src - OpenBSD base system

Age	Commit message (Collapse)	Author
2009-10-22	gc unused global	David Gwynne

2009-09-13	Unchecked copyout() calls cause premature death in furry animals, so	Miod Vallat
	check their return values in the ATA_S_COMPLETE ioctl handler. dlg@ agrees.
2009-02-16	Extend the scsi_adapter minphys() callback to take a struct scsi_link *	Miod Vallat
	as additional argument. This will allow intermediate layers between scsi devices such as sd and scsi host adapters to take appropriate action if necessary.
2008-09-25	Initialize 'capacity' to 0 and use tabs to indent. "won't hurt" miod@.	Kenneth R Westerback

2008-05-05	Don't set the adaptor template link's flags and quirks when trying to	Kenneth R Westerback
	set the flags and quirks for a device. Should fix weird issues where the device after an ATAPI device gets marked as ATAPI. First reported on misc@ by Daniel Wade, who also provided some useful testing. ok dlg@
2008-03-26	plug a mem leak. the array of ports wasnt freed on detach.	David Gwynne

2007-12-30	dont freeze lock or set caching modes on ata devices that arent disks, its	David Gwynne
	not a good idea. found by simon@ who plugged a sata dvd drive into ahci.
2007-12-29	the scsi layer always had function pointers for asking the hba about a	David Gwynne
	device before issuing scsi commands to it, but it was never implemented, never used, and no hba actually filled them in. i came along and added another two function pointers for the same thing. this cleans up the extra pointers. ok krw@ marco@ miod@ deraadt@
2007-12-28	provide vpd page 0, which lists which vpd pages we provide.	David Gwynne

2007-12-28	rename some functions to make it clear theyre used to emulate vpd responses	David Gwynne

2007-12-28	implement emulation of vpd page 83, the device identification page. if the	David Gwynne
	disk provides a wwn, use it, otherwise do our best to generate an id from the model and serial number.
2007-12-28	massage the way vpds are defined. rename the page used for the devices	David Gwynne
	serial number and split the vpd header out for use in other places. while here define the device identification page bits too. ok krw@ marco@
2007-12-09	unify synchronous scsi xfer completion handling. this started out as a fix	David Gwynne
	to the completions for some scsi commands we simply ack in the disk_cmd handler.
2007-12-09	remove the nosleep argument to ata_get_xfer. it is a relic from when i	David Gwynne
	used to allocated xfers out of a pool rather than as part of the hbas ccb structure.
2007-12-06	Ask for write cache and read look ahead to be turned	Jonathan Gray
	on if supported as per the wd changes. Some drives don't do this for us and it helps performance by a large amount. ok dlg@
2007-11-28	when we probe a port send an ata inquiry to the device and cache	David Gwynne
	the result in the ata_port struct. use this cached inquiry data when building the replies to scsi commands rather than fetching a new copy of ata inq every time. this shrinks the code by 100 lines. it also avoids a malloc in a scsi io path, which could be issued at a time when malloc isnt guaranteed to succeed, but a real scsi io definitely should.
2007-11-28	make ata controllers protect their own command lists so atascsi doesnt have	David Gwynne
	to continually go to splbio to ensure its safe to work on them. shrinks code a little.
2007-11-26	implement atascsi_probe_dev and atascsi_detach_dev. they let a hba tell	David Gwynne
	the midlayer(s) what to do.
2007-11-26	drive port probes from the scsi midlayer now that it will ask the adapter	David Gwynne
	if a device is there before doing any scsi commands. also implement a free path for when devices are detached. software hotplug has been tested on sili, and ahci is still working according to claudio@
2007-11-23	starting bits to implement hotplug of sili controllers. this adds the code	David Gwynne
	to clean up the atascsi stuff, and has sili call it when its going away. tested on an expresscard sili variant by kettenis@
2007-11-23	use copyin and copyout to get the actual ata data in and out of the kernel	David Gwynne
	for the ata ioctl used by atactl. i dont know how this worked before.
2007-11-23	implement translation of the scsi vpd page for fetching a disks serial	David Gwynne
	number. lets bioctl against an atascsi disk see the serial number. continually requested by henning@
2007-11-23	t10 has a whole spec on doing SCSI to ATA translation called SAT. i wish	David Gwynne
	id known about it when i wrote atascsi. anyway. it says we should claim SPC-3 when we fake INQUIRY.
2007-11-16	accept ATA ioctls so tools like atactl work on disks behind the atascsi	David Gwynne
	layer. i intended to write better ways to deal with disks, but i have been too slack. shame on me. this is less code than the equivilent in wdc/pciide, which amuses me greatly.
2007-10-01	More easy bzero() -> M_ZERO. Use 'p = malloc(sizeof(*p) ...' where	Kenneth R Westerback
	obvious.
2007-04-22	when an ata command returns an error, dont fill dmesg with messages about	David Gwynne
	it. returning an error to the scsi midlayer is all we need to do, which in turn will do the right thing.
2007-04-12	Do a security freeze lock like we do in wd(4) to deal with systems	Jonathan Gray
	that leave the security set on by default. "looks sane" grange@, "looks ok" pascoe@, "put it in" dlg@
2007-04-10	endian fixes for the read capacity handling.	David Gwynne

2007-03-24	Complete the right scsi_xfer when processing a disk sync request.	Christopher Pascoe

2007-03-23	Track the ATA xfer's state more closely - may help us identify why some	Christopher Pascoe
	are seeing a panic at halt time.
2007-03-23	Really clear the device register on the flush cache command. Last commit	Christopher Pascoe
	adjusted the packet and identify commands.
2007-03-23	Clear the device register on the flush cache command, in case something	Christopher Pascoe
	that shouldn't care about don't care bits does care.
2007-03-23	Catch timed out disk commands.	Christopher Pascoe

2007-03-22	Let atascsi issue NCQ commands if the controller supports it.	Christopher Pascoe
	This includes a nasty hack to reduce openings and throw away command slots if the device supports a lower queue depth than the host controller does. Yes, we're thinking about a better solution.
2007-03-21	A polled xfer has completed and been put by the time ata_cmd returns, so	Christopher Pascoe
	don't use xa->flags afterwards.
2007-03-21	atascsi sends h2d fis's, not d2h.	David Gwynne

2007-03-21	replace the array of u_int8_ts for the fis/registers with a struct to make	David Gwynne
	it a bit easier to read. there are variations on this, but they can be defined later.
2007-03-20	Be sure to reset ata_xfer flags for all PACKET transfers, otherwise we	Christopher Pascoe
	could pick up a previous ATA_F_POLL from a previous non-data command.
2007-03-20	Basic FLUSH CACHE support.	Christopher Pascoe
	Note that the ATA spec says we're should retry after error until we see no more errors. This is not (yet) implemented.
2007-03-20	Tag ATAPI transfers that have no data phase correctly, otherwise the command	Christopher Pascoe
	isn't sent to the device.
2007-03-20	Oops, remove some unused variables and unmangle a command name.	Christopher Pascoe

2007-03-20	Cap the device size we present to 2TB, just in case someone attaches	Christopher Pascoe
	something really big. The SCSI layer isn't yet ready for it.
2007-03-20	Move common completion handling into ata_exec itself. Removes a double free	Christopher Pascoe
	in the inquiry error paths, as the HBA will have already completed the ata_xfer with an error. Reminder and ok dlg@
2007-03-20	Add support for ATA PACKET commands. This should make ATAPI cdrom drives,	Christopher Pascoe
	etc, just work.
2007-03-20	move ata command definitions from atascsi.c to atascsi.h	David Gwynne

2007-03-20	nothing from wdc, wd, or any of the existing ata stuff pls. this gets rid	David Gwynne
	of the wdcreg.h include in atascsi. "feel free to remove" pascoe@
2007-03-20	Request "descriptor processed" interrupts only for PIO requests (inquiry,	Christopher Pascoe
	etc), and request that the D2H FIS at the completion of all (DMA) commands generate an interrupt. This makes each data transfer generate one interrupt instead of two and should guarantee that the interrupt that is generated actually arrives after the command has completed.
2007-03-20	Calculate the residual and copy it back to the scsi_xfer, so that RW actually	Christopher Pascoe
	works. From dlg@
2007-03-20	Add support for issuing sector read/write commands.	Christopher Pascoe

2007-03-20	Instead of having a copy of every register that we will issue in the ata_cmd,	Christopher Pascoe
	simply include a pointer to the command FIS that we will issue to the device. Include another space where we can copy back an error register set from a failed command. This means that we can now build and issue arbitrary commands from atascsi, and retrieve errors back.