
 /*
  * unistr.c - NTFS Unicode string handling. Part of the Linux-NTFS project.
  *
- * Copyright (c) 2001-2004 Anton Altaparmakov
+ * Copyright (c) 2001-2005 Anton Altaparmakov
  *
  * This program/include file is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License as published
diff --git a/include/asm-arm/arch-h720x/system.h b/include/asm-arm/arch-h720x/system.h
index 0b025e2..09eda84 100644
--- a/include/asm-arm/arch-h720x/system.h
+++ b/include/asm-arm/arch-h720x/system.h
@@ -17,9 +17,11 @@
 static void arch_idle(void)
 {
 	CPU_REG (PMU_BASE, PMU_MODE) = PMU_MODE_IDLE;
-	__asm__ __volatile__(
-	"mov	r0, r0\n\t"
-	"mov	r0, r0");
+	nop();
+	nop();
+	CPU_REG (PMU_BASE, PMU_MODE) = PMU_MODE_RUN;
+	nop();
+	nop();
 }
 
 
diff --git a/include/asm-arm/arch-imx/imx-regs.h b/include/asm-arm/arch-imx/imx-regs.h
index 93b840e..229f700 100644
--- a/include/asm-arm/arch-imx/imx-regs.h
+++ b/include/asm-arm/arch-imx/imx-regs.h
@@ -76,6 +76,7 @@
 #define GPIO_PIN_MASK 0x1f
 #define GPIO_PORT_MASK (0x3 << 5)
 
+#define GPIO_PORT_SHIFT 5
 #define GPIO_PORTA (0<<5)
 #define GPIO_PORTB (1<<5)
 #define GPIO_PORTC (2<<5)
@@ -88,24 +89,37 @@
 #define GPIO_PF    (0<<9)
 #define GPIO_AF    (1<<9)
 
+#define GPIO_OCR_SHIFT 10
 #define GPIO_OCR_MASK (3<<10)
 #define GPIO_AIN   (0<<10)
 #define GPIO_BIN   (1<<10)
 #define GPIO_CIN   (2<<10)
-#define GPIO_GPIO  (3<<10)
+#define GPIO_DR    (3<<10)
 
-#define GPIO_AOUT  (1<<12)
-#define GPIO_BOUT  (1<<13)
+#define GPIO_AOUT_SHIFT 12
+#define GPIO_AOUT_MASK (3<<12)
+#define GPIO_AOUT     (0<<12)
+#define GPIO_AOUT_ISR (1<<12)
+#define GPIO_AOUT_0   (2<<12)
+#define GPIO_AOUT_1   (3<<12)
+
+#define GPIO_BOUT_SHIFT 14
+#define GPIO_BOUT_MASK (3<<14)
+#define GPIO_BOUT      (0<<14)
+#define GPIO_BOUT_ISR  (1<<14)
+#define GPIO_BOUT_0    (2<<14)
+#define GPIO_BOUT_1    (3<<14)
+
+#define GPIO_GIUS      (1<<16)
 
 /* assignements for GPIO alternate/primary functions */
 
 /* FIXME: This list is not completed. The correct directions are
  * missing on some (many) pins
  */
-#define PA0_PF_A24           ( GPIO_PORTA | GPIO_PF | 0 )
-#define PA0_AIN_SPI2_CLK     ( GPIO_PORTA | GPIO_OUT | GPIO_AIN | 0 )
+#define PA0_AIN_SPI2_CLK     ( GPIO_GIUS | GPIO_PORTA | GPIO_OUT | 0 )
 #define PA0_AF_ETMTRACESYNC  ( GPIO_PORTA | GPIO_AF | 0 )
-#define PA1_AOUT_SPI2_RXD    ( GPIO_PORTA | GPIO_IN | GPIO_AOUT | 1 )
+#define PA1_AOUT_SPI2_RXD    ( GPIO_GIUS | GPIO_PORTA | GPIO_IN | 1 )
 #define PA1_PF_TIN           ( GPIO_PORTA | GPIO_PF | 1 )
 #define PA2_PF_PWM0          ( GPIO_PORTA | GPIO_OUT | GPIO_PF | 2 )
 #define PA3_PF_CSI_MCLK      ( GPIO_PORTA | GPIO_PF | 3 )
@@ -123,7 +137,7 @@
 #define PA15_PF_I2C_SDA      ( GPIO_PORTA | GPIO_OUT | GPIO_PF | 15 )
 #define PA16_PF_I2C_SCL      ( GPIO_PORTA | GPIO_OUT | GPIO_PF | 16 )
 #define PA17_AF_ETMTRACEPKT4 ( GPIO_PORTA | GPIO_AF | 17 )
-#define PA17_AIN_SPI2_SS     ( GPIO_PORTA | GPIO_AIN | 17 )
+#define PA17_AIN_SPI2_SS     ( GPIO_GIUS | GPIO_PORTA | GPIO_OUT | 17 )
 #define PA18_AF_ETMTRACEPKT5 ( GPIO_PORTA | GPIO_AF | 18 )
 #define PA19_AF_ETMTRACEPKT6 ( GPIO_PORTA | GPIO_AF | 19 )
 #define PA20_AF_ETMTRACEPKT7 ( GPIO_PORTA | GPIO_AF | 20 )
@@ -191,19 +205,27 @@
 #define PC15_PF_SPI1_SS      ( GPIO_PORTC | GPIO_PF | 15 )
 #define PC16_PF_SPI1_MISO    ( GPIO_PORTC | GPIO_PF | 16 )
 #define PC17_PF_SPI1_MOSI    ( GPIO_PORTC | GPIO_PF | 17 )
+#define PC24_BIN_UART3_RI    ( GPIO_GIUS | GPIO_PORTC | GPIO_OUT | GPIO_BIN | 24 )
+#define PC25_BIN_UART3_DSR   ( GPIO_GIUS | GPIO_PORTC | GPIO_OUT | GPIO_BIN | 25 )
+#define PC26_AOUT_UART3_DTR  ( GPIO_GIUS | GPIO_PORTC | GPIO_IN | 26 )
+#define PC27_BIN_UART3_DCD   ( GPIO_GIUS | GPIO_PORTC | GPIO_OUT | GPIO_BIN | 27 )
+#define PC28_BIN_UART3_CTS   ( GPIO_GIUS | GPIO_PORTC | GPIO_OUT | GPIO_BIN | 28 )
+#define PC29_AOUT_UART3_RTS  ( GPIO_GIUS | GPIO_PORTC | GPIO_IN | 29 )
+#define PC30_BIN_UART3_TX    ( GPIO_GIUS | GPIO_PORTC | GPIO_BIN | 30 )
+#define PC31_AOUT_UART3_RX   ( GPIO_GIUS | GPIO_PORTC | GPIO_IN | 31)
 #define PD6_PF_LSCLK         ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 6 )
 #define PD7_PF_REV           ( GPIO_PORTD | GPIO_PF | 7 )
 #define PD7_AF_UART2_DTR     ( GPIO_PORTD | GPIO_IN | GPIO_AF | 7 )
-#define PD7_AIN_SPI2_SCLK    ( GPIO_PORTD | GPIO_AIN | 7 )
+#define PD7_AIN_SPI2_SCLK    ( GPIO_GIUS | GPIO_PORTD | GPIO_AIN | 7 )
 #define PD8_PF_CLS           ( GPIO_PORTD | GPIO_PF | 8 )
 #define PD8_AF_UART2_DCD     ( GPIO_PORTD | GPIO_OUT | GPIO_AF | 8 )
-#define PD8_AIN_SPI2_SS      ( GPIO_PORTD | GPIO_AIN | 8 )
+#define PD8_AIN_SPI2_SS      ( GPIO_GIUS | GPIO_PORTD | GPIO_AIN | 8 )
 #define PD9_PF_PS            ( GPIO_PORTD | GPIO_PF | 9 )
 #define PD9_AF_UART2_RI      ( GPIO_PORTD | GPIO_OUT | GPIO_AF | 9 )
-#define PD9_AOUT_SPI2_RXD    ( GPIO_PORTD | GPIO_IN | GPIO_AOUT | 9 )
+#define PD9_AOUT_SPI2_RXD    ( GPIO_GIUS | GPIO_PORTD | GPIO_IN | 9 )
 #define PD10_PF_SPL_SPR      ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 10 )
 #define PD10_AF_UART2_DSR    ( GPIO_PORTD | GPIO_OUT | GPIO_AF | 10 )
-#define PD10_AIN_SPI2_TXD    ( GPIO_PORTD | GPIO_OUT | GPIO_AIN | 10 )
+#define PD10_AIN_SPI2_TXD    ( GPIO_GIUS | GPIO_PORTD | GPIO_OUT | 10 )
 #define PD11_PF_CONTRAST     ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 11 )
 #define PD12_PF_ACD_OE       ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 12 )
 #define PD13_PF_LP_HSYNC     ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 13 )
@@ -225,7 +247,7 @@
 #define PD29_PF_LD14         ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 29 )
 #define PD30_PF_LD15         ( GPIO_PORTD | GPIO_OUT | GPIO_PF | 30 )
 #define PD31_PF_TMR2OUT      ( GPIO_PORTD | GPIO_PF | 31 )
-#define PD31_BIN_SPI2_TXD    ( GPIO_PORTD | GPIO_BIN | 31 )
+#define PD31_BIN_SPI2_TXD    ( GPIO_GIUS | GPIO_PORTD | GPIO_BIN | 31 )
 
 /*
  * PWM controller
diff --git a/include/asm-arm/arch-ixp4xx/platform.h b/include/asm-arm/arch-ixp4xx/platform.h
index d13ee7f..f14ed63 100644
--- a/include/asm-arm/arch-ixp4xx/platform.h
+++ b/include/asm-arm/arch-ixp4xx/platform.h
@@ -93,7 +93,7 @@
 
 static inline void gpio_line_config(u8 line, u32 direction)
 {
-	if (direction == IXP4XX_GPIO_OUT)
+	if (direction == IXP4XX_GPIO_IN)
 		*IXP4XX_GPIO_GPOER |= (1 << line);
 	else
 		*IXP4XX_GPIO_GPOER &= ~(1 << line);
diff --git a/include/asm-sparc/btfixup.h b/include/asm-sparc/btfixup.h
index b84c96c..c2868d0 100644
--- a/include/asm-sparc/btfixup.h
+++ b/include/asm-sparc/btfixup.h
@@ -49,17 +49,17 @@
 /* Put bottom 13bits into some register variable */
 
 #define BTFIXUPDEF_SIMM13(__name)							\
-	extern unsigned int ___sf_##__name(void) __attribute_const__;		\
+	static inline unsigned int ___sf_##__name(void) __attribute_const__;		\
 	extern unsigned ___ss_##__name[2];						\
-	extern __inline__ unsigned int ___sf_##__name(void) {				\
+	static inline unsigned int ___sf_##__name(void) {				\
 		unsigned int ret;							\
 		__asm__ ("or %%g0, ___s_" #__name ", %0" : "=r"(ret));			\
 		return ret;								\
 	}
 #define BTFIXUPDEF_SIMM13_INIT(__name,__val)						\
-	extern unsigned int ___sf_##__name(void) __attribute_const__;		\
+	static inline unsigned int ___sf_##__name(void) __attribute_const__;		\
 	extern unsigned ___ss_##__name[2];						\
-	extern __inline__ unsigned int ___sf_##__name(void) {				\
+	static inline unsigned int ___sf_##__name(void) {				\
 		unsigned int ret;							\
 		__asm__ ("or %%g0, ___s_" #__name "__btset_" #__val ", %0" : "=r"(ret));\
 		return ret;								\
@@ -71,17 +71,17 @@
  */
 
 #define BTFIXUPDEF_HALF(__name)								\
-	extern unsigned int ___af_##__name(void) __attribute_const__;		\
+	static inline unsigned int ___af_##__name(void) __attribute_const__;		\
 	extern unsigned ___as_##__name[2];						\
-	extern __inline__ unsigned int ___af_##__name(void) {				\
+	static inline unsigned int ___af_##__name(void) {				\
 		unsigned int ret;							\
 		__asm__ ("or %%g0, ___a_" #__name ", %0" : "=r"(ret));			\
 		return ret;								\
 	}
 #define BTFIXUPDEF_HALF_INIT(__name,__val)						\
-	extern unsigned int ___af_##__name(void) __attribute_const__;		\
+	static inline unsigned int ___af_##__name(void) __attribute_const__;		\
 	extern unsigned ___as_##__name[2];						\
-	extern __inline__ unsigned int ___af_##__name(void) {				\
+	static inline unsigned int ___af_##__name(void) {				\
 		unsigned int ret;							\
 		__asm__ ("or %%g0, ___a_" #__name "__btset_" #__val ", %0" : "=r"(ret));\
 		return ret;								\
@@ -90,17 +90,17 @@
 /* Put upper 22 bits into some register variable */
 
 #define BTFIXUPDEF_SETHI(__name)							\
-	extern unsigned int ___hf_##__name(void) __attribute_const__;		\
+	static inline unsigned int ___hf_##__name(void) __attribute_const__;		\
 	extern unsigned ___hs_##__name[2];						\
-	extern __inline__ unsigned int ___hf_##__name(void) {				\
+	static inline unsigned int ___hf_##__name(void) {				\
 		unsigned int ret;							\
 		__asm__ ("sethi %%hi(___h_" #__name "), %0" : "=r"(ret));		\
 		return ret;								\
 	}
 #define BTFIXUPDEF_SETHI_INIT(__name,__val)						\
-	extern unsigned int ___hf_##__name(void) __attribute_const__;		\
+	static inline unsigned int ___hf_##__name(void) __attribute_const__;		\
 	extern unsigned ___hs_##__name[2];						\
-	extern __inline__ unsigned int ___hf_##__name(void) {				\
+	static inline unsigned int ___hf_##__name(void) {				\
 		unsigned int ret;							\
 		__asm__ ("sethi %%hi(___h_" #__name "__btset_" #__val "), %0" : 	\
 			 "=r"(ret));							\
diff --git a/include/asm-sparc/cache.h b/include/asm-sparc/cache.h
index e6316fd..a10522c 100644
--- a/include/asm-sparc/cache.h
+++ b/include/asm-sparc/cache.h
@@ -27,7 +27,7 @@
  */
 
 /* First, cache-tag access. */
-extern __inline__ unsigned int get_icache_tag(int setnum, int tagnum)
+static inline unsigned int get_icache_tag(int setnum, int tagnum)
 {
 	unsigned int vaddr, retval;
 
@@ -38,7 +38,7 @@
 	return retval;
 }
 
-extern __inline__ void put_icache_tag(int setnum, int tagnum, unsigned int entry)
+static inline void put_icache_tag(int setnum, int tagnum, unsigned int entry)
 {
 	unsigned int vaddr;
 
@@ -51,7 +51,7 @@
 /* Second cache-data access.  The data is returned two-32bit quantities
  * at a time.
  */
-extern __inline__ void get_icache_data(int setnum, int tagnum, int subblock,
+static inline void get_icache_data(int setnum, int tagnum, int subblock,
 				       unsigned int *data)
 {
 	unsigned int value1, value2, vaddr;
@@ -67,7 +67,7 @@
 	data[0] = value1; data[1] = value2;
 }
 
-extern __inline__ void put_icache_data(int setnum, int tagnum, int subblock,
+static inline void put_icache_data(int setnum, int tagnum, int subblock,
 				       unsigned int *data)
 {
 	unsigned int value1, value2, vaddr;
@@ -92,35 +92,35 @@
  */
 
 /* Flushes which clear out both the on-chip and external caches */
-extern __inline__ void flush_ei_page(unsigned int addr)
+static inline void flush_ei_page(unsigned int addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_PAGE) :
 			     "memory");
 }
 
-extern __inline__ void flush_ei_seg(unsigned int addr)
+static inline void flush_ei_seg(unsigned int addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_SEG) :
 			     "memory");
 }
 
-extern __inline__ void flush_ei_region(unsigned int addr)
+static inline void flush_ei_region(unsigned int addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_REGION) :
 			     "memory");
 }
 
-extern __inline__ void flush_ei_ctx(unsigned int addr)
+static inline void flush_ei_ctx(unsigned int addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_CTX) :
 			     "memory");
 }
 
-extern __inline__ void flush_ei_user(unsigned int addr)
+static inline void flush_ei_user(unsigned int addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_USER) :
diff --git a/include/asm-sparc/cypress.h b/include/asm-sparc/cypress.h
index fc92fc8..99599533 100644
--- a/include/asm-sparc/cypress.h
+++ b/include/asm-sparc/cypress.h
@@ -48,25 +48,25 @@
 #define CYPRESS_NFAULT    0x00000002
 #define CYPRESS_MENABLE   0x00000001
 
-extern __inline__ void cypress_flush_page(unsigned long page)
+static inline void cypress_flush_page(unsigned long page)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (page), "i" (ASI_M_FLUSH_PAGE));
 }
 
-extern __inline__ void cypress_flush_segment(unsigned long addr)
+static inline void cypress_flush_segment(unsigned long addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_SEG));
 }
 
-extern __inline__ void cypress_flush_region(unsigned long addr)
+static inline void cypress_flush_region(unsigned long addr)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t" : :
 			     "r" (addr), "i" (ASI_M_FLUSH_REGION));
 }
 
-extern __inline__ void cypress_flush_context(void)
+static inline void cypress_flush_context(void)
 {
 	__asm__ __volatile__("sta %%g0, [%%g0] %0\n\t" : :
 			     "i" (ASI_M_FLUSH_CTX));
diff --git a/include/asm-sparc/delay.h b/include/asm-sparc/delay.h
index 6edf2cb..7ec8e9f 100644
--- a/include/asm-sparc/delay.h
+++ b/include/asm-sparc/delay.h
@@ -10,7 +10,7 @@
 #include <linux/config.h>
 #include <asm/cpudata.h>
 
-extern __inline__ void __delay(unsigned long loops)
+static inline void __delay(unsigned long loops)
 {
 	__asm__ __volatile__("cmp %0, 0\n\t"
 			     "1: bne 1b\n\t"
diff --git a/include/asm-sparc/dma.h b/include/asm-sparc/dma.h
index 07e6368..8ec206a 100644
--- a/include/asm-sparc/dma.h
+++ b/include/asm-sparc/dma.h
@@ -198,7 +198,7 @@
 /* Pause until counter runs out or BIT isn't set in the DMA condition
  * register.
  */
-extern __inline__ void sparc_dma_pause(struct sparc_dma_registers *regs,
+static inline void sparc_dma_pause(struct sparc_dma_registers *regs,
 				       unsigned long bit)
 {
 	int ctr = 50000;   /* Let's find some bugs ;) */
diff --git a/include/asm-sparc/iommu.h b/include/asm-sparc/iommu.h
index 8171362..70c589c 100644
--- a/include/asm-sparc/iommu.h
+++ b/include/asm-sparc/iommu.h
@@ -108,12 +108,12 @@
 	struct bit_map usemap;
 };
 
-extern __inline__ void iommu_invalidate(struct iommu_regs *regs)
+static inline void iommu_invalidate(struct iommu_regs *regs)
 {
 	regs->tlbflush = 0;
 }
 
-extern __inline__ void iommu_invalidate_page(struct iommu_regs *regs, unsigned long ba)
+static inline void iommu_invalidate_page(struct iommu_regs *regs, unsigned long ba)
 {
 	regs->pageflush = (ba & PAGE_MASK);
 }
diff --git a/include/asm-sparc/kdebug.h b/include/asm-sparc/kdebug.h
index 3ea4916..fba9248 100644
--- a/include/asm-sparc/kdebug.h
+++ b/include/asm-sparc/kdebug.h
@@ -46,7 +46,7 @@
 extern struct kernel_debug *linux_dbvec;
 
 /* Use this macro in C-code to enter the debugger. */
-extern __inline__ void sp_enter_debugger(void)
+static inline void sp_enter_debugger(void)
 {
 	__asm__ __volatile__("jmpl %0, %%o7\n\t"
 			     "nop\n\t" : :
diff --git a/include/asm-sparc/mbus.h b/include/asm-sparc/mbus.h
index 5f27490..ecacdf4 100644
--- a/include/asm-sparc/mbus.h
+++ b/include/asm-sparc/mbus.h
@@ -83,7 +83,7 @@
  */
 #define TBR_ID_SHIFT            20
 
-extern __inline__ int get_cpuid(void)
+static inline int get_cpuid(void)
 {
 	register int retval;
 	__asm__ __volatile__("rd %%tbr, %0\n\t"
@@ -93,7 +93,7 @@
 	return (retval & 3);
 }
 
-extern __inline__ int get_modid(void)
+static inline int get_modid(void)
 {
 	return (get_cpuid() | 0x8);
 }
diff --git a/include/asm-sparc/msi.h b/include/asm-sparc/msi.h
index b69543d..ff72cbd 100644
--- a/include/asm-sparc/msi.h
+++ b/include/asm-sparc/msi.h
@@ -19,7 +19,7 @@
 #define MSI_ASYNC_MODE  0x80000000	/* Operate the MSI asynchronously */
 
 
-extern __inline__ void msi_set_sync(void)
+static inline void msi_set_sync(void)
 {
 	__asm__ __volatile__ ("lda [%0] %1, %%g3\n\t"
 			      "andn %%g3, %2, %%g3\n\t"
diff --git a/include/asm-sparc/mxcc.h b/include/asm-sparc/mxcc.h
index 60ef9d6..128fe97 100644
--- a/include/asm-sparc/mxcc.h
+++ b/include/asm-sparc/mxcc.h
@@ -85,7 +85,7 @@
 
 #ifndef __ASSEMBLY__
 
-extern __inline__ void mxcc_set_stream_src(unsigned long *paddr)
+static inline void mxcc_set_stream_src(unsigned long *paddr)
 {
 	unsigned long data0 = paddr[0];
 	unsigned long data1 = paddr[1];
@@ -98,7 +98,7 @@
 			      "i" (ASI_M_MXCC) : "g2", "g3");
 }
 
-extern __inline__ void mxcc_set_stream_dst(unsigned long *paddr)
+static inline void mxcc_set_stream_dst(unsigned long *paddr)
 {
 	unsigned long data0 = paddr[0];
 	unsigned long data1 = paddr[1];
@@ -111,7 +111,7 @@
 			      "i" (ASI_M_MXCC) : "g2", "g3");
 }
 
-extern __inline__ unsigned long mxcc_get_creg(void)
+static inline unsigned long mxcc_get_creg(void)
 {
 	unsigned long mxcc_control;
 
@@ -125,7 +125,7 @@
 	return mxcc_control;
 }
 
-extern __inline__ void mxcc_set_creg(unsigned long mxcc_control)
+static inline void mxcc_set_creg(unsigned long mxcc_control)
 {
 	__asm__ __volatile__("sta %0, [%1] %2\n\t" : :
 			     "r" (mxcc_control), "r" (MXCC_CREG),
diff --git a/include/asm-sparc/obio.h b/include/asm-sparc/obio.h
index 62e1d77..47854a2 100644
--- a/include/asm-sparc/obio.h
+++ b/include/asm-sparc/obio.h
@@ -98,7 +98,7 @@
 
 #ifndef __ASSEMBLY__
 
-extern __inline__ int bw_get_intr_mask(int sbus_level)
+static inline int bw_get_intr_mask(int sbus_level)
 {
 	int mask;
 	
@@ -109,7 +109,7 @@
 	return mask;
 }
 
-extern __inline__ void bw_clear_intr_mask(int sbus_level, int mask)
+static inline void bw_clear_intr_mask(int sbus_level, int mask)
 {
 	__asm__ __volatile__ ("stha %0, [%1] %2" : :
 			      "r" (mask),
@@ -117,7 +117,7 @@
 			      "i" (ASI_M_CTL));
 }
 
-extern __inline__ unsigned bw_get_prof_limit(int cpu)
+static inline unsigned bw_get_prof_limit(int cpu)
 {
 	unsigned limit;
 	
@@ -128,7 +128,7 @@
 	return limit;
 }
 
-extern __inline__ void bw_set_prof_limit(int cpu, unsigned limit)
+static inline void bw_set_prof_limit(int cpu, unsigned limit)
 {
 	__asm__ __volatile__ ("sta %0, [%1] %2" : :
 			      "r" (limit),
@@ -136,7 +136,7 @@
 			      "i" (ASI_M_CTL));
 }
 
-extern __inline__ unsigned bw_get_ctrl(int cpu)
+static inline unsigned bw_get_ctrl(int cpu)
 {
 	unsigned ctrl;
 	
@@ -147,7 +147,7 @@
 	return ctrl;
 }
 
-extern __inline__ void bw_set_ctrl(int cpu, unsigned ctrl)
+static inline void bw_set_ctrl(int cpu, unsigned ctrl)
 {
 	__asm__ __volatile__ ("sta %0, [%1] %2" : :
 			      "r" (ctrl),
@@ -157,7 +157,7 @@
 
 extern unsigned char cpu_leds[32];
 
-extern __inline__ void show_leds(int cpuid)
+static inline void show_leds(int cpuid)
 {
 	cpuid &= 0x1e;
 	__asm__ __volatile__ ("stba %0, [%1] %2" : :
@@ -166,7 +166,7 @@
 			      "i" (ASI_M_CTL));
 }
 
-extern __inline__ unsigned cc_get_ipen(void)
+static inline unsigned cc_get_ipen(void)
 {
 	unsigned pending;
 	
@@ -177,7 +177,7 @@
 	return pending;
 }
 
-extern __inline__ void cc_set_iclr(unsigned clear)
+static inline void cc_set_iclr(unsigned clear)
 {
 	__asm__ __volatile__ ("stha %0, [%1] %2" : :
 			      "r" (clear),
@@ -185,7 +185,7 @@
 			      "i" (ASI_M_MXCC));
 }
 
-extern __inline__ unsigned cc_get_imsk(void)
+static inline unsigned cc_get_imsk(void)
 {
 	unsigned mask;
 	
@@ -196,7 +196,7 @@
 	return mask;
 }
 
-extern __inline__ void cc_set_imsk(unsigned mask)
+static inline void cc_set_imsk(unsigned mask)
 {
 	__asm__ __volatile__ ("stha %0, [%1] %2" : :
 			      "r" (mask),
@@ -204,7 +204,7 @@
 			      "i" (ASI_M_MXCC));
 }
 
-extern __inline__ unsigned cc_get_imsk_other(int cpuid)
+static inline unsigned cc_get_imsk_other(int cpuid)
 {
 	unsigned mask;
 	
@@ -215,7 +215,7 @@
 	return mask;
 }
 
-extern __inline__ void cc_set_imsk_other(int cpuid, unsigned mask)
+static inline void cc_set_imsk_other(int cpuid, unsigned mask)
 {
 	__asm__ __volatile__ ("stha %0, [%1] %2" : :
 			      "r" (mask),
@@ -223,7 +223,7 @@
 			      "i" (ASI_M_CTL));
 }
 
-extern __inline__ void cc_set_igen(unsigned gen)
+static inline void cc_set_igen(unsigned gen)
 {
 	__asm__ __volatile__ ("sta %0, [%1] %2" : :
 			      "r" (gen),
@@ -239,7 +239,7 @@
 #define IGEN_MESSAGE(bcast, devid, sid, levels) \
 	(((bcast) << 31) | ((devid) << 23) | ((sid) << 15) | (levels))
             
-extern __inline__ void sun4d_send_ipi(int cpu, int level)
+static inline void sun4d_send_ipi(int cpu, int level)
 {
 	cc_set_igen(IGEN_MESSAGE(0, cpu << 3, 6 + ((level >> 1) & 7), 1 << (level - 1)));
 }
diff --git a/include/asm-sparc/pci.h b/include/asm-sparc/pci.h
index 97052ba..3864474 100644
--- a/include/asm-sparc/pci.h
+++ b/include/asm-sparc/pci.h
@@ -15,12 +15,12 @@
 
 #define PCI_IRQ_NONE		0xffffffff
 
-extern inline void pcibios_set_master(struct pci_dev *dev)
+static inline void pcibios_set_master(struct pci_dev *dev)
 {
 	/* No special bus mastering setup handling */
 }
 
-extern inline void pcibios_penalize_isa_irq(int irq, int active)
+static inline void pcibios_penalize_isa_irq(int irq, int active)
 {
 	/* We don't do dynamic PCI IRQ allocation */
 }
@@ -137,7 +137,7 @@
  * only drive the low 24-bits during PCI bus mastering, then
  * you would pass 0x00ffffff as the mask to this function.
  */
-extern inline int pci_dma_supported(struct pci_dev *hwdev, u64 mask)
+static inline int pci_dma_supported(struct pci_dev *hwdev, u64 mask)
 {
 	return 1;
 }
diff --git a/include/asm-sparc/pgtable.h b/include/asm-sparc/pgtable.h
index 8395ad2..a14e986 100644
--- a/include/asm-sparc/pgtable.h
+++ b/include/asm-sparc/pgtable.h
@@ -154,7 +154,7 @@
 BTFIXUPDEF_CALL(void, pte_clear, pte_t *)
 BTFIXUPDEF_CALL(int, pte_read, pte_t)
 
-extern __inline__ int pte_none(pte_t pte)
+static inline int pte_none(pte_t pte)
 {
 	return !(pte_val(pte) & ~BTFIXUP_SETHI(none_mask));
 }
@@ -167,7 +167,7 @@
 BTFIXUPDEF_CALL_CONST(int, pmd_present, pmd_t)
 BTFIXUPDEF_CALL(void, pmd_clear, pmd_t *)
 
-extern __inline__ int pmd_none(pmd_t pmd)
+static inline int pmd_none(pmd_t pmd)
 {
 	return !(pmd_val(pmd) & ~BTFIXUP_SETHI(none_mask));
 }
@@ -194,20 +194,20 @@
 BTFIXUPDEF_HALF(pte_dirtyi)
 BTFIXUPDEF_HALF(pte_youngi)
 
-extern int pte_write(pte_t pte) __attribute_const__;
-extern __inline__ int pte_write(pte_t pte)
+static int pte_write(pte_t pte) __attribute_const__;
+static inline int pte_write(pte_t pte)
 {
 	return pte_val(pte) & BTFIXUP_HALF(pte_writei);
 }
 
-extern int pte_dirty(pte_t pte) __attribute_const__;
-extern __inline__ int pte_dirty(pte_t pte)
+static int pte_dirty(pte_t pte) __attribute_const__;
+static inline int pte_dirty(pte_t pte)
 {
 	return pte_val(pte) & BTFIXUP_HALF(pte_dirtyi);
 }
 
-extern int pte_young(pte_t pte) __attribute_const__;
-extern __inline__ int pte_young(pte_t pte)
+static int pte_young(pte_t pte) __attribute_const__;
+static inline int pte_young(pte_t pte)
 {
 	return pte_val(pte) & BTFIXUP_HALF(pte_youngi);
 }
@@ -217,8 +217,8 @@
  */
 BTFIXUPDEF_HALF(pte_filei)
 
-extern int pte_file(pte_t pte) __attribute_const__;
-extern __inline__ int pte_file(pte_t pte)
+static int pte_file(pte_t pte) __attribute_const__;
+static inline int pte_file(pte_t pte)
 {
 	return pte_val(pte) & BTFIXUP_HALF(pte_filei);
 }
@@ -229,20 +229,20 @@
 BTFIXUPDEF_HALF(pte_mkcleani)
 BTFIXUPDEF_HALF(pte_mkoldi)
 
-extern pte_t pte_wrprotect(pte_t pte) __attribute_const__;
-extern __inline__ pte_t pte_wrprotect(pte_t pte)
+static pte_t pte_wrprotect(pte_t pte) __attribute_const__;
+static inline pte_t pte_wrprotect(pte_t pte)
 {
 	return __pte(pte_val(pte) & ~BTFIXUP_HALF(pte_wrprotecti));
 }
 
-extern pte_t pte_mkclean(pte_t pte) __attribute_const__;
-extern __inline__ pte_t pte_mkclean(pte_t pte)
+static pte_t pte_mkclean(pte_t pte) __attribute_const__;
+static inline pte_t pte_mkclean(pte_t pte)
 {
 	return __pte(pte_val(pte) & ~BTFIXUP_HALF(pte_mkcleani));
 }
 
-extern pte_t pte_mkold(pte_t pte) __attribute_const__;
-extern __inline__ pte_t pte_mkold(pte_t pte)
+static pte_t pte_mkold(pte_t pte) __attribute_const__;
+static inline pte_t pte_mkold(pte_t pte)
 {
 	return __pte(pte_val(pte) & ~BTFIXUP_HALF(pte_mkoldi));
 }
@@ -278,8 +278,8 @@
 
 BTFIXUPDEF_INT(pte_modify_mask)
 
-extern pte_t pte_modify(pte_t pte, pgprot_t newprot) __attribute_const__;
-extern __inline__ pte_t pte_modify(pte_t pte, pgprot_t newprot)
+static pte_t pte_modify(pte_t pte, pgprot_t newprot) __attribute_const__;
+static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
 {
 	return __pte((pte_val(pte) & BTFIXUP_INT(pte_modify_mask)) |
 		pgprot_val(newprot));
@@ -386,13 +386,13 @@
 
 #define NO_CONTEXT     -1
 
-extern __inline__ void remove_from_ctx_list(struct ctx_list *entry)
+static inline void remove_from_ctx_list(struct ctx_list *entry)
 {
 	entry->next->prev = entry->prev;
 	entry->prev->next = entry->next;
 }
 
-extern __inline__ void add_to_ctx_list(struct ctx_list *head, struct ctx_list *entry)
+static inline void add_to_ctx_list(struct ctx_list *head, struct ctx_list *entry)
 {
 	entry->next = head;
 	(entry->prev = head->prev)->next = entry;
@@ -401,7 +401,7 @@
 #define add_to_free_ctxlist(entry) add_to_ctx_list(&ctx_free, entry)
 #define add_to_used_ctxlist(entry) add_to_ctx_list(&ctx_used, entry)
 
-extern __inline__ unsigned long
+static inline unsigned long
 __get_phys (unsigned long addr)
 {
 	switch (sparc_cpu_model){
@@ -416,7 +416,7 @@
 	}
 }
 
-extern __inline__ int
+static inline int
 __get_iospace (unsigned long addr)
 {
 	switch (sparc_cpu_model){
diff --git a/include/asm-sparc/pgtsrmmu.h b/include/asm-sparc/pgtsrmmu.h
index ee3b9d9..edeb981 100644
--- a/include/asm-sparc/pgtsrmmu.h
+++ b/include/asm-sparc/pgtsrmmu.h
@@ -148,7 +148,7 @@
 #define __nocache_fix(VADDR) __va(__nocache_pa(VADDR))
 
 /* Accessing the MMU control register. */
-extern __inline__ unsigned int srmmu_get_mmureg(void)
+static inline unsigned int srmmu_get_mmureg(void)
 {
         unsigned int retval;
 	__asm__ __volatile__("lda [%%g0] %1, %0\n\t" :
@@ -157,14 +157,14 @@
 	return retval;
 }
 
-extern __inline__ void srmmu_set_mmureg(unsigned long regval)
+static inline void srmmu_set_mmureg(unsigned long regval)
 {
 	__asm__ __volatile__("sta %0, [%%g0] %1\n\t" : :
 			     "r" (regval), "i" (ASI_M_MMUREGS) : "memory");
 
 }
 
-extern __inline__ void srmmu_set_ctable_ptr(unsigned long paddr)
+static inline void srmmu_set_ctable_ptr(unsigned long paddr)
 {
 	paddr = ((paddr >> 4) & SRMMU_CTX_PMASK);
 	__asm__ __volatile__("sta %0, [%1] %2\n\t" : :
@@ -173,7 +173,7 @@
 			     "memory");
 }
 
-extern __inline__ unsigned long srmmu_get_ctable_ptr(void)
+static inline unsigned long srmmu_get_ctable_ptr(void)
 {
 	unsigned int retval;
 
@@ -184,14 +184,14 @@
 	return (retval & SRMMU_CTX_PMASK) << 4;
 }
 
-extern __inline__ void srmmu_set_context(int context)
+static inline void srmmu_set_context(int context)
 {
 	__asm__ __volatile__("sta %0, [%1] %2\n\t" : :
 			     "r" (context), "r" (SRMMU_CTX_REG),
 			     "i" (ASI_M_MMUREGS) : "memory");
 }
 
-extern __inline__ int srmmu_get_context(void)
+static inline int srmmu_get_context(void)
 {
 	register int retval;
 	__asm__ __volatile__("lda [%1] %2, %0\n\t" :
@@ -201,7 +201,7 @@
 	return retval;
 }
 
-extern __inline__ unsigned int srmmu_get_fstatus(void)
+static inline unsigned int srmmu_get_fstatus(void)
 {
 	unsigned int retval;
 
@@ -211,7 +211,7 @@
 	return retval;
 }
 
-extern __inline__ unsigned int srmmu_get_faddr(void)
+static inline unsigned int srmmu_get_faddr(void)
 {
 	unsigned int retval;
 
@@ -222,7 +222,7 @@
 }
 
 /* This is guaranteed on all SRMMU's. */
-extern __inline__ void srmmu_flush_whole_tlb(void)
+static inline void srmmu_flush_whole_tlb(void)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t": :
 			     "r" (0x400),        /* Flush entire TLB!! */
@@ -231,7 +231,7 @@
 }
 
 /* These flush types are not available on all chips... */
-extern __inline__ void srmmu_flush_tlb_ctx(void)
+static inline void srmmu_flush_tlb_ctx(void)
 {
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t": :
 			     "r" (0x300),        /* Flush TLB ctx.. */
@@ -239,7 +239,7 @@
 
 }
 
-extern __inline__ void srmmu_flush_tlb_region(unsigned long addr)
+static inline void srmmu_flush_tlb_region(unsigned long addr)
 {
 	addr &= SRMMU_PGDIR_MASK;
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t": :
@@ -249,7 +249,7 @@
 }
 
 
-extern __inline__ void srmmu_flush_tlb_segment(unsigned long addr)
+static inline void srmmu_flush_tlb_segment(unsigned long addr)
 {
 	addr &= SRMMU_REAL_PMD_MASK;
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t": :
@@ -258,7 +258,7 @@
 
 }
 
-extern __inline__ void srmmu_flush_tlb_page(unsigned long page)
+static inline void srmmu_flush_tlb_page(unsigned long page)
 {
 	page &= PAGE_MASK;
 	__asm__ __volatile__("sta %%g0, [%0] %1\n\t": :
@@ -267,7 +267,7 @@
 
 }
 
-extern __inline__ unsigned long srmmu_hwprobe(unsigned long vaddr)
+static inline unsigned long srmmu_hwprobe(unsigned long vaddr)
 {
 	unsigned long retval;
 
@@ -279,7 +279,7 @@
 	return retval;
 }
 
-extern __inline__ int
+static inline int
 srmmu_get_pte (unsigned long addr)
 {
 	register unsigned long entry;
diff --git a/include/asm-sparc/processor.h b/include/asm-sparc/processor.h
index 5a7a1a8..6fbb3f0 100644
--- a/include/asm-sparc/processor.h
+++ b/include/asm-sparc/processor.h
@@ -79,7 +79,7 @@
 extern unsigned long thread_saved_pc(struct task_struct *t);
 
 /* Do necessary setup to start up a newly executed thread. */
-extern __inline__ void start_thread(struct pt_regs * regs, unsigned long pc,
+static inline void start_thread(struct pt_regs * regs, unsigned long pc,
 				    unsigned long sp)
 {
 	register unsigned long zero asm("g1");
diff --git a/include/asm-sparc/psr.h b/include/asm-sparc/psr.h
index 9778b8c..19c9780 100644
--- a/include/asm-sparc/psr.h
+++ b/include/asm-sparc/psr.h
@@ -38,7 +38,7 @@
 
 #ifndef __ASSEMBLY__
 /* Get the %psr register. */
-extern __inline__ unsigned int get_psr(void)
+static inline unsigned int get_psr(void)
 {
 	unsigned int psr;
 	__asm__ __volatile__(
@@ -53,7 +53,7 @@
 	return psr;
 }
 
-extern __inline__ void put_psr(unsigned int new_psr)
+static inline void put_psr(unsigned int new_psr)
 {
 	__asm__ __volatile__(
 		"wr	%0, 0x0, %%psr\n\t"
@@ -72,7 +72,7 @@
 
 extern unsigned int fsr_storage;
 
-extern __inline__ unsigned int get_fsr(void)
+static inline unsigned int get_fsr(void)
 {
 	unsigned int fsr = 0;
 
diff --git a/include/asm-sparc/sbi.h b/include/asm-sparc/sbi.h
index 739ccac..86a603a 100644
--- a/include/asm-sparc/sbi.h
+++ b/include/asm-sparc/sbi.h
@@ -65,7 +65,7 @@
 
 #ifndef __ASSEMBLY__
 
-extern __inline__ int acquire_sbi(int devid, int mask)
+static inline int acquire_sbi(int devid, int mask)
 {
 	__asm__ __volatile__ ("swapa [%2] %3, %0" :
 			      "=r" (mask) :
@@ -75,7 +75,7 @@
 	return mask;
 }
 
-extern __inline__ void release_sbi(int devid, int mask)
+static inline void release_sbi(int devid, int mask)
 {
 	__asm__ __volatile__ ("sta %0, [%1] %2" : :
 			      "r" (mask),
@@ -83,7 +83,7 @@
 			      "i" (ASI_M_CTL));
 }
 
-extern __inline__ void set_sbi_tid(int devid, int targetid)
+static inline void set_sbi_tid(int devid, int targetid)
 {
 	__asm__ __volatile__ ("sta %0, [%1] %2" : :
 			      "r" (targetid),
@@ -91,7 +91,7 @@
 			      "i" (ASI_M_CTL));
 }
 
-extern __inline__ int get_sbi_ctl(int devid, int cfgno)
+static inline int get_sbi_ctl(int devid, int cfgno)
 {
 	int cfg;
 	
@@ -102,7 +102,7 @@
 	return cfg;
 }
 
-extern __inline__ void set_sbi_ctl(int devid, int cfgno, int cfg)
+static inline void set_sbi_ctl(int devid, int cfgno, int cfg)
 {
 	__asm__ __volatile__ ("sta %0, [%1] %2" : :
 			      "r" (cfg),
diff --git a/include/asm-sparc/sbus.h b/include/asm-sparc/sbus.h
index 3a8b390..a13cddc 100644
--- a/include/asm-sparc/sbus.h
+++ b/include/asm-sparc/sbus.h
@@ -28,12 +28,12 @@
  * numbers + offsets, and vice versa.
  */
 
-extern __inline__ unsigned long sbus_devaddr(int slotnum, unsigned long offset)
+static inline unsigned long sbus_devaddr(int slotnum, unsigned long offset)
 {
   return (unsigned long) (SUN_SBUS_BVADDR+((slotnum)<<25)+(offset));
 }
 
-extern __inline__ int sbus_dev_slot(unsigned long dev_addr)
+static inline int sbus_dev_slot(unsigned long dev_addr)
 {
   return (int) (((dev_addr)-SUN_SBUS_BVADDR)>>25);
 }
@@ -80,7 +80,7 @@
 
 extern struct sbus_bus *sbus_root;
 
-extern __inline__ int
+static inline int
 sbus_is_slave(struct sbus_dev *dev)
 {
 	/* XXX Have to write this for sun4c's */
diff --git a/include/asm-sparc/smp.h b/include/asm-sparc/smp.h
index 4f96d83..580c51d 100644
--- a/include/asm-sparc/smp.h
+++ b/include/asm-sparc/smp.h
@@ -60,22 +60,22 @@
 #define smp_cross_call(func,arg1,arg2,arg3,arg4,arg5) BTFIXUP_CALL(smp_cross_call)(func,arg1,arg2,arg3,arg4,arg5)
 #define smp_message_pass(target,msg,data,wait) BTFIXUP_CALL(smp_message_pass)(target,msg,data,wait)
 
-extern __inline__ void xc0(smpfunc_t func) { smp_cross_call(func, 0, 0, 0, 0, 0); }
-extern __inline__ void xc1(smpfunc_t func, unsigned long arg1)
+static inline void xc0(smpfunc_t func) { smp_cross_call(func, 0, 0, 0, 0, 0); }
+static inline void xc1(smpfunc_t func, unsigned long arg1)
 { smp_cross_call(func, arg1, 0, 0, 0, 0); }
-extern __inline__ void xc2(smpfunc_t func, unsigned long arg1, unsigned long arg2)
+static inline void xc2(smpfunc_t func, unsigned long arg1, unsigned long arg2)
 { smp_cross_call(func, arg1, arg2, 0, 0, 0); }
-extern __inline__ void xc3(smpfunc_t func, unsigned long arg1, unsigned long arg2,
+static inline void xc3(smpfunc_t func, unsigned long arg1, unsigned long arg2,
 			   unsigned long arg3)
 { smp_cross_call(func, arg1, arg2, arg3, 0, 0); }
-extern __inline__ void xc4(smpfunc_t func, unsigned long arg1, unsigned long arg2,
+static inline void xc4(smpfunc_t func, unsigned long arg1, unsigned long arg2,
 			   unsigned long arg3, unsigned long arg4)
 { smp_cross_call(func, arg1, arg2, arg3, arg4, 0); }
-extern __inline__ void xc5(smpfunc_t func, unsigned long arg1, unsigned long arg2,
+static inline void xc5(smpfunc_t func, unsigned long arg1, unsigned long arg2,
 			   unsigned long arg3, unsigned long arg4, unsigned long arg5)
 { smp_cross_call(func, arg1, arg2, arg3, arg4, arg5); }
 
-extern __inline__ int smp_call_function(void (*func)(void *info), void *info, int nonatomic, int wait)
+static inline int smp_call_function(void (*func)(void *info), void *info, int nonatomic, int wait)
 {
 	xc1((smpfunc_t)func, (unsigned long)info);
 	return 0;
@@ -84,16 +84,16 @@
 extern __volatile__ int __cpu_number_map[NR_CPUS];
 extern __volatile__ int __cpu_logical_map[NR_CPUS];
 
-extern __inline__ int cpu_logical_map(int cpu)
+static inline int cpu_logical_map(int cpu)
 {
 	return __cpu_logical_map[cpu];
 }
-extern __inline__ int cpu_number_map(int cpu)
+static inline int cpu_number_map(int cpu)
 {
 	return __cpu_number_map[cpu];
 }
 
-extern __inline__ int hard_smp4m_processor_id(void)
+static inline int hard_smp4m_processor_id(void)
 {
 	int cpuid;
 
@@ -104,7 +104,7 @@
 	return cpuid;
 }
 
-extern __inline__ int hard_smp4d_processor_id(void)
+static inline int hard_smp4d_processor_id(void)
 {
 	int cpuid;
 
@@ -114,7 +114,7 @@
 }
 
 #ifndef MODULE
-extern __inline__ int hard_smp_processor_id(void)
+static inline int hard_smp_processor_id(void)
 {
 	int cpuid;
 
@@ -136,7 +136,7 @@
 	return cpuid;
 }
 #else
-extern __inline__ int hard_smp_processor_id(void)
+static inline int hard_smp_processor_id(void)
 {
 	int cpuid;
 	
diff --git a/include/asm-sparc/smpprim.h b/include/asm-sparc/smpprim.h
index 9b9c28e..e7b6d34 100644
--- a/include/asm-sparc/smpprim.h
+++ b/include/asm-sparc/smpprim.h
@@ -15,7 +15,7 @@
  * atomic.
  */
 
-extern __inline__ __volatile__ char test_and_set(void *addr)
+static inline __volatile__ char test_and_set(void *addr)
 {
 	char state = 0;
 
@@ -27,7 +27,7 @@
 }
 
 /* Initialize a spin-lock. */
-extern __inline__ __volatile__ smp_initlock(void *spinlock)
+static inline __volatile__ smp_initlock(void *spinlock)
 {
 	/* Unset the lock. */
 	*((unsigned char *) spinlock) = 0;
@@ -36,7 +36,7 @@
 }
 
 /* This routine spins until it acquires the lock at ADDR. */
-extern __inline__ __volatile__ smp_lock(void *addr)
+static inline __volatile__ smp_lock(void *addr)
 {
 	while(test_and_set(addr) == 0xff)
 		;
@@ -46,7 +46,7 @@
 }
 
 /* This routine releases the lock at ADDR. */
-extern __inline__ __volatile__ smp_unlock(void *addr)
+static inline __volatile__ smp_unlock(void *addr)
 {
 	*((unsigned char *) addr) = 0;
 }
diff --git a/include/asm-sparc/spinlock.h b/include/asm-sparc/spinlock.h
index 111727a..e344c98 100644
--- a/include/asm-sparc/spinlock.h
+++ b/include/asm-sparc/spinlock.h
@@ -17,7 +17,7 @@
 #define __raw_spin_unlock_wait(lock) \
 	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
 
-extern __inline__ void __raw_spin_lock(raw_spinlock_t *lock)
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__(
 	"\n1:\n\t"
@@ -37,7 +37,7 @@
 	: "g2", "memory", "cc");
 }
 
-extern __inline__ int __raw_spin_trylock(raw_spinlock_t *lock)
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
 {
 	unsigned int result;
 	__asm__ __volatile__("ldstub [%1], %0"
@@ -47,7 +47,7 @@
 	return (result == 0);
 }
 
-extern __inline__ void __raw_spin_unlock(raw_spinlock_t *lock)
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
 {
 	__asm__ __volatile__("stb %%g0, [%0]" : : "r" (lock) : "memory");
 }
@@ -78,7 +78,7 @@
  *
  * Unfortunately this scheme limits us to ~16,000,000 cpus.
  */
-extern __inline__ void __read_lock(raw_rwlock_t *rw)
+static inline void __read_lock(raw_rwlock_t *rw)
 {
 	register raw_rwlock_t *lp asm("g1");
 	lp = rw;
@@ -98,7 +98,7 @@
 	local_irq_restore(flags); \
 } while(0)
 
-extern __inline__ void __read_unlock(raw_rwlock_t *rw)
+static inline void __read_unlock(raw_rwlock_t *rw)
 {
 	register raw_rwlock_t *lp asm("g1");
 	lp = rw;
diff --git a/include/asm-sparc/system.h b/include/asm-sparc/system.h
index 3557781..1f6b71f 100644
--- a/include/asm-sparc/system.h
+++ b/include/asm-sparc/system.h
@@ -204,7 +204,7 @@
 BTFIXUPDEF_CALL(void, ___xchg32, void)
 #endif
 
-extern __inline__ unsigned long xchg_u32(__volatile__ unsigned long *m, unsigned long val)
+static inline unsigned long xchg_u32(__volatile__ unsigned long *m, unsigned long val)
 {
 #ifdef CONFIG_SMP
 	__asm__ __volatile__("swap [%2], %0"
diff --git a/include/asm-sparc/traps.h b/include/asm-sparc/traps.h
index 6690ab9..f62c7f8 100644
--- a/include/asm-sparc/traps.h
+++ b/include/asm-sparc/traps.h
@@ -22,7 +22,7 @@
 /* We set this to _start in system setup. */
 extern struct tt_entry *sparc_ttable;
 
-extern __inline__ unsigned long get_tbr(void)
+static inline unsigned long get_tbr(void)
 {
 	unsigned long tbr;
 
diff --git a/include/asm-um/processor-generic.h b/include/asm-um/processor-generic.h
index 2d24236..075771c 100644
--- a/include/asm-um/processor-generic.h
+++ b/include/asm-um/processor-generic.h
@@ -13,6 +13,7 @@
 #include "linux/config.h"
 #include "asm/ptrace.h"
 #include "choose-mode.h"
+#include "registers.h"
 
 struct mm_struct;
 
@@ -136,19 +137,15 @@
 #define current_cpu_data boot_cpu_data
 #endif
 
-#define KSTK_EIP(tsk) (PT_REGS_IP(&tsk->thread.regs))
-#define KSTK_ESP(tsk) (PT_REGS_SP(&tsk->thread.regs))
+
+#ifdef CONFIG_MODE_SKAS
+#define KSTK_REG(tsk, reg) \
+	({ union uml_pt_regs regs; \
+	   get_thread_regs(&regs, tsk->thread.mode.skas.switch_buf); \
+	   UPT_REG(&regs, reg); })
+#else
+#define KSTK_REG(tsk, reg) (0xbadbabe)
+#endif
 #define get_wchan(p) (0)
 
 #endif
-
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/include/asm-um/processor-i386.h b/include/asm-um/processor-i386.h
index 431bad3..4108a57 100644
--- a/include/asm-um/processor-i386.h
+++ b/include/asm-um/processor-i386.h
@@ -43,17 +43,10 @@
 #define ARCH_IS_STACKGROW(address) \
        (address + 32 >= UPT_SP(&current->thread.regs.regs))
 
+#define KSTK_EIP(tsk) KSTK_REG(tsk, EIP)
+#define KSTK_ESP(tsk) KSTK_REG(tsk, UESP)
+#define KSTK_EBP(tsk) KSTK_REG(tsk, EBP)
+
 #include "asm/processor-generic.h"
 
 #endif
-
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/include/asm-um/processor-x86_64.h b/include/asm-um/processor-x86_64.h
index 0beb9a4..e1e1255 100644
--- a/include/asm-um/processor-x86_64.h
+++ b/include/asm-um/processor-x86_64.h
@@ -36,17 +36,9 @@
 #define ARCH_IS_STACKGROW(address) \
         (address + 128 >= UPT_SP(&current->thread.regs.regs))
 
+#define KSTK_EIP(tsk) KSTK_REG(tsk, RIP)
+#define KSTK_ESP(tsk) KSTK_REG(tsk, RSP)
+
 #include "asm/processor-generic.h"
 
 #endif
-
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/include/linux/ata.h b/include/linux/ata.h
index a5b74ef..ecb7346 100644
--- a/include/linux/ata.h
+++ b/include/linux/ata.h
@@ -132,6 +132,7 @@
 	ATA_CMD_PACKET		= 0xA0,
 	ATA_CMD_VERIFY		= 0x40,
 	ATA_CMD_VERIFY_EXT	= 0x42,
+	ATA_CMD_INIT_DEV_PARAMS	= 0x91,
 
 	/* SETFEATURES stuff */
 	SETFEATURES_XFER	= 0x03,
@@ -146,14 +147,14 @@
 	XFER_MW_DMA_2		= 0x22,
 	XFER_MW_DMA_1		= 0x21,
 	XFER_MW_DMA_0		= 0x20,
+	XFER_SW_DMA_2		= 0x12,
+	XFER_SW_DMA_1		= 0x11,
+	XFER_SW_DMA_0		= 0x10,
 	XFER_PIO_4		= 0x0C,
 	XFER_PIO_3		= 0x0B,
 	XFER_PIO_2		= 0x0A,
 	XFER_PIO_1		= 0x09,
 	XFER_PIO_0		= 0x08,
-	XFER_SW_DMA_2		= 0x12,
-	XFER_SW_DMA_1		= 0x11,
-	XFER_SW_DMA_0		= 0x10,
 	XFER_PIO_SLOW		= 0x00,
 
 	/* ATAPI stuff */
@@ -181,6 +182,7 @@
 	ATA_TFLAG_ISADDR	= (1 << 1), /* enable r/w to nsect/lba regs */
 	ATA_TFLAG_DEVICE	= (1 << 2), /* enable r/w to device reg */
 	ATA_TFLAG_WRITE		= (1 << 3), /* data dir: host->dev==1 (write) */
+	ATA_TFLAG_LBA		= (1 << 4), /* enable LBA */
 };
 
 enum ata_tf_protocols {
@@ -250,6 +252,18 @@
 	  ((u64) (id)[(n) + 1] << 16) |	\
 	  ((u64) (id)[(n) + 0]) )
 
+static inline int ata_id_current_chs_valid(u16 *id)
+{
+	/* For ATA-1 devices, if the INITIALIZE DEVICE PARAMETERS command 
+	   has not been issued to the device then the values of 
+	   id[54] to id[56] are vendor specific. */
+	return (id[53] & 0x01) && /* Current translation valid */
+		id[54] &&  /* cylinders in current translation */
+		id[55] &&  /* heads in current translation */
+		id[55] <= 16 &&
+		id[56];    /* sectors in current translation */
+}
+
 static inline int atapi_cdb_len(u16 *dev_id)
 {
 	u16 tmp = dev_id[0] & 0x3;
diff --git a/include/linux/atmdev.h b/include/linux/atmdev.h
index 9f374cf..f1fd849 100644
--- a/include/linux/atmdev.h
+++ b/include/linux/atmdev.h
@@ -457,7 +457,7 @@
 
 int atm_charge(struct atm_vcc *vcc,int truesize);
 struct sk_buff *atm_alloc_charge(struct atm_vcc *vcc,int pdu_size,
-    int gfp_flags);
+    unsigned int __nocast gfp_flags);
 int atm_pcr_goal(struct atm_trafprm *tp);
 
 void vcc_release_async(struct atm_vcc *vcc, int reply);
diff --git a/include/linux/bfs_fs.h b/include/linux/bfs_fs.h
index c1237aa..8ed6dfdc 100644
--- a/include/linux/bfs_fs.h
+++ b/include/linux/bfs_fs.h
@@ -20,19 +20,19 @@
 
 /* BFS inode layout on disk */
 struct bfs_inode {
-	__u16 i_ino;
+	__le16 i_ino;
 	__u16 i_unused;
-	__u32 i_sblock;
-	__u32 i_eblock;
-	__u32 i_eoffset;
-	__u32 i_vtype;
-	__u32 i_mode;
-	__s32 i_uid;
-	__s32 i_gid;
-	__u32 i_nlink;
-	__u32 i_atime;
-	__u32 i_mtime;
-	__u32 i_ctime;
+	__le32 i_sblock;
+	__le32 i_eblock;
+	__le32 i_eoffset;
+	__le32 i_vtype;
+	__le32 i_mode;
+	__le32 i_uid;
+	__le32 i_gid;
+	__le32 i_nlink;
+	__le32 i_atime;
+	__le32 i_mtime;
+	__le32 i_ctime;
 	__u32 i_padding[4];
 };
 
@@ -41,17 +41,17 @@
 #define BFS_DIRS_PER_BLOCK	32
 
 struct bfs_dirent {
-	__u16 ino;
+	__le16 ino;
 	char name[BFS_NAMELEN];
 };
 
 /* BFS superblock layout on disk */
 struct bfs_super_block {
-	__u32 s_magic;
-	__u32 s_start;
-	__u32 s_end;
-	__s32 s_from;
-	__s32 s_to;
+	__le32 s_magic;
+	__le32 s_start;
+	__le32 s_end;
+	__le32 s_from;
+	__le32 s_to;
 	__s32 s_bfrom;
 	__s32 s_bto;
 	char  s_fsname[6];
@@ -66,15 +66,15 @@
 #define BFS_INO2OFF(ino) \
 	((__u32)(((ino) - BFS_ROOT_INO) * sizeof(struct bfs_inode)) + BFS_BSIZE)
 #define BFS_NZFILESIZE(ip) \
-        ((cpu_to_le32((ip)->i_eoffset) + 1) -  cpu_to_le32((ip)->i_sblock) * BFS_BSIZE)
+        ((le32_to_cpu((ip)->i_eoffset) + 1) -  le32_to_cpu((ip)->i_sblock) * BFS_BSIZE)
 
 #define BFS_FILESIZE(ip) \
         ((ip)->i_sblock == 0 ? 0 : BFS_NZFILESIZE(ip))
 
 #define BFS_FILEBLOCKS(ip) \
-        ((ip)->i_sblock == 0 ? 0 : (cpu_to_le32((ip)->i_eblock) + 1) -  cpu_to_le32((ip)->i_sblock))
+        ((ip)->i_sblock == 0 ? 0 : (le32_to_cpu((ip)->i_eblock) + 1) -  le32_to_cpu((ip)->i_sblock))
 #define BFS_UNCLEAN(bfs_sb, sb)	\
-	((cpu_to_le32(bfs_sb->s_from) != -1) && (cpu_to_le32(bfs_sb->s_to) != -1) && !(sb->s_flags & MS_RDONLY))
+	((le32_to_cpu(bfs_sb->s_from) != -1) && (le32_to_cpu(bfs_sb->s_to) != -1) && !(sb->s_flags & MS_RDONLY))
 
 
 #endif	/* _LINUX_BFS_FS_H */
diff --git a/include/linux/connector.h b/include/linux/connector.h
index 86d4b0a..96582c9 100644
--- a/include/linux/connector.h
+++ b/include/linux/connector.h
@@ -149,7 +149,7 @@
 
 int cn_add_callback(struct cb_id *, char *, void (*callback) (void *));
 void cn_del_callback(struct cb_id *);
-int cn_netlink_send(struct cn_msg *, u32, int);
+int cn_netlink_send(struct cn_msg *, u32, unsigned int __nocast);
 
 int cn_queue_add_callback(struct cn_queue_dev *dev, char *name, struct cb_id *id, void (*callback)(void *));
 void cn_queue_del_callback(struct cn_queue_dev *dev, struct cb_id *id);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 3ab6762..0f89f41 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -97,6 +97,7 @@
 	ATA_DFLAG_LBA48		= (1 << 0), /* device supports LBA48 */
 	ATA_DFLAG_PIO		= (1 << 1), /* device currently in PIO mode */
 	ATA_DFLAG_LOCK_SECTORS	= (1 << 2), /* don't adjust max_sectors */
+	ATA_DFLAG_LBA		= (1 << 3), /* device supports LBA */
 
 	ATA_DEV_UNKNOWN		= 0,	/* unknown device */
 	ATA_DEV_ATA		= 1,	/* ATA device */
@@ -158,17 +159,21 @@
 	/* size of buffer to pad xfers ending on unaligned boundaries */
 	ATA_DMA_PAD_SZ		= 4,
 	ATA_DMA_PAD_BUF_SZ	= ATA_DMA_PAD_SZ * ATA_MAX_QUEUE,
+	
+	/* Masks for port functions */
+	ATA_PORT_PRIMARY	= (1 << 0),
+	ATA_PORT_SECONDARY	= (1 << 1),
 };
 
-enum pio_task_states {
-	PIO_ST_UNKNOWN,
-	PIO_ST_IDLE,
-	PIO_ST_POLL,
-	PIO_ST_TMOUT,
-	PIO_ST,
-	PIO_ST_LAST,
-	PIO_ST_LAST_POLL,
-	PIO_ST_ERR,
+enum hsm_task_states {
+	HSM_ST_UNKNOWN,
+	HSM_ST_IDLE,
+	HSM_ST_POLL,
+	HSM_ST_TMOUT,
+	HSM_ST,
+	HSM_ST_LAST,
+	HSM_ST_LAST_POLL,
+	HSM_ST_ERR,
 };
 
 /* forward declarations */
@@ -291,6 +296,11 @@
 	u8			xfer_protocol;	/* taskfile xfer protocol */
 	u8			read_cmd;	/* opcode to use on read */
 	u8			write_cmd;	/* opcode to use on write */
+
+	/* for CHS addressing */
+	u16			cylinders;	/* Number of cylinders */
+	u16			heads;		/* Number of heads */
+	u16			sectors;	/* Number of sectors per track */
 };
 
 struct ata_port {
@@ -331,7 +341,7 @@
 	struct work_struct	packet_task;
 
 	struct work_struct	pio_task;
-	unsigned int		pio_task_state;
+	unsigned int		hsm_task_state;
 	unsigned long		pio_task_timeout;
 
 	void			*private_data;
@@ -412,6 +422,8 @@
 extern int ata_scsi_error(struct Scsi_Host *host);
 extern int ata_scsi_release(struct Scsi_Host *host);
 extern unsigned int ata_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc);
+extern int ata_ratelimit(void);
+
 /*
  * Default driver ops implementations
  */
@@ -464,7 +476,7 @@
 
 extern void ata_pci_host_stop (struct ata_host_set *host_set);
 extern struct ata_probe_ent *
-ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port);
+ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int portmask);
 extern int pci_test_config_bits(struct pci_dev *pdev, struct pci_bits *bits);
 
 #endif /* CONFIG_PCI */
diff --git a/include/linux/textsearch.h b/include/linux/textsearch.h
index 941f45a..1a4990e 100644
--- a/include/linux/textsearch.h
+++ b/include/linux/textsearch.h
@@ -158,7 +158,8 @@
 #define TS_PRIV_ALIGNTO	8
 #define TS_PRIV_ALIGN(len) (((len) + TS_PRIV_ALIGNTO-1) & ~(TS_PRIV_ALIGNTO-1))
 
-static inline struct ts_config *alloc_ts_config(size_t payload, int gfp_mask)
+static inline struct ts_config *alloc_ts_config(size_t payload,
+						unsigned int __nocast gfp_mask)
 {
 	struct ts_config *conf;
 
diff --git a/include/net/dn_nsp.h b/include/net/dn_nsp.h
index 6bbeafa..8a0891e 100644
--- a/include/net/dn_nsp.h
+++ b/include/net/dn_nsp.h
@@ -19,9 +19,9 @@
 extern void dn_nsp_send_oth_ack(struct sock *sk);
 extern void dn_nsp_delayed_ack(struct sock *sk);
 extern void dn_send_conn_ack(struct sock *sk);
-extern void dn_send_conn_conf(struct sock *sk, int gfp);
+extern void dn_send_conn_conf(struct sock *sk, unsigned int __nocast gfp);
 extern void dn_nsp_send_disc(struct sock *sk, unsigned char type, 
-				unsigned short reason, int gfp);
+			unsigned short reason, unsigned int __nocast gfp);
 extern void dn_nsp_return_disc(struct sk_buff *skb, unsigned char type,
 				unsigned short reason);
 extern void dn_nsp_send_link(struct sock *sk, unsigned char lsflags, char fcval);
@@ -29,14 +29,14 @@
 
 extern void dn_nsp_output(struct sock *sk);
 extern int dn_nsp_check_xmit_queue(struct sock *sk, struct sk_buff *skb, struct sk_buff_head *q, unsigned short acknum);
-extern void dn_nsp_queue_xmit(struct sock *sk, struct sk_buff *skb, int gfp, int oob);
+extern void dn_nsp_queue_xmit(struct sock *sk, struct sk_buff *skb, unsigned int __nocast gfp, int oob);
 extern unsigned long dn_nsp_persist(struct sock *sk);
 extern int dn_nsp_xmit_timeout(struct sock *sk);
 
 extern int dn_nsp_rx(struct sk_buff *);
 extern int dn_nsp_backlog_rcv(struct sock *sk, struct sk_buff *skb);
 
-extern struct sk_buff *dn_alloc_skb(struct sock *sk, int size, int pri);
+extern struct sk_buff *dn_alloc_skb(struct sock *sk, int size, unsigned int __nocast pri);
 extern struct sk_buff *dn_alloc_send_skb(struct sock *sk, size_t *size, int noblock, long timeo, int *err);
 
 #define NSP_REASON_OK 0		/* No error */
diff --git a/include/net/dn_route.h b/include/net/dn_route.h
index d084721..11fe973 100644
--- a/include/net/dn_route.h
+++ b/include/net/dn_route.h
@@ -15,7 +15,7 @@
     GNU General Public License for more details.
 *******************************************************************************/
 
-extern struct sk_buff *dn_alloc_skb(struct sock *sk, int size, int pri);
+extern struct sk_buff *dn_alloc_skb(struct sock *sk, int size, unsigned int __nocast pri);
 extern int dn_route_output_sock(struct dst_entry **pprt, struct flowi *, struct sock *sk, int flags);
 extern int dn_cache_dump(struct sk_buff *skb, struct netlink_callback *cb);
 extern int dn_cache_getroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg);
diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 35f49e6..f50f959 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -40,7 +40,7 @@
 struct inet_ehash_bucket {
 	rwlock_t	  lock;
 	struct hlist_head chain;
-} __attribute__((__aligned__(8)));
+};
 
 /* There are a few simple rules, which allow for local port reuse by
  * an application.  In essence:
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index 06b4235..ecb2b06 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -832,7 +832,7 @@
 
 extern int ip_vs_app_pkt_out(struct ip_vs_conn *, struct sk_buff **pskb);
 extern int ip_vs_app_pkt_in(struct ip_vs_conn *, struct sk_buff **pskb);
-extern int ip_vs_skb_replace(struct sk_buff *skb, int pri,
+extern int ip_vs_skb_replace(struct sk_buff *skb, unsigned int __nocast pri,
 			     char *o_buf, int o_len, char *n_buf, int n_len);
 extern int ip_vs_app_init(void);
 extern void ip_vs_app_cleanup(void);
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index a9d0d8c..b6e72f8 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -875,7 +875,7 @@
 } 
 #endif
 
-struct xfrm_policy *xfrm_policy_alloc(int gfp);
+struct xfrm_policy *xfrm_policy_alloc(unsigned int __nocast gfp);
 extern int xfrm_policy_walk(int (*func)(struct xfrm_policy *, int, int, void*), void *);
 int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl);
 struct xfrm_policy *xfrm_policy_bysel(int dir, struct xfrm_selector *sel,
@@ -931,4 +931,9 @@
 	}
 }
 
+static inline int xfrm_policy_id2dir(u32 index)
+{
+	return index & 7;
+}
+
 #endif	/* _NET_XFRM_H */
diff --git a/include/rxrpc/call.h b/include/rxrpc/call.h
index f48f27e..8118731 100644
--- a/include/rxrpc/call.h
+++ b/include/rxrpc/call.h
@@ -203,7 +203,7 @@
 				 size_t sioc,
 				 struct kvec *siov,
 				 uint8_t rxhdr_flags,
-				 int alloc_flags,
+				 unsigned int __nocast alloc_flags,
 				 int dup_data,
 				 size_t *size_sent);
 
diff --git a/include/rxrpc/message.h b/include/rxrpc/message.h
index 3a59df6..983d9f9 100644
--- a/include/rxrpc/message.h
+++ b/include/rxrpc/message.h
@@ -63,7 +63,7 @@
 			     uint8_t type,
 			     int count,
 			     struct kvec *diov,
-			     int alloc_flags,
+			     unsigned int __nocast alloc_flags,
 			     struct rxrpc_message **_msg);
 
 extern int rxrpc_conn_sendmsg(struct rxrpc_connection *conn, struct rxrpc_message *msg);
diff --git a/lib/ts_bm.c b/lib/ts_bm.c
index 2cc79112..1b61fce 100644
--- a/lib/ts_bm.c
+++ b/lib/ts_bm.c
@@ -127,7 +127,7 @@
 }
 
 static struct ts_config *bm_init(const void *pattern, unsigned int len,
-				 int gfp_mask)
+				 unsigned int __nocast gfp_mask)
 {
 	struct ts_config *conf;
 	struct ts_bm *bm;
diff --git a/lib/ts_fsm.c b/lib/ts_fsm.c
index d27c0a0..ef9779e 100644
--- a/lib/ts_fsm.c
+++ b/lib/ts_fsm.c
@@ -258,7 +258,7 @@
 }
 
 static struct ts_config *fsm_init(const void *pattern, unsigned int len,
-				     int gfp_mask)
+				     unsigned int __nocast gfp_mask)
 {
 	int i, err = -EINVAL;
 	struct ts_config *conf;
diff --git a/lib/ts_kmp.c b/lib/ts_kmp.c
index 73266b9..e45f0f0 100644
--- a/lib/ts_kmp.c
+++ b/lib/ts_kmp.c
@@ -87,7 +87,7 @@
 }
 
 static struct ts_config *kmp_init(const void *pattern, unsigned int len,
-				  int gfp_mask)
+				  unsigned int __nocast gfp_mask)
 {
 	struct ts_config *conf;
 	struct ts_kmp *kmp;
diff --git a/net/atm/atm_misc.c b/net/atm/atm_misc.c
index b2113c3..71abc99 100644
--- a/net/atm/atm_misc.c
+++ b/net/atm/atm_misc.c
@@ -25,7 +25,7 @@
 
 
 struct sk_buff *atm_alloc_charge(struct atm_vcc *vcc,int pdu_size,
-    int gfp_flags)
+    unsigned int __nocast gfp_flags)
 {
 	struct sock *sk = sk_atm(vcc);
 	int guess = atm_guess_pdu2truesize(pdu_size);
diff --git a/net/ax25/ax25_in.c b/net/ax25/ax25_in.c
index 810c9c7..73cfc34 100644
--- a/net/ax25/ax25_in.c
+++ b/net/ax25/ax25_in.c
@@ -123,7 +123,7 @@
 		}
 
 		skb_pull(skb, 1);	/* Remove PID */
-		skb->h.raw    = skb->data;
+		skb->mac.raw  = skb->nh.raw;
 		skb->nh.raw   = skb->data;
 		skb->dev      = ax25->ax25_dev->dev;
 		skb->pkt_type = PACKET_HOST;
diff --git a/net/decnet/af_decnet.c b/net/decnet/af_decnet.c
index 348f36b..34d4128 100644
--- a/net/decnet/af_decnet.c
+++ b/net/decnet/af_decnet.c
@@ -452,7 +452,8 @@
 	.obj_size = sizeof(struct dn_sock),
 };
 
-static struct sock *dn_alloc_sock(struct socket *sock, int gfp)
+static struct sock *dn_alloc_sock(struct socket *sock,
+				  unsigned int __nocast gfp)
 {
 	struct dn_scp *scp;
 	struct sock *sk = sk_alloc(PF_DECnet, gfp, &dn_proto, 1);
@@ -804,7 +805,8 @@
 	return rv;
 }
 
-static int dn_confirm_accept(struct sock *sk, long *timeo, int allocation)
+static int dn_confirm_accept(struct sock *sk, long *timeo,
+				unsigned int __nocast allocation)
 {
 	struct dn_scp *scp = DN_SK(sk);
 	DEFINE_WAIT(wait);
diff --git a/net/decnet/dn_nsp_out.c b/net/decnet/dn_nsp_out.c
index 53633d3..cd08244 100644
--- a/net/decnet/dn_nsp_out.c
+++ b/net/decnet/dn_nsp_out.c
@@ -117,7 +117,8 @@
  * The eventual aim is for each socket to have a cached header size
  * for its outgoing packets, and to set hdr from this when sk != NULL.
  */
-struct sk_buff *dn_alloc_skb(struct sock *sk, int size, int pri)
+struct sk_buff *dn_alloc_skb(struct sock *sk, int size,
+			     unsigned int __nocast pri)
 {
 	struct sk_buff *skb;
 	int hdr = 64;
@@ -210,7 +211,8 @@
  *
  * Returns: The number of times the packet has been sent previously
  */
-static inline unsigned dn_nsp_clone_and_send(struct sk_buff *skb, int gfp)
+static inline unsigned dn_nsp_clone_and_send(struct sk_buff *skb,
+					     unsigned int __nocast gfp)
 {
 	struct dn_skb_cb *cb = DN_SKB_CB(skb);
 	struct sk_buff *skb2;
@@ -350,7 +352,8 @@
 	return ptr;
 }
 
-void dn_nsp_queue_xmit(struct sock *sk, struct sk_buff *skb, int gfp, int oth)
+void dn_nsp_queue_xmit(struct sock *sk, struct sk_buff *skb,
+			unsigned int __nocast gfp, int oth)
 {
 	struct dn_scp *scp = DN_SK(sk);
 	struct dn_skb_cb *cb = DN_SKB_CB(skb);
@@ -517,7 +520,7 @@
 	return 0;
 }
 
-void dn_send_conn_conf(struct sock *sk, int gfp)
+void dn_send_conn_conf(struct sock *sk, unsigned int __nocast gfp)
 {
 	struct dn_scp *scp = DN_SK(sk);
 	struct sk_buff *skb = NULL;
@@ -549,7 +552,8 @@
 
 
 static __inline__ void dn_nsp_do_disc(struct sock *sk, unsigned char msgflg, 
-			unsigned short reason, int gfp, struct dst_entry *dst,
+			unsigned short reason, unsigned int __nocast gfp,
+			struct dst_entry *dst,
 			int ddl, unsigned char *dd, __u16 rem, __u16 loc)
 {
 	struct sk_buff *skb = NULL;
@@ -591,7 +595,7 @@
 
 
 void dn_nsp_send_disc(struct sock *sk, unsigned char msgflg, 
-			unsigned short reason, int gfp)
+			unsigned short reason, unsigned int __nocast gfp)
 {
 	struct dn_scp *scp = DN_SK(sk);
 	int ddl = 0;
@@ -612,7 +616,7 @@
 {
 	struct dn_skb_cb *cb = DN_SKB_CB(skb);
 	int ddl = 0;
-	int gfp = GFP_ATOMIC;
+	unsigned int __nocast gfp = GFP_ATOMIC;
 
 	dn_nsp_do_disc(NULL, msgflg, reason, gfp, skb->dst, ddl, 
 			NULL, cb->src_port, cb->dst_port);
@@ -624,7 +628,7 @@
 	struct dn_scp *scp = DN_SK(sk);
 	struct sk_buff *skb;
 	unsigned char *ptr;
-	int gfp = GFP_ATOMIC;
+	unsigned int __nocast gfp = GFP_ATOMIC;
 
 	if ((skb = dn_alloc_skb(sk, DN_MAX_NSP_DATA_HEADER + 2, gfp)) == NULL)
 		return;
@@ -659,7 +663,8 @@
 	unsigned char menuver;
 	struct dn_skb_cb *cb;
 	unsigned char type = 1;
-	int allocation = (msgflg == NSP_CI) ? sk->sk_allocation : GFP_ATOMIC;
+	unsigned int __nocast allocation =
+			(msgflg == NSP_CI) ? sk->sk_allocation : GFP_ATOMIC;
 	struct sk_buff *skb = dn_alloc_skb(sk, 200, allocation);
 
 	if (!skb)
diff --git a/net/ieee80211/ieee80211_tx.c b/net/ieee80211/ieee80211_tx.c
index c9aaff3..ecdf9f7 100644
--- a/net/ieee80211/ieee80211_tx.c
+++ b/net/ieee80211/ieee80211_tx.c
@@ -207,7 +207,7 @@
 }
 
 static struct ieee80211_txb *ieee80211_alloc_txb(int nr_frags, int txb_size,
-						 int gfp_mask)
+						 unsigned int __nocast gfp_mask)
 {
 	struct ieee80211_txb *txb;
 	int i;
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 50c0519..0093ea0 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -286,6 +286,8 @@
 
 static int halve_threshold = 25;
 static int inflate_threshold = 50;
+static int halve_threshold_root = 15;
+static int inflate_threshold_root = 25; 
 
 
 static void __alias_free_mem(struct rcu_head *head)
@@ -449,6 +451,8 @@
 	int i;
 	int err = 0;
 	struct tnode *old_tn;
+	int inflate_threshold_use;
+	int halve_threshold_use;
 
  	if (!tn)
 		return NULL;
@@ -541,10 +545,17 @@
 
 	check_tnode(tn);
 
+	/* Keep root node larger  */
+
+	if(!tn->parent)
+		inflate_threshold_use = inflate_threshold_root;
+	else 
+		inflate_threshold_use = inflate_threshold;
+
 	err = 0;
 	while ((tn->full_children > 0 &&
 	       50 * (tn->full_children + tnode_child_length(tn) - tn->empty_children) >=
-				inflate_threshold * tnode_child_length(tn))) {
+				inflate_threshold_use * tnode_child_length(tn))) {
 
 		old_tn = tn;
 		tn = inflate(t, tn);
@@ -564,10 +575,18 @@
 	 * node is above threshold.
 	 */
 
+
+	/* Keep root node larger  */
+
+	if(!tn->parent)
+		halve_threshold_use = halve_threshold_root;
+	else 
+		halve_threshold_use = halve_threshold;
+
 	err = 0;
 	while (tn->bits > 1 &&
 	       100 * (tnode_child_length(tn) - tn->empty_children) <
-	       halve_threshold * tnode_child_length(tn)) {
+	       halve_threshold_use * tnode_child_length(tn)) {
 
 		old_tn = tn;
 		tn = halve(t, tn);
diff --git a/net/ipv4/ipvs/ip_vs_app.c b/net/ipv4/ipvs/ip_vs_app.c
index 6e092da..b942ff3 100644
--- a/net/ipv4/ipvs/ip_vs_app.c
+++ b/net/ipv4/ipvs/ip_vs_app.c
@@ -604,7 +604,7 @@
 /*
  *	Replace a segment of data with a new segment
  */
-int ip_vs_skb_replace(struct sk_buff *skb, int pri,
+int ip_vs_skb_replace(struct sk_buff *skb, unsigned int __nocast pri,
 		      char *o_buf, int o_len, char *n_buf, int n_len)
 {
 	struct iphdr *iph;
diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
index 2cd7e7d..a765972 100644
--- a/net/ipv4/netfilter/Kconfig
+++ b/net/ipv4/netfilter/Kconfig
@@ -141,7 +141,7 @@
 	tristate  'PPTP protocol support'
 	help
 	  This module adds support for PPTP (Point to Point Tunnelling
-	  Protocol, RFC2637) conncection tracking and NAT. 
+	  Protocol, RFC2637) connection tracking and NAT. 
 	
 	  If you are running PPTP sessions over a stateful firewall or NAT
 	  box, you may want to enable this feature.  
diff --git a/net/ipv4/tcp_bic.c b/net/ipv4/tcp_bic.c
index b940346..6d80e06 100644
--- a/net/ipv4/tcp_bic.c
+++ b/net/ipv4/tcp_bic.c
@@ -136,7 +136,7 @@
 		else if (cwnd < ca->last_max_cwnd + max_increment*(BICTCP_B-1))
 			/* slow start */
 			ca->cnt = (cwnd * (BICTCP_B-1))
-				/ cwnd-ca->last_max_cwnd;
+				/ (cwnd - ca->last_max_cwnd);
 		else
 			/* linear increase */
 			ca->cnt = cwnd / max_increment;
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 519899f..39a96c7 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -1393,7 +1393,7 @@
 
 static int grec_size(struct ifmcaddr6 *pmc, int type, int gdel, int sdel)
 {
-	return sizeof(struct mld2_grec) + 4*mld_scount(pmc,type,gdel,sdel);
+	return sizeof(struct mld2_grec) + 16 * mld_scount(pmc,type,gdel,sdel);
 }
 
 static struct sk_buff *add_grhead(struct sk_buff *skb, struct ifmcaddr6 *pmc,
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 555a313..305d9ee 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1450,7 +1450,7 @@
 
 static void pndisc_redo(struct sk_buff *skb)
 {
-	ndisc_rcv(skb);
+	ndisc_recv_ns(skb);
 	kfree_skb(skb);
 }
 
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index e4cad11..bf95193 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -99,7 +99,7 @@
 		next:;
 		}
 		result = best;
-		for(;; result += UDP_HTABLE_SIZE) {
+		for(i = 0; i < (1 << 16) / UDP_HTABLE_SIZE; i++, result += UDP_HTABLE_SIZE) {
 			if (result > sysctl_local_port_range[1])
 				result = sysctl_local_port_range[0]
 					+ ((result - sysctl_local_port_range[0]) &
@@ -107,6 +107,8 @@
 			if (!udp_lport_inuse(result))
 				break;
 		}
+		if (i >= (1 << 16) / UDP_HTABLE_SIZE)
+			goto fail;
 gotit:
 		udp_port_rover = snum = result;
 	} else {
diff --git a/net/key/af_key.c b/net/key/af_key.c
index 4879743..bbf0f69 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -185,7 +185,7 @@
 }
 
 static int pfkey_broadcast_one(struct sk_buff *skb, struct sk_buff **skb2,
-			       int allocation, struct sock *sk)
+			       unsigned int __nocast allocation, struct sock *sk)
 {
 	int err = -ENOBUFS;
 
@@ -217,7 +217,7 @@
 #define BROADCAST_ONE		1
 #define BROADCAST_REGISTERED	2
 #define BROADCAST_PROMISC_ONLY	4
-static int pfkey_broadcast(struct sk_buff *skb, int allocation,
+static int pfkey_broadcast(struct sk_buff *skb, unsigned int __nocast allocation,
 			   int broadcast_flags, struct sock *one_sk)
 {
 	struct sock *sk;
@@ -1416,7 +1416,8 @@
 	return 0;
 }
 
-static struct sk_buff *compose_sadb_supported(struct sadb_msg *orig, int allocation)
+static struct sk_buff *compose_sadb_supported(struct sadb_msg *orig,
+					      unsigned int __nocast allocation)
 {
 	struct sk_buff *skb;
 	struct sadb_msg *hdr;
@@ -2153,6 +2154,7 @@
 
 static int pfkey_spdget(struct sock *sk, struct sk_buff *skb, struct sadb_msg *hdr, void **ext_hdrs)
 {
+	unsigned int dir;
 	int err;
 	struct sadb_x_policy *pol;
 	struct xfrm_policy *xp;
@@ -2161,7 +2163,11 @@
 	if ((pol = ext_hdrs[SADB_X_EXT_POLICY-1]) == NULL)
 		return -EINVAL;
 
-	xp = xfrm_policy_byid(0, pol->sadb_x_policy_id,
+	dir = xfrm_policy_id2dir(pol->sadb_x_policy_id);
+	if (dir >= XFRM_POLICY_MAX)
+		return -EINVAL;
+
+	xp = xfrm_policy_byid(dir, pol->sadb_x_policy_id,
 			      hdr->sadb_msg_type == SADB_X_SPDDELETE2);
 	if (xp == NULL)
 		return -ENOENT;
@@ -2173,9 +2179,9 @@
 	if (hdr->sadb_msg_type == SADB_X_SPDDELETE2) {
 		c.data.byid = 1;
 		c.event = XFRM_MSG_DELPOLICY;
-		km_policy_notify(xp, pol->sadb_x_policy_dir-1, &c);
+		km_policy_notify(xp, dir, &c);
 	} else {
-		err = key_pol_get_resp(sk, xp, hdr, pol->sadb_x_policy_dir-1);
+		err = key_pol_get_resp(sk, xp, hdr, dir);
 	}
 
 	xfrm_pol_put(xp);
diff --git a/net/netfilter/nfnetlink.c b/net/netfilter/nfnetlink.c
index 49a3900..34d6719 100644
--- a/net/netfilter/nfnetlink.c
+++ b/net/netfilter/nfnetlink.c
@@ -195,7 +195,8 @@
 
 int nfnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo)
 {
-	int allocation = in_interrupt() ? GFP_ATOMIC : GFP_KERNEL;
+	unsigned int __nocast allocation =
+			in_interrupt() ? GFP_ATOMIC : GFP_KERNEL;
 	int err = 0;
 
 	NETLINK_CB(skb).dst_group = group;
diff --git a/net/netrom/nr_dev.c b/net/netrom/nr_dev.c
index 4e66eef..509afdd 100644
--- a/net/netrom/nr_dev.c
+++ b/net/netrom/nr_dev.c
@@ -58,7 +58,7 @@
 
 	/* Spoof incoming device */
 	skb->dev      = dev;
-	skb->h.raw    = skb->data;
+	skb->mac.raw  = skb->nh.raw;
 	skb->nh.raw   = skb->data;
 	skb->pkt_type = PACKET_HOST;
 
diff --git a/net/rxrpc/call.c b/net/rxrpc/call.c
index 5cfd4ca..86f7770 100644
--- a/net/rxrpc/call.c
+++ b/net/rxrpc/call.c
@@ -1923,7 +1923,7 @@
 			  size_t sioc,
 			  struct kvec *siov,
 			  u8 rxhdr_flags,
-			  int alloc_flags,
+			  unsigned int __nocast alloc_flags,
 			  int dup_data,
 			  size_t *size_sent)
 {
diff --git a/net/rxrpc/connection.c b/net/rxrpc/connection.c
index 61463c7..be4b2be 100644
--- a/net/rxrpc/connection.c
+++ b/net/rxrpc/connection.c
@@ -522,7 +522,7 @@
 		      uint8_t type,
 		      int dcount,
 		      struct kvec diov[],
-		      int alloc_flags,
+		      unsigned int __nocast alloc_flags,
 		      struct rxrpc_message **_msg)
 {
 	struct rxrpc_message *msg;
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index f310403..ade730e 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -719,7 +719,7 @@
 void *
 rpc_malloc(struct rpc_task *task, size_t size)
 {
-	int	gfp;
+	unsigned int __nocast	gfp;
 
 	if (task->tk_flags & RPC_TASK_SWAPPER)
 		gfp = GFP_ATOMIC;
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index fda737d..061b44c 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -163,7 +163,7 @@
 	if (xp->dead)
 		goto out;
 
-	dir = xp->index & 7;
+	dir = xfrm_policy_id2dir(xp->index);
 
 	if (xp->lft.hard_add_expires_seconds) {
 		long tmo = xp->lft.hard_add_expires_seconds +
@@ -225,7 +225,7 @@
  * SPD calls.
  */
 
-struct xfrm_policy *xfrm_policy_alloc(int gfp)
+struct xfrm_policy *xfrm_policy_alloc(unsigned int __nocast gfp)
 {
 	struct xfrm_policy *policy;
 
@@ -417,7 +417,7 @@
 	struct xfrm_policy *pol, **p;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = &xfrm_policy_list[id & 7]; (pol=*p)!=NULL; p = &pol->next) {
+	for (p = &xfrm_policy_list[dir]; (pol=*p)!=NULL; p = &pol->next) {
 		if (pol->index == id) {
 			xfrm_pol_hold(pol);
 			if (delete)