This gives +10% performance on simple tests, so definitly worth it.
A few percent more could be had by not using M_ZERO'd alloc's, but
we then need to clear fields all over the place to be safe, and
that was deemed not worth the trouble (and it makes life dangerous).
of the leftovers from the old version that really doesn't work anymore.
Add a reset function for host-end of the ATA channel. This is needed
for the SiI3112 in order to whack it back to reality if a device
locks up the SATA interface (thereby preventing that we can reset the
device). The result is that ATA now recovers from the timeouts that
happens with the SiI3112A and more or less all disks based on old
PATA electronics with a Marvell PATA->SATA converter. This includes
lots of the popular SATA dongles and the WDC Raptor disks..
in connection with Marvell based SATA->PATA dongles.
The problem was caused by a combination of things working
together to make it hard to spot...
The ATA driver has always started the ATA command, then build
the SG list for DMA and then finally started the DMA engine.
While this is according to specs, it poses a potential
problem as some controllers apparently do not allow for unlimitted
time between starting the ATA command and starting the DMA engine.
At about the same time as ATAng was committed there were lots
of other changes applied, some of which was locking in parts
that causes the busdma load functions to take significantly
longer to load the SG list.
This pushed the time spent between starting the ATA command and
starting the DMA engine over the hill for some controllers
(especially the Silicon Image DS3112a) and caused what looked
like lost interrupts.
The solution is to get all the SG list work or rather all
busdma related stuff done before we even try to start anything.
This has the nice side effect of seperating busdma out the
way it should be, so the working of the ATA machinery is not
cluttered up with busdma droppings, making the code easier
to read and understand.
Restructure the way ATA/ATAPI commands are processed, use a common
ata_request structure for both. This centralises the way requests
are handled so locking is much easier to handle.
The driver is now layered much more cleanly to seperate the lowlevel
HW access so it can be tailored to specific controllers without touching
the upper layers. This is needed to support some of the newer
semi-intelligent ATA controllers showing up.
The top level drivers (disk, ATAPI devices) are more or less still
the same with just corrections to use the new interface.
Pull ATA out from under Gaint now that locking can be done in a sane way.
Add support for a the National Geode SC1100. Thanks to Soekris engineering
for sponsoring a Soekris 4801 to make this support.
Fixed alot of small bugs in the chipset code for various chips now
we are around in that corner anyways.
Add two new arguments to bus_dma_tag_create(): lockfunc and lockfuncarg.
Lockfunc allows a driver to provide a function for managing its locking
semantics while using busdma. At the moment, this is used for the
asynchronous busdma_swi and callback mechanism. Two lockfunc implementations
are provided: busdma_lock_mutex() performs standard mutex operations on the
mutex that is specified from lockfuncarg. dftl_lock() is a panic
implementation and is defaulted to when NULL, NULL are passed to
bus_dma_tag_create(). The only time that NULL, NULL should ever be used is
when the driver ensures that bus_dmamap_load() will not be deferred.
Drivers that do not provide their own locking can pass
busdma_lock_mutex,&Giant args in order to preserve the former behaviour.
sparc64 and powerpc do not provide real busdma_swi functions, so this is
largely a noop on those platforms. The busdma_swi on is64 is not properly
locked yet, so warnings will be emitted on this platform when busdma
callback deferrals happen.
If anyone gets panics or warnings from dflt_lock() being called, please
let me know right away.
Reviewed by: tmm, gibbs
Clean up the DMA interface too much unneeded stuff crept in with
the busdma code back when.
Modify the ATA_IN* / ATA_OUT* macros so that resource and offset
are gotten from a table. That allows for new chipsets that doesn't
nessesarily have things ordered the good old way. This also removes
the need for the wierd PC98 resource functions.
Tested on: i386, PC98, Alpha, Sparc64
This moves all chipset specific code to a new file 'ata-chipset.c'.
Extensive use of tables and pointers to avoid having the same switch
on chipset type in several places, and to allow substituting various
functions for different HW arch needs.
Added PIO mode setup and all DMA modes.
Support for all known SiS chipsets. Thanks to Christoph Kukulies for
sponsoring a nice ASUS P4S8X SiS648 based board for this work!
Tested on: i386, PC98, alpha and sparc64
This mostly consists of functionality to serialize accesses to
the two ATA channels (which can also be used to "fix" certain
PCI based controllers).
Add support for Acard controllers.
Enable the ATA driver in PC98 GENERIC, and add device hints.
Update man page with latest support.
The PC98 core team has kindly provided me with a PC98
machine that made this all possible, thanks to all that
contributed to that effort, without that this would
probably newer have been possible..
Approved by: re@
Link if only ATAPI device in kernel config
Remove unused #includes
Rearrange a bit in ata-raid to make diff against -stable smaller
Enable wc as default again, dunne how this happend...
Add code to properly detach/attach disks that are part of a RAID.
Mark a disk that is attached on an ATA channel belonging to a
RAID as a spare disk that can be used for rebuilding failed RAID1's.
Add support for rebuilding failed RAID1's.
Several fixes to the detach/attach code.
For replacing a disk in a failed RAID1 do the following:
Find the controller channel# of the failed disk.
Exec 'atacontrol detach <channel#>' to free the disk from the system.
Replace the failed disk with a new one of at least the same size.
If your have your disks in drawers/enclosures this can be done with
the system still running.
Exec 'atacontrol attach <channel#>' to add the disk to the system and
mark it as a valid spare for rebuild.
Exec 'atacontrol rebuild <array#>'
The system will rebuild the array on the fly, the array can still
be used during this, although with slower performance.
Please let me know of any problems with this!
Sponsored by: Advanis Inc.
MFC after: 2 weeks
Overhaul of the attach/detach code and structures, there were some nasty
bugs in the old implementation. This made it possible to collapse the
ATA/ATAPI device control structures into one generic structure.
A note here, the kernel is NOT ready for detach of active devices,
it fails all over in random places, but for inactive devices it works.
However for ATA RAID this works, since the RAID abstration layer
insulates the buggy^H^H^H^H^H^Hfragile device subsystem from the
physical disks.
Proberly detect the RAID's from the BIOS, and mark critical RAID1
arrays as such, but continue if there is enough of the mirror left
to do so.
Properly fail arrays on a live system. For RAID0 that means return EIO,
and for RAID1 it means continue on the still working part of the mirror
if possible, else return EIO.
If the state changes, log this to the console.
Allow for Promise & Highpoint controllers/arrays to coexist on the
same machine. It is not possible to distribute arrays over different
makes of controllers though.
If Promise SuperSwap enclosures are used, signal disk state on the
status LED on the front.
Misc fixes that I had lying around for various minor bugs.
Sponsored by: Advanis Inc.
