Commit Graph

154 Commits

Author SHA1 Message Date
Jason Evans
0384fff8c5 Major update to the way synchronization is done in the kernel. Highlights
include:

* Mutual exclusion is used instead of spl*().  See mutex(9).  (Note: The
  alpha port is still in transition and currently uses both.)

* Per-CPU idle processes.

* Interrupts are run in their own separate kernel threads and can be
  preempted (i386 only).

Partially contributed by:	BSDi (BSD/OS)
Submissions by (at least):	cp, dfr, dillon, grog, jake, jhb, sheldonh
2000-09-07 01:33:02 +00:00
Doug Rabson
21c3015a24 * Completely rewrite the alpha busspace to hide the implementation from
the drivers.
* Remove legacy inx/outx support from chipset and replace with macros
  which call busspace.
* Rework pci config accesses to route through the pcib device instead of
  calling a MD function directly.

With these changes it is possible to cleanly support machines which have
more than one independantly numbered PCI busses. As a bonus, the new
busspace implementation should be measurably faster than the old one.
2000-08-28 21:48:13 +00:00
Søren Schmidt
b0bb0aeb95 Rearrange the probe a bit, hopefully this will help to eleminate
some of the fake devices sometimes seen on single device ATA
channels.

Proberly fail on failures in ata-disk.c, retry instead of hang.

Cleanup the VIA probe/init code a bit.

Fix a couble of missing free's in atapi-cd.c in the changer code.
2000-08-16 07:09:04 +00:00
Søren Schmidt
710591b796 Fix off by one error in track# calculation. 2000-08-13 09:08:19 +00:00
Søren Schmidt
ae97e232fb Backout the tsleep on probe, it breaks in resume mode as tsleep
not always has a context to sleep on.
Back to the drawing board...
2000-08-11 10:34:22 +00:00
Søren Schmidt
071e6e581e Add support for accessing individual tracks on a CD.
This is done by misusing the device minor a bit to encode the
track no there.

So to read track #4 just use /dev/acdNt4 where N is the device #.

The driver no automatically sets the blocksize (sectorsize) to
what the track is set to in the TOC.

This has the nice effect that you can now rip audioi tracks
by simply doing:

dd if=/dev/acdNt2 of=audiotrack2.raw bs=2352

it cant be much simpler than that :)

NOTE: the original acdNa & acdNc device still work as usual,
except the blocksize is set according to track0.
2000-08-11 09:21:36 +00:00
Søren Schmidt
4466a3504d Fix the ICH2 ATA100 support, and proberly downgrade the chip setup when
going to slower DMA modes on all Intel chips.
2000-08-10 07:34:49 +00:00
Søren Schmidt
bfc0a42650 Fix braino in the tsleep call, ata_delayed_attach might be a NULL
pointer..
2000-08-08 14:57:36 +00:00
Søren Schmidt
ab418d7db3 Add experimental code for ATA100 support on:
Promise Ultra100 / Fasttrak100
HighPoint HPT370 controllers (fx Abit KA7-100 onboard ctrl, Abit HotRod 100)
Intel ICH2 (Intel 815E based motherboards)

So far I can read >90MB/s on the Promise and the HPT370.
I can write >64MB/s on the promise and >50MB/s on the HPT370 so it seems
writing is still done in ATA66 mode :(
The ICH2 support is untested as of yet...
2000-08-06 19:51:58 +00:00
Søren Schmidt
19095fd23c Use tsleep instead od DELAY in probe when not in boot.
This fixes the uptil 30s hangs on PCCARD ata device probes.
2000-08-06 19:10:05 +00:00
Søren Schmidt
0ec4d7b5b0 Remove sequence checking in the burner support.
This was really a leftover from the wormcontrol days, it is no longer
needed when using burncd.
2000-08-06 18:03:42 +00:00
Søren Schmidt
e7d1a5c2b1 Be a bit more restrictive in using multisector transfers. 2000-08-06 18:01:47 +00:00
Steve Price
bf11d3dcda Spell the word 'limited' with only one 't'. 2000-07-30 17:54:34 +00:00
Poul-Henning Kamp
77978ab8bc Previous commit changing SYSCTL_HANDLER_ARGS violated KNF.
Pointed out by:	bde
2000-07-04 11:25:35 +00:00
Poul-Henning Kamp
82d9ae4e32 Style police catches up with rev 1.26 of src/sys/sys/sysctl.h:
Sanitize SYSCTL_HANDLER_ARGS so that simplistic tools can grog our
sources:

        -sysctl_vm_zone SYSCTL_HANDLER_ARGS
        +sysctl_vm_zone (SYSCTL_HANDLER_ARGS)
2000-07-03 09:35:31 +00:00
Poul-Henning Kamp
4bd02a5609 Add disk_enumerate() for finding names of disks. Vinum and libh will
need this RSN.

Remove a pointless warning in the root device locating code.

Remove the "wd" compatibility name from the "ad" driver.

WARNING: If you have not updated to use /dev/wd* in your /etc/fstab
and modern bootblocks, it would be a very good idea to do so BEFORE
you upgrade your kernel.
2000-06-15 20:30:53 +00:00
Doug Rabson
1536418a84 Brucify the pmap_enter_temporary() changes. 2000-05-29 19:21:01 +00:00
Doug Rabson
31891bc2bd Add a new pmap entry point, pmap_enter_temporary() to be used during
dumps to create temporary page mappings. This replaces the use of CADDR1
which is fairly x86 specific.

Reviewed by: dillon
2000-05-28 15:49:55 +00:00
Søren Schmidt
724a60faf8 Cosmetics: dont say Unknown but Generic chipset in probe.. 2000-05-28 07:51:23 +00:00
Søren Schmidt
5158543d6a Fix alignment problem on the alpha reported by several parties. 2000-05-28 07:50:19 +00:00
Jake Burkholder
e39756439c Back out the previous change to the queue(3) interface.
It was not discussed and should probably not happen.

Requested by:		msmith and others
2000-05-26 02:09:24 +00:00
Jake Burkholder
740a1973a6 Change the way that the queue(3) structures are declared; don't assume that
the type argument to *_HEAD and *_ENTRY is a struct.

Suggested by:	phk
Reviewed by:	phk
Approved by:	mdodd
2000-05-23 20:41:01 +00:00
Søren Schmidt
5996ceaaa9 Fix a nasty bug in ata_intr, parens are a good thing if used right.
Make the error recovery code a little more obvious.

Inform the user if UDMA66 mode couldn't be selected due to a
non ATA66 compliant 80pin cable.

Minor cosmetics.
2000-05-23 19:05:56 +00:00
Peter Wemm
0385347c1a Implement an optimization of the VM<->pmap API. Pass vm_page_t's directly
to various pmap_*() functions instead of looking up the physical address
and passing that.  In many cases, the first thing the pmap code was doing
was going to a lot of trouble to get back the original vm_page_t, or
it's shadow pv_table entry.

Inspired by: John Dyson's 1998 patches.

Also:
Eliminate pv_table as a seperate thing and build it into a machine
dependent part of vm_page_t.  This eliminates having a seperate set of
structions that shadow each other in a 1:1 fashion that we often went to
a lot of trouble to translate from one to the other. (see above)
This happens to save 4 bytes of physical memory for each page in the
system.  (8 bytes on the Alpha).

Eliminate the use of the phys_avail[] array to determine if a page is
managed (ie: it has pv_entries etc).  Store this information in a flag.
Things like device_pager set it because they create vm_page_t's on the
fly that do not have pv_entries.  This makes it easier to "unmanage" a
page of physical memory (this will be taken advantage of in subsequent
commits).

Add a function to add a new page to the freelist.  This could be used
for reclaiming the previously wasted pages left over from preloaded
loader(8) files.

Reviewed by:	dillon
2000-05-21 12:50:18 +00:00
Poul-Henning Kamp
9626b608de Separate the struct bio related stuff out of <sys/buf.h> into
<sys/bio.h>.

<sys/bio.h> is now a prerequisite for <sys/buf.h> but it shall
not be made a nested include according to bdes teachings on the
subject of nested includes.

Diskdrivers and similar stuff below specfs::strategy() should no
longer need to include <sys/buf.> unless they need caching of data.

Still a few bogus uses of struct buf to track down.

Repocopy by:    peter
2000-05-05 09:59:14 +00:00
Søren Schmidt
af5bd99e9a Fix for the HP burners (and possibly other broken drives to)
that fails to proberly close the disk.
The problem seems to be that the HP burners sometimes return
ready when they actually are not, the solution is to not use
immediate mode on the closing commands. This is suboptimal
for real burners, in that they now hog the ATA bus for possibly
minutes, where its really not nessesary, *sigh*.
2000-05-01 07:30:28 +00:00
Poul-Henning Kamp
67f3c95cf9 Clone the {b|bio}_offset field, and make sure it is always initialized
in struct bio.  Eventually, bio_offset will probably obsolete the
bio_blkno and bio_pblkno fields.

Remove the special hack in atapi-cd.c to determine of bio_offset was valid.
2000-04-25 10:51:18 +00:00
Poul-Henning Kamp
ed6aff7387 Remove unneeded <sys/buf.h> includes.
Due to some interesting cpp tricks in lockmgr, the LINT kernel shrinks
by 924 bytes.
2000-04-18 15:15:39 +00:00
Poul-Henning Kamp
8177437d85 Complete the bio/buf divorce for all code below devfs::strategy
Exceptions:
        Vinum untouched.  This means that it cannot be compiled.
        Greg Lehey is on the case.

        CCD not converted yet, casts to struct buf (still safe)

        atapi-cd casts to struct buf to examine B_PHYS
2000-04-15 05:54:02 +00:00
Søren Schmidt
4175758a2d Add support for ATA66 on newer revisions of the VIA 82C596 2000-04-09 15:08:27 +00:00
Poul-Henning Kamp
c244d2de43 Move B_ERROR flag to b_ioflags and call it BIO_ERROR.
(Much of this done by script)

Move B_ORDERED flag to b_ioflags and call it BIO_ORDERED.

Move b_pblkno and b_iodone_chain to struct bio while we transition, they
will be obsoleted once bio structs chain/stack.

Add bio_queue field for struct bio aware disksort.

Address a lot of stylistic issues brought up by bde.
2000-04-02 15:24:56 +00:00
Søren Schmidt
eae50c69b3 Correct a mishap in phk's BIO_* changes. 2000-03-26 12:01:27 +00:00
Søren Schmidt
8c918526a5 Fix the panic occuring on non-changer devices. 2000-03-24 07:44:20 +00:00
Søren Schmidt
c481f1fa74 Make ATAPI CD changer devices work, hopefully better than before. 2000-03-21 13:26:54 +00:00
Poul-Henning Kamp
21144e3bf1 Remove B_READ, B_WRITE and B_FREEBUF and replace them with a new
field in struct buf: b_iocmd.  The b_iocmd is enforced to have
exactly one bit set.

B_WRITE was bogusly defined as zero giving rise to obvious coding
mistakes.

Also eliminate the redundant struct buf flag B_CALL, it can just
as efficiently be done by comparing b_iodone to NULL.

Should you get a panic or drop into the debugger, complaining about
"b_iocmd", don't continue.  It is likely to write on your disk
where it should have been reading.

This change is a step in the direction towards a stackable BIO capability.

A lot of this patch were machine generated (Thanks to style(9) compliance!)

Vinum users:  Greg has not had time to test this yet, be careful.
2000-03-20 10:44:49 +00:00
Søren Schmidt
2a14fd50d9 Remove the old DSC code, it doesn't work anymore. 2000-03-18 22:42:45 +00:00
Søren Schmidt
0d611a6c29 Cleanup sweep. 2000-03-13 12:27:11 +00:00
Søren Schmidt
cfb21bbb7d Enable disksort the right way both on ATA disks and ATAPI devices. 2000-03-13 12:04:58 +00:00
Søren Schmidt
8048917432 Backout the previous fix, it is _not_ that easy too fix.
Leave it as is for 4.0 the real solution comes later.
2000-03-13 09:10:25 +00:00
Søren Schmidt
4c435787ab Dont call ad_start unless the controller is idle.
This effectively nullified the usefullness of disksort().
2000-03-12 13:56:05 +00:00
Søren Schmidt
924f6734a5 Cosmetics.. 2000-03-10 22:10:34 +00:00
Søren Schmidt
643b70ec9d Dont call make_dev until we are sure the device is there, this prevents
panic'ing when the device is tried accessed later.
2000-03-10 08:09:53 +00:00
Søren Schmidt
aaa29cf209 Fix the CD driver so that the last blocks can be read even if
a blocksize != 2k is used.
Update the timeout code to try fallback to PIO if problems
arise in DMA mode.
2000-03-05 16:52:26 +00:00
Søren Schmidt
b744c0fdc6 Spell construct right. 2000-03-01 14:39:51 +00:00
Søren Schmidt
7caf73fb3c Add DVD-RAM support.
The driver constructs a fake disklabel that makes the 'a' partition
cover the entire DVD-RAM disk. This cannot be changed from the user
side. This solution was chosen because most DVD-RAM will have a
UDF (or until we have that CD9660) filesystem on it covering the
entire disk, its not really thought as a real random access device.

This might change over time, but for now this is what we have, and
it is compatible with CDROM's etc, that makes using the minidisk
subsystem less than ideal, because of !modulo BDEV_SIZE blocks.
2000-02-29 22:00:53 +00:00
Søren Schmidt
41ef40c93d Only read user data when blocksize != 2048, this "read the works"
sneaked in sometime which is not what the average user wants.
2000-02-29 16:11:18 +00:00
Søren Schmidt
cb66e8d4e3 Fix the problem that caused the boot to fail when modules were loaded.
Real braino, confuses two different softc types, I wonder how this
could ever work :(
2000-02-27 12:41:07 +00:00
Søren Schmidt
1a488af60f Add (partial) support for the Cypress ATA controller.
Call intr_teardown on detach.

Always add non masterdevice from unit 2 upwards.

Update to the pccard code, at least some cards are now working,
more testing to follow.
2000-02-25 09:48:23 +00:00
Søren Schmidt
9bd1c89ad3 Add UDMA66 support to the Intel ICH chip.
Submitted by: sakichan@lares.dti.ne.jp
2000-02-22 20:37:01 +00:00
Søren Schmidt
456b5ad253 Fix the changer code, dont wait for device ready on unload, it
wont become ready as there is no media - doh.

Update the command name reporting.
2000-02-21 18:41:05 +00:00