Some Legacy megaraid cards can't actually cope with the scatter/gather
version of the READ CAPACITY command (which is what we now send them
since altering all SCSI internal I/O to go via the block layer).  Fix
this (and a few other broken megaraid driver assumptions) by sending
the non-sg version of the command if the sg list only has a single
element.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

In these drivers, scsi_remove_host() is called too late, at the point
it is called, the driver has already shut down too far to accept any
I/O that the shutdown might generate.  Any generated I/O actually
triggers a panic.

Fix this by calling scsi_remove_host() as early as possible and not
calling scsi_host_put() until just before we kfree the ahc_softc.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

There's a problem in our host release in that it calls
scsi_proc_hostdir_rm(). However, if you hold a reference to the host as
you remove the module, the host template (which proc uses) will be freed
and the system will panic when the host device is finally released.

Fix this by moving scsi_proc_hostdir_rm() to where it should be: in
scsi_remove_host().
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

A bunch of create_proc_dir_entry() calls creating directories had crept
in since the last sweep; converted to proc_mkdir().
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Received from Mark Salyzyn from Adaptec.

High Priority Queues have *never* been used in the entire history of the
aac based adapters. Associated with this, aac_insert_entry can be
removed, SavedIrql can be removed & padding variable can be removed.
With the movement of SavedIrql out & replaced with an automatic variable
qflags, the locking can be refined somewhat. The sparse warnings did not
catch the need for byte swapping in the 'dprintk' debugging print
macros, so fixed this up when this code was moved outside of the now
refined locking.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn from Adaptec.

The size of the command packet's scatter gather list maximum size was
miscalculated in the low range leading to the driver initialization
limiting the maximum i/o size that could go to the Adapter. There were
no negative operational side effects resulting from this bad math, only
a subtle limit in performance of the Adapter at the top end of the
range.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn from Adaptec.

In the rare instances where the adapter, or the motherboard, is
misbehaving; driver initialization or shutdown becomes problematic. By
introducing a 3 minute timeout on the first interrupt driven command
during initialization, or the issuance of the adapter shutdown command
during driver unload, we can resolve the lockup problems induced by
common (but rare) hardware misbehaviors.

The timeout during initialization, should it occur, is accompanied by a
message presented to the console and the logs indicating that the user
should inspect and resolve problems with interrupt routing.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch adds some additional error return checking and error return
value propagation during initialization. Also, the deprecation of
pci_module_init with pci_register_driver along with the change in return
values.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn from Adaptec.

Hotplug sniffs the AIFs (events) from the adapter and if a container
change resulting in the device going offline (container zero), online
(container zero completed) or changing capacity (morph) it will take
actions by calling the appropriate API.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Recevied from Mark Salyzyn from Adaptec.

Aif pre-allocation is used to pull the kmalloc outside of the locks.

Applies to the scsi-misc-2.6 git tree.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn from Adaptec.

There are a few adapters that are capable of creating devices with this large
of a capacity, but now that we have the large fib support in, the management
applications will be capable of generating them.  The problem is, once they are
created, the driver will not be able to access the devices correctly without
this patch.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NSreenivas Bagalkote <Sreenivas.Bagalkote@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Brown paperbag bug:  sas_rphy_delete was ordered completely
wrong.  Fix it up to be the same order as sas_phy_delete or
fc_rport_terminate and fix rphy objects that leaked after module
removal.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch fixes a hole in the rport unblock handling when processing
fabric events via the ADISC/PLOGI device state machine.  Original code
would not properly 'unblock' the port upon the port reloging into the
fabric.
Signed-off-by: NAndrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Currently we just ignore the device, which means there are a few
arrays out there that we don't find.

This patch updates the scsi_report_lun_scan() to take a target instead
of a device so it can be called on a return of
SCSI_SCAN_TARGET_PRESENT, which is what a PQ 3 device returns.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NAdam Radford <linuxraid@amcc.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The PowerMac mesh SCSI driver had some missing error handling which would
trigger warnings due to lack of handling of return value from
scsi_add_host.  This fixes it.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: James Bottomley <James.Bottomley@SteelEye.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NAndy Currid <acurrid@nvidia.com>
Signed-off-by: NJeff Garzik <jgarzik@pobox.com>

gcc-2.95.x doesn't do anonymous unions.
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

From: 	Alan Stern <stern@rowland.harvard.edu>

This patch (as561) fixes the error handler's thread-exit code.  The
kthread_stop call won't wake the thread from a down_interruptible, so
the patch gets rid of the semaphore and simply does

        set_current_state(TASK_INTERRUPTIBLE);
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>

Modified to simplify the termination loop and correct the sleep condition.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

We fix the oops by enforcing the host state model.  There have also
been two extra states added: SHOST_CANCEL_RECOVERY and
SHOST_DEL_RECOVERY so we can take the model through host removal while
the recovery thread is active.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch (as545) fixes the list traversals in __scsi_remove_target and
scsi_forget_host.  In each case the existing code list_for_each_entry_safe
in an _unsafe_ manner, because the list was not protected from outside
modification while the iteration was running.

The new scsi_forget_host routine takes the moderately controversial step
of iterating over devices for removal rather than iterating over targets.
This makes more sense to me because the current scheme treats targets as
second-class citizens, created and removed on demand, rather than as
objects corresponding to actual hardware.  (Also I couldn't figure out any
safe way to iterate over the target list, since it's not so easy to tell
when a target has already been removed.)
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

I found one other thing that needs to be fixed.  The call to
scsi_release_buffers in scsi_unprep_request causes an oops, because the
sgtable has already been freed in scsi_io_completion.  The following patch
is needed.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

PCI_ROM_ADDRESS is a 32 bit register and as such should be accessed using
pci_bus_{read,write}_config_dword().  A recent audit of drivers/ turned up
several cases of byte- and word-sized accesses.  The harmful ones were fixed
by Linus directly.  This patches up one of the remaining
harmless-but-still-wrong cases caught in the dragnet.
Signed-off-by: NAdam Kropelin <akropel1@rochester.rr.com>
Cc: Andrew Vasquez <andrew.vasquez@qlogic.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Make sure we that completion is the final action we take; prior to this
change, another CPU may have changed ap->pio_task_state before we tested
it a final time.

Spotted by, and original patch by Albert Lee @ IBM.

Also includes a minor optimization:  eliminate a ton of unnecessary
queue_work() calls, simply by jumping to the beginning of the FSM
function ata_pio_task().

Signed-off-by: NAlan Cox <alan@redhat.com>
Signed-off-by: NJeff Garzik <jgarzik@pobox.com>

Signed-off-by: NAlan Cox <alan@redhat.com>
Signed-off-by: NJeff Garzik <jgarzik@pobox.com>

Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

From: Alan Cox <alan@redhat.com>

The virt_to_bus() wasn't correctly taken out of this driver.  It needs
to be able to track both physical and virtual addresses for its prd table.
Update the driver to do this with separate tracking entries.
Signed-off-by: NAlan Cox <alan@redhat.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

On Wed, 2005-09-14 at 18:06 +1000, Anton Blanchard wrote:
> And in particular it looks like the scsi_unprep_request in
> scsi_queue_insert is causing it. The following patch fixes the boot
> problems on the vscsi machine:

OK, my fault.  Your fix is almost correct .. I was going to do this
eventually, honest, because there's no need to unprep and reprep a
command that comes in through scsi_queue_insert().

However, I decided to leave it in to exercise the scsi_unprep_request()
path just to make sure it was working.  What's happening, I think, is
that we also use this path for retries.  Since we kill and reget the
command each time, the retries decrement is never seen, so we're
retrying forever.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Modules need a license to prevent kernel tainting.
Signed-off-by: NRandy Dunlap <rdunlap@xenotime.net>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This fixes an issue in scsi command initialization from a request
where sd, sr, st, and scsi_lib all fail to copy the request's
cmd_len to the scsi command's cmd_len field.
Signed-off-by: NTimothy Thelin <timothy.thelin@wdc.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

There is an uninitialized variable issue in sata_sis.c
Signed-off-by: NJeff Garzik <jgarzik@pobox.com>

This patch fixes a nasty typo I introduced in my previous patch (commit
f2c853bc). The right offset of the
second port in pure sata mode is 64 and not 0x64.
Thanks to Martin Schuster for pointing this to me
Signed-off-by: NArnaud Patard <apatard@mandriva.com>
 ---
Signed-off-by: NJeff Garzik <jgarzik@pobox.com>

This patch moves aic7xxx over to the dma_get_required_mask() API and
dumps its open coded memory check.

It also appears from this bug:

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167049

That 39 bit addressing doesn't work on older cards.  I surmise that the
AHC_LARGE_SCBS flag is the one that marks cards capable of using 39 bit
addressing, so I also folded that check into the code.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Linda Xie ever so gently pointed out that she had a patch
to preserve compatibility with older SLES targets, and I told
her we didn't need to push it to mainline.

This patch explicitly checks the version of the IBMVSCSI target
and ensures that large scatterlists are not sent to older
targets.
Signed-off-by: NLinda Xie <lxie@us.ibm.com>
Signed-off-by: NDave Boutcher <boutcher@us.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

They report being SCSI-3 but seem to give back rubbish to a
REPORT_LUNS command.  Force them to be sequentially scanned.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

set DID_NO_CONNECT for the BLKPREP_KILL case and correct a few
BLKPREP_DEFER cases that weren't checking for the need to plug the
queue.
Signed-Off-By: NMike Christie <michaelc@cs.wisc.edu>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The original API returned either an ERR_PTR() or a refcounted sdev.
Unfortunately, if it's successful, you need to do a scsi_device_put() on
the sdev otherwise the refcounting is wrong.

Everyone seems to expect that scsi_add_device() should be callable
without doing the ref put, so alter the API so it is (we still have
__scsi_add_device with the original behaviour).

The only actual caller that needs altering is the one in firewire ...
not because it gets this right, but because it acts on the error if one
is returned.
Acked-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>