No reason to get overzealous about recovered comm and data errors.
Some PHYs habitually sets them w/o no good reason and being draconian
about these soft error conditions doesn't seem to help anybody.

If need ever rises, we might need to add soft PHY error condition, say
AC_ERR_MAYBE_ATA_BUS and use it only to determine whether speed down
is necessary but I don't think that's very likely to happen.  It's far
more likely we'll get timeouts or fatal transmission errors if
recovered errors are so prominent that they hamper operation.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Originally, whole reset processing was done while the port is frozen
and SError was cleared during @postreset().  This had two race
conditions.  1: hotplug could occur after reset but before SError is
cleared and libata won't know about it.  2: hotplug could occur after
all the reset is complete but before the port is thawed.  As all
events are cleared on thaw, the hotplug event would be lost.

Commit ac371987 kills the first race
by clearing SError during link resume but before link onlineness test.
However, this doesn't fix race #2 and in some cases clearing SError
after SRST is a good idea.

This patch solves this problem by cross checking link onlineness with
classification result after SError is cleared and port is thawed.
Reset is retried if link is online but all devices attached to the
link are unknown.  As all devices will be revalidated, this one-way
check is enough to ensure that all devices are detected and
revalidated reliably.

This, luckily, also fixes the cases where host controller returns
bogus status while harddrive is spinning up after hotplug making
classification run before the device sends the first FIS and thus
causes misdetection.

Low level drivers can bypass the logic by setting class explicitly to
ATA_DEV_NONE if ever necessary (currently none requires this).
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Previously reset freeze/thaw handling lived outside of ata_eh_reset()
mainly because the original PMP reset code needed the port frozen
while resetting all the fan-out ports, which is no longer the case.

This patch moves freeze/thaw handling into ata_eh_reset().
@prereset() and @postreset() are now called w/o freezing the port
although @prereset() an be called frozen if the port is frozen prior
to entering ata_eh_reset().

This makes code simpler and will help removing hotplug event related
races.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Reorganize ata_eh_reset() such that @prereset() is called even when no
reset method is available and if block is used instead of goto to skip
actual reset.  This makes no reset case behave better (readiness wait)
and future changes easier.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Export ata_eh_analyze_ncq_error() for subsequent use by sata_mv,
as suggested by Tejun.
Signed-off-by: NMark Lord <mlord@pobox.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Fix mis-reporting of NCQ errors by ensuring that result_tf->flags
is properly initialized in libata-eh.  This allows ata_gen_ata_sense()
to report the failed block number correctly to SCSI after a media error
during NCQ.

This patch may also be a candidate for backporting to earlier kernels.
Without this fix, SCSI will fail I/O on the entire request rather
than just the bad sector.  That can be bad for a request that was
merged from many independent read reads from different tasks.
Signed-off-by: NMark Lord <mlord@pobox.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

When no reset method is available, libata currently oopses.  Although
the condition can't happen unless there's a bug in a low level driver,
oopsing isn't the best way to report the error condition.  Complain,
dump stack and fail reset instead.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Currently, SATA softresets should do link onlineness check before
actually performing SRST protocol but it doesn't really belong to
softreset.

This patch moves onlineness check in softreset to ata_eh_reset() and
ata_eh_followup_srst_needed() to clean up code and help future sata_mv
changes which need clear separation between SCR and TF accesses.

sata_fsl is peculiar in that its softreset really isn't softreset but
combination of hardreset and softreset.  This patch adds dummy private
->prereset to keep the current behavior but the driver really should
implement separate hard and soft resets and return -EAGAIN from
hardreset if it should be follwed by softreset.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Some code paths which had been made obsolete by recent reset
simplification were still around.  Kill them.

* ata_eh_reset() checked for ATA_DEV_UNKNOWN to determine
  classification failure.  This is no longer applicable.

* ata_do_reset() should convert ATA_DEV_UNKNOWN to ATA_DEV_NONE
  regardless of reset result (e.g. -EAGAIN).

* LLDs don't need to convert ATA_DEV_UNKNOWN to ATA_DEV_NONE.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

Implement helpers to test whether PMP is supported, attached and
determine pmp number to use when issuing SRST to a link.  While at it,
move ata_is_host_link() so that it's together with the two new PMP
helpers.

This change simplifies LLDs and helps making PMP support optional.
Signed-off-by: NTejun Heo <htejun@gmail.com>

Previously, there were two ways to trigger follow-up SRST from
hardreset method - returning -EAGAIN and leaving all device classes
unmodified.  Drivers never used the latter mechanism and the only use
case for the former was when hardreset couldn't classify.

Drop the latter mechanism and let -EAGAIN mean "perform follow-up SRST
if classification is required".  This change removes unnecessary
follow-up SRSTs and simplifies reset implementations.
Signed-off-by: NTejun Heo <htejun@gmail.com>

If PMP fan-out reset fails and SCR isn't accessible, PMP should be
reset.  This used to be tested by sata_pmp_std_hardreset() and
communicated to EH by -ERESTART.  However, this logic is generic and
doesn't really have much to do with specific hardreset implementation.

This patch moves SCR access failure detection logic to ata_eh_reset()
where it belongs.  As this makes sata_pmp_std_hardreset() identical to
sata_std_hardreset(), the function is killed and replaced with the
standard method.
Signed-off-by: NTejun Heo <htejun@gmail.com>

Implement sata_std_hardreset(), which simply wraps around
sata_link_hardreset().  sata_std_hardreset() becomes new standard
hardreset method for sata_port_ops and sata_sff_hardreset() moves from
ata_base_port_ops to ata_sff_port_ops, which is where it really
belongs.

ata_is_builtin_hardreset() is added so that both
ata_std_error_handler() and ata_sff_error_handler() skip both builtin
hardresets if SCR isn't accessible.

piix_sidpr_hardreset() in ata_piix.c is identical to
sata_std_hardreset() in functionality and got replaced with the
standard function.
Signed-off-by: NTejun Heo <htejun@gmail.com>

SFF functions have confusing names.  Some have sff prefix, some have
bmdma, some std, some pci and some none.  Unify the naming by...

* SFF functions which are common to both BMDMA and non-BMDMA are
  prefixed with ata_sff_.

* SFF functions which are specific to BMDMA are prefixed with
  ata_bmdma_.

* SFF functions which are specific to PCI but apply to both BMDMA and
  non-BMDMA are prefixed with ata_pci_sff_.

* SFF functions which are specific to PCI and BMDMA are prefixed with
  ata_pci_bmdma_.

* Drop generic prefixes from LLD specific routines.  For example,
  bfin_std_dev_select -> bfin_dev_select.

The following renames are noteworthy.

  ata_qc_issue_prot() -> ata_sff_qc_issue()
  ata_pci_default_filter() -> ata_bmdma_mode_filter()
  ata_dev_try_classify() -> ata_sff_dev_classify()

This rename is in preparation of separating SFF support out of libata
core layer.  This patch strictly renames functions and doesn't
introduce any behavior difference.
Signed-off-by: NTejun Heo <htejun@gmail.com>

Currently whether a command should be retried after failure is
determined inside ata_eh_finish().  Add ATA_QCFLAG_RETRY and move the
logic into ata_eh_autopsy().  This makes things clearer and helps
extending retry determination logic.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Currently reset methods are not specified directly in the
ata_port_operations table.  If a LLD wants to use custom reset
methods, it should construct and use a error_handler which uses those
reset methods.  It's done this way for two reasons.

First, the ops table already contained too many methods and adding
four more of them would noticeably increase the amount of necessary
boilerplate code all over low level drivers.

Second, as ->error_handler uses those reset methods, it can get
confusing.  ie. By overriding ->error_handler, those reset ops can be
made useless making layering a bit hazy.

Now that ops table uses inheritance, the first problem doesn't exist
anymore.  The second isn't completely solved but is relieved by
providing default values - most drivers can just override what it has
implemented and don't have to concern itself about higher level
callbacks.  In fact, there currently is no driver which actually
modifies error handling behavior.  Drivers which override
->error_handler just wraps the standard error handler only to prepare
the controller for EH.  I don't think making ops layering strict has
any noticeable benefit.

This patch makes ->prereset, ->softreset, ->hardreset, ->postreset and
their PMP counterparts propoer ops.  Default ops are provided in the
base ops tables and drivers are converted to override individual reset
methods instead of creating custom error_handler.

* ata_std_error_handler() doesn't use sata_std_hardreset() if SCRs
  aren't accessible.  sata_promise doesn't need to use separate
  error_handlers for PATA and SATA anymore.

* softreset is broken for sata_inic162x and sata_sx4.  As libata now
  always prefers hardreset, this doesn't really matter but the ops are
  forced to NULL using ATA_OP_NULL for documentation purpose.

* pata_hpt374 needs to use different prereset for the first and second
  PCI functions.  This used to be done by branching from
  hpt374_error_handler().  The proper way to do this is to use
  separate ops and port_info tables for each function.  Converted.
Signed-off-by: NTejun Heo <htejun@gmail.com>

ata_ehi_schedule_probe() was created to hide details of link-resuming
reset magic.  Now that all the softreset workarounds are gone,
scheduling probe is very simple - set probe_mask and request RESET.
Kill ata_ehi_schedule_probe() and open code it.  This also increases
consistency as ata_ehi_schedule_probe() couldn't cover individual
device probings so they were open-coded even when the helper existed.

While at it, define ATA_ALL_DEVICES as mask of all possible devices on
a link and always use it when requesting probe on link level for
simplicity and consistency.  Setting extra bits in the probe_mask
doesn't hurt anybody.
Signed-off-by: NTejun Heo <htejun@gmail.com>

ATA_EHI_RESUME_LINK has two functions - promote reset to hardreset if
ATA_LFLAG_HRST_TO_RESUME is set and preventing EH from shortcutting
reset action when probing is requested.  The former is gone now and
the latter can easily be achieved by making EH to perform at least one
reset if reset is requested, which also makes more sense than
depending on RESUME_LINK flag.

As ATA_EHI_RESUME_LINK was the only EHI reset modifier, this also
kills reset modifier handling.
Signed-off-by: NTejun Heo <htejun@gmail.com>

When both soft and hard resets are available, libata preferred
softreset till now.  The logic behind it was to be softer to devices;
however, this doesn't really help much.  Rationales for the change:

* BIOS may freeze lock certain things during boot and softreset can't
  unlock those.  This by itself is okay but during operation PHY event
  or other error conditions can trigger hardreset and the device may
  end up with different configuration.

  For example, after a hardreset, previously unlockable HPA can be
  unlocked resulting in different device size and thus revalidation
  failure.  Similar condition can occur during or after resume.

* Certain ATAPI devices require hardreset to recover after certain
  error conditions.  On PATA, this is done by issuing the DEVICE RESET
  command.  On SATA, COMRESET has equivalent effect.  The problem is
  that DEVICE RESET needs its own execution protocol.

  For SFF controllers with bare TF access, it can be easily
  implemented but more advanced controllers (e.g. ahci and sata_sil24)
  require specialized implementations.  Simply using hardreset solves
  the problem nicely.

* COMRESET initialization sequence is the norm in SATA land and many
  SATA devices don't work properly if only SRST is used.  For example,
  some PMPs behave this way and libata works around by always issuing
  hardreset if the host supports PMP.

  Like the above example, libata has developed a number of mechanisms
  aiming to promote softreset to hardreset if softreset is not going
  to work.  This approach is time consuming and error prone.

  Also, note that, dependingon how you read the specs, it could be
  argued that PMP fan-out ports require COMRESET to start operation.
  In fact, all the PMPs on the market except one don't work properly
  if COMRESET is not issued to fan-out ports after PMP reset.

* COMRESET is an integral part of SATA connection and any working
  device should be able to handle COMRESET properly.  After all, it's
  the way to signal hardreset during reboot.  This is the most used
  and recommended (at least by the ahci spec) method of resetting
  devices.

So, this patch makes libata prefer hardreset over softreset by making
the following changes.

* Rename ATA_EH_RESET_MASK to ATA_EH_RESET and use it whereever
  ATA_EH_{SOFT|HARD}RESET used to be used.  ATA_EH_{SOFT|HARD}RESET is
  now only used to tell prereset whether soft or hard reset will be
  issued.

* Strip out now unneeded promote-to-hardreset logics from
  ata_eh_reset(), ata_std_prereset(), sata_pmp_std_prereset() and
  other places.
Signed-off-by: NTejun Heo <htejun@gmail.com>

EH actions are ATA_EH_* not ATA_EHI_*.  Rename ATA_EHI_LPM to
ATA_EH_LPM.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Cc: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Some old SFF controllers don't have any way to reset the channel.
Currently, this isn't supported and libata EH causes an oops.  Allow
LLDs w/o any reset method and just assume ATA class in such cases.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

This patch implements libata.force module parameter which can
selectively override ATA port, link and device configurations
including cable type, SATA PHY SPD limit, transfer mode and NCQ.

For example, you can say "use 1.5Gbps for all fan-out ports attached
to the second port but allow 3.0Gbps for the PMP device itself, oh,
the device attached to the third fan-out port chokes on NCQ and
shouldn't go over UDMA4" by the following.

 libata.force=2:1.5g,2.15:3.0g,2.03:noncq,udma4
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

ATA_ECAT_DUBIOUS_BASE was too high by one and thus all DUBIOUS error
categorizations were wrong.  This passed test because only ATA_BUS and
UNK_DEV were used during testing and the ones after them - ATA_BUS and
an overflowed entry - behaved similarly.

This patch fixes the problem by adding DUBIOUS_NONE category and use
it as base.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

ATA_PROT_ATAPI_* are ugly and naming schemes between ATA_PROT_* and
ATA_PROT_ATAPI_* are inconsistent causing confusion.  Rename them to
ATAPI_PROT_* and make them consistent with ATA counterpart.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

drivers/ata/libata-eh.c: In function `ata_port_pbar_desc':
drivers/ata/libata-eh.c:215: warning: long long unsigned int format, long unsigned int arg (arg 4)

Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

It's very likely that the configured data transfer mode is the wrong
one if device fails data transfers right after initial data transfer
mode configuration (including NCQ on/off and xfermode).  libata EH
needs to speed down fast before upper layers give up on probing.

This patch implement fast speed down rules to handle such cases
better.  Error occured while data transfer hasn't been verified
trigger fast back-to-back speed down actions until data transfer
works.

This change will make cable mis-detection and other initial
configuration problems corrected before partition scanning code gives
up.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

ATA_DFLAG_DUBIOUS_XFER is set whenever data transfer speed or method
changes and gets cleared when data transfer command succeeds in the
newly configured transfer mode.

This will be used to improve speed down logic.

Signed-off-by: Tejun Heo <htejun@gmail.com<
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Speed down rules were too conservative.  Adjust them a bit.

* More than 10 timeouts can't happen in 5 minutes as command timeout
  is 30secs.  Lower the limit for rule #1 to 6.

* 10 timeouts is too high for rule #3 too.  Lower it to 6.

* SATAPI can benefit from falling back to PIO too.  Allow SATAPI
  devices to fall back to PIO.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Clean up EH speed down implementation.

* is_io boolean variable is replaced eflags.  is_io is ATA_EFLAG_IS_IO.

* Error categories now have names.

* Better comments.

* Reorder 5min and 10min rules in ata_eh_speed_down_verdict()

* Use local variable @link to cache @dev->link in ata_eh_speed_down()

These changes are to improve readability and ease further changes.
This patch doesn't introduce any behavior change.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Move ata_set_mode() to libata-eh.c.  ata_set_mode() is surely an EH
action and will be more tightly coupled with the rest of error
handling.  Move it to libata-eh.c.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Factor out ata_eh_schedule_probe() from ata_eh_handle_dev_fail() and
ata_eh_recover().  This is to improve maintainability and make future
changes easier.

In the previous revision, ata_dev_enabled() test was accidentally
dropped while factoring out.  This problem was spotted by Bartlomiej.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

ACPI spec (ver 3.0a, p289) requires IDE power on/off executes ACPI _PSx
methods. As recently most PATA drivers use libata, this patch adds _PSx
method support in libata. ACPI spec doesn't mention if SATA requires the
same _PSx method.
Signed-off-by: NShaohua Li <shaohua.li@intel.com>
Acked-by: NLen Brown <len.brown@intel.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

After non-classifying reset, ehc->classes[] could contain
ATA_DEV_UNKNOWN which used to be normalized to ATA_DEV_NONE for
consistency.  However, this causes unfortunate side effect for drivers
which have non-classifying hardresets (e.g. sata_nv) by making
hardreset report ATA_DEV_NONE for non-classifying resets and thus
makes EH believe that the port is unoccupied and recovery can be
skipped.  The end result is that after a device is swapped with
another one, the new device isn't attached after the old one is
detached.

This patch makes ata_eh_reset() not normalize UNKNOWN to NONE after
non-classifying resets.  This fixes the above problem.  As UNKNOWN and
NONE are handled differently by only EH hotplug logic, this doesn't
cause other behavior changes.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Cc: Robert Hancock <hancockr@shaw.ca>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Timeout on downstream command may indicate transmission problem on
host link.  Propagate timeouts to host link.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

While updating lbam/h for ATAPI commands, atapi_eh_request_sense() was
left out.  Update it.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Protocol and CDB allocation size field are important in determining
what went wrong with ATAPI commands.  Report them on failure.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

This patch removes the following obsolete functions:
- libata-core.c: __sata_phy_reset()
- libata-core.c: sata_phy_reset()
- libata-eh.c: ata_qc_timeout()
- libata-eh.c: ata_eng_timeout()
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NTejun Heo <htejun@gmail.com>

libata EH used to ignore errors not associated with commands when
determining whether speed down is necessary or not.  This leads to the
following problems.

* Errors not associated with commands can occur indefinitely without
  libata EH taking corrective actions.

* Upstream link errors don't trigger speed down when PMP is attached
  to it and commands issued to downstream device trigger errors on the
  upstream link.

This patch makes ata_eh_link_autopsy() consider errors not associated
with command for speed down.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Reset failure is a critical error.  It results in disabling the link
requiring user intervention to re-enable it.  Make reset failure
handling more robust such that libata EH doesn't give up too early.

* Temporary glitches during hardreset may lead to classification
  failure when there's no softreset available.  Retry instead of
  giving up.

* Initial softreset or follow up softreset may fail classification.
  Move classification error handling block out of followup softreset
  block such that both cases are handled and retry instead of giving
  up.  Also, on the last try, give ATA class a blind shot.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>