way it attaches multiple PCI buses directly to the CPU, instead of having
them hanging off from PCI to PCI bridges. This code is a hack, and will
be obsoleted by the planned rework of the PCI code, which will change the
dealing with PCI to PCI bridges and other special devices significantly.
The patch also adds a kern_devconf entry for PCI bus 0 which is assumed
to be a child of cpu0. The new PCI code will make it possible to hand out
the kern_devconf structure to a pci device being attached, since this is
(regretably, IMHO) required by a few ISA devices.
Finally there are new PCI ids for some Intel chip set devices, which had
already been known to 2.1.5R, but did not make it into -current. This closes
"kern/1558: PCI probe seems to have lost a device in -current".
logic clock signal, which had been erroneously commented out by the
previous commit. This will re-enable support for sync. transfer negotiation,
which depends on one of those values.
calculate an optimum value from (constant) parameters.
This should set the SCNTL3 register of the 53c860 and 53c875 to twice
the divider it used to be, since cards based on those chips seem to use
an 80MHz clock instead of the Clock Doubler feature and a 40MHz clock.
This code applies to several systems with integrated Ethernet
chip, for example from HP or Compaq. It should also support
PCI Ethernet cards based on the AMD PCI Lance chip.
This code has been reviewed (visually) by Paul Richards and
tested (using an ISA Lance board) by Joerg Wunsch.
Since the parameters to nearly each and every single function
had to be changed (generally from unit number to lnc_soft*),
there is some potential for buglets having crept in ...
BEWARE: If you had lnc0 configured to have the ISA probe find
your PCI Lance, then it should now be found by the PCI probe,
and should be automatically configured as pci1 (!!! note the "1").
Reviewed by: paul, joerg
is only used by the icu support modules and by a few drivers that know
too much about the icu (most only use it to convert `n' to `IRQn'). isa.h
is only used by ioconf.c and by a few drivers that know too much about
isa addresses (a few have to, because config is deficient).
All new code is "#ifdef PC98"ed so this should make no difference to
PC/AT (and its clones) users.
Ok'd by: core
Submitted by: FreeBSD(98) development team
NetBSD/OpenBSD support Submitted by:Noriyuki Soda <soda@sra.co.jp>,
Pete Bentley <pete@demon.net>,
Charles M. Hannum <mycroft@mit.edu>,
Theo de Raadt <deraadt@theos.com>
I spent the better part of a day trying to figure out why my
experiment didn't work the way I expected, only to find out that
the router was dropping huge numbers of packets because of PCI bus
priblems. This does not fix the bug that errors are counted as
input packets because my patch doesn't apply cleanly.
is enabled by having an "device ed0 at isa? [...]" config line.
The first PCI card will get a unit number one higher than the highest
defined for any ISA card of the ED type, e.g. if ed0 and ed1 are
configured, then the PCI cards will be ed2, ed3, ...
BEWARE: If you have configured your kernel as ed0 with the port address
as assigned by the PCI BIOS, then your card will be found by both the
PCI and ISA probes, and bad things may happen. Make sure to restore
the original port address form the GENERIC kernel for the ed0 device!
Reviewed by: davidg
1) A spelling error pointed out by Paco Hope.
2) A bug in the range checking routing pointed out by Jim Bray.
3) Enables the setting of frames per second.
Submitted-By: Jim Lowe <james@miller.cs.uwm.edu>
a BIOS was not installed, this will still be true by the time we probe
the chip. We use this heuristic to determine if we should use the left
over scratch ram target settings for controllers that don't have an
SEEPROM. We also "snapshot" the host adapter SCSI id and whether ultra
is enabled or not and use these values if a BIOS was installed. The card
will act as if a BIOS was installed even if there wasn't one if you warm
reboot, but since the scratch ram area is still valid in this case, its
hardly worth the effort of writing a shutdown routing that clears out
the scratch ram. This should make users of motherboard controllers
happy.
should be <= than subordinate, not the other way around.
They are both true if the bridge is not cascaded (i.e., twin-channel
scsi/e-net adapters won't be affected by this bug), which is probably why
it was unnoticed until today.
- always use pci_conf_read() and pci_conf_write(). (This is required to
simulate non-existant devices in my system for PCI bridge code tests.)
- reorder some functions (put the main functions at the end).
- correct off by one bug in the code dealing with unitialized PCI to PCI
bridge chips. (Bug found by ASAMI Satoshi.)
- print function number for multi-function devices.
Use new XS_SELTIMEOUT error code for selection timeouts.
aic7870.c:
Move SCB walking code to aic7xxx.c and make it work for all card types.
The flag AHC_EXTSCB is no longer needed since the SCBs are walked in
all cases now.
vm_offset_t is currently unsigned long but should probably be plain
unsigned for i386's to match the choice of minimal types to represent
for fixed-width types in Lite2. Anyway, it shouldn't be assumed
to be unsigned long.
I only fixed the type mismatches that were detected when I changed
vm_offset_t to unsigned. Only pointer type mismatches were detected.
- Do not enable tagged commands by default
- Probe only 1 LUN
- Do not negotiate sync. transfer with CDROM drives
Defining FAILSAFE will result in a driver that will tolerate
marginal hardware for the price of a slight loss of performance
It is intended for use in install kernels.
Cleanse the SCSI subsystem of its internally defined types
u_int32, u_int16, u_int8, int32, int16, int8.
Use the system defined *_t types instead.
aic7870.c:
Handle Seeprom data a little better.
the S-Video input. It also has code in the driver for the meteor RGB support
and some other bug fixes. I don't have a meteor RGB but I have been told
that it works.
Submitted by: Jim Lowe <james@miller.cs.uwm.edu>
port addresses (even though the PC architecture doesn't support them).
Add code to limit the I/O map size based on the lowest set bit of the
address. This cures the problem with the BT946C only having a 16 bit
map register, in voiolation of the PCI specs, without giving up the
general support of >65K port regions.
- fill in and use ifp->if_softc
- use if_bpf rather than private cookie variables
- change bpf interface to take advantage of this
- call ether_ifattach() directly from Ethernet drivers
- delete kludge in if_attach() that did this indirectly
that don't announce support for command queues.
SCSI_NCR_DFLT_TAGS can be specified in the kernel config file
and sets the default number of tags per disk drive.
A value of 0 means "no tags".
Minor correction in debug messages: Values from the msg_in
buffer were being printed in the msg_out trace message ...
feature in the header type register, though it is required by the PCI spec.
This should correctly probe both functions of the Intel 82371FB chip,
without the need for a special case based on the device ID.
chipset. This does not attempt to do anything special with the timing
on the hope that the BIOS will have done the right thing already. The
actual interface from the wd driver to the new facility is not
implemented yet (this commit being an attempt at prodding someone else
to do it because looking at the wd driver always confuses the h*** out of me).
includes a hack in the probe code: the 82371FB is a multifuction
device, but doesn't properly set the configuration bit which
indicates this. So, we just hard-wire all 82371FBs as multifunction
devices.
This does not actually make the bus-master IDE stuff work, although
if anyone wants to work on that, I have the databooks that tell
how to use it.
(To be replaced by sysctl accesses some time ...)
Remove a backward jump from the NCR script, which allowed a SCSI target
to receive any number of NOP messages it desired. If a target indeed
does such a silly thing, make it fail at the next instruction, instead
of causing a timeout a few seconds later.
Simplify the initialization of adapters by pulling all card specific
initialization to the card specific modules.
Set the Latency timer and Burst len to good values if thery are not
initialized during post or are reset during chip reset.
Properly identify and handle external SCB SRAM. The code was false
id'ing 255 SCBs on aic7880 chips.
Reviewed by: David Greenman <davidg@FreeBSd.org>
#includes to get prototypes.
pci now uses a different interrupt handler type for interrupts that it
dispatches and the isa interrupt handler type for the interrupts that
it handles.
Reduce default value of pcicb_membase to 0x2000000 (from 0x4000000)
since this seems to be the lower bound used by many systems.
Submitted by: Mihoko Tanaka <m_tanaka@pa.yokogawa.co.jp>
all the other bt_XXX() functions in i386/scsi/bt*.
This the important effect of forcing a link error if the user is
still using the old "vector btintr" which is dangerously wrong
after Justin's updates to the driver.
The correct isa vector line for the bt driver is "vector bt_isa_intr".
Justin mentioned this in the commit message and updated LINT and
GENERIC. This change is to enforce that.. :-)
most devsw referenced functions are now static, as they are
in the same file as their devsw structure. I've also added DEVFS
support for nearly every device in the system, however
many of the devices have 'incorrect' names under DEVFS
because I couldn't quickly work out the correct naming conventions.
(but devfs won't be coming on line for a month or so anyhow so that doesn't
matter)
If you "OWN" a device which would normally have an entry in /dev
then search for the devfs_add_devsw() entries and munge to make them right..
check out similar devices to see what I might have done in them in you
can't see what's going on..
for a laugh compare conf.c conf.h defore and after... :)
I have not doen DEVFS entries for any DISKSLICE devices yet as that will be
a much more complicated job.. (pass 5 :)
pass 4 will be to make the devsw tables of type (cdevsw * )
rather than (cdevsw)
seems to work here..
complaints to the usual places.. :)
Removed ifnet.if_init and ifnet.if_reset as they are generally unused.
Change the parameter passed to if_watchdog to be a ifnet * rather than
a unit number. All of this is an attempt to move toward not needing an
array of softc pointers (which is usually static in size) to point to
the driver softc.
if_ed.c:
Changed some of the argument passing to some functions to make a little
more sense.
if_ep.c, if_vx.c:
Killed completely bogus use of if_timer. It was being set in such a way
that the interface was being reset once per second (blech!).
o Add signed/unsigned functionality to the matrox meteor device driver.
o Apply a few fixes to the sound driver.
o Add a ``SPIGOT_UNSECURE'' compile time definition so, if one defines
SPIGOT_UNSECURE in their conf file, then they can use the spigot w/o
root. There is a warning that this allows users access to the IO
page which is probably not secure.
Submitted by: james
Convert the remaining sysctl stuff to the new way of doing things.
the devconf stuff is the reason for the large number of files.
Cleaned up some compiler warnings while I were there.
incompatible with the type of a PCI interrupt handler.
Fixed the type of pdc_pci_ifintr(). The type of a PCI interrupt handler
is too generic to pass arbitrary struct pointers.
incompatible with the type of a PCI interrupt handler. A new entry
point `ahc_pci_intr()' is used for PCI. ISA and PCI interrupts are
penalized equally (:-) by calling a common handler `ahc_intr()'. This
should be reorganized. Some strings now name the wrong function...
Submitted by: fgray@rice.edu
this driver hasn't been checked but as a separate module, bringing it in won't
break anything else and it't the best way of testing it......
julian
If RAMENB is set in devconfig, walk the external SCBs. Some Intel Xpress
motherboards set this bit.
For external SCBs for the 3940. It doesn't set RAMPS or RAMENB, but does
have the ram.
"I screwed the initialization of the burstsize. Right now it will default
to 0 (which can cause corruption problems on high latency PCI buses). It
should be set to 8 longwords to avoid problems with certain PCI chipsets."
Submitted by: Matt Thomas <matt@lkg.dec.com>
for now since I don't have any documentation on this card yet and it
is software compatible with the earlier cards.
The 2940 *Ultra* is Adaptec's push to take SCSI to 20MHz bus rates. Its
based on the aic7880 chip which is the successor to the aic7870.
and for the 53c810ap, the improved version of the 53c810.
The driver should work with all those controllers, but doesn't know
about any of their advanced features.
of NCR script labels and of command control block structure components.
This allows for easy modification of the actual virtual to physical
mapping operator used ...
Make all applicable references to physical address values use the above
macros instead of calls to vtophys().
They shouldn't affect FreeBSD, since they are within #ifdef NETBSD
directives.
Changed the HAD_ERROR return code into COMPLETE according to a comment
in "/sys/scsi/scsiconf.h"
Submitted by: Andreas Wrede <andreas@planix.com>
This seems to work fine on my 53c810, but really should be tested on
a 53c825 with at least one target set to an ID >= 8.
The script is now copied to memory mapped using vm_page_alloc_contig(),
since it has to be physically contigous. This must be changed, if the
driver is converted into a loadable module !
Two of the probe messages are suppressed, unless "bootverbose" is set.
contiguous memory in virtual space, but doesn't guarantee any particular
physical layout of pages, the script may be not contguous in physical
memory. This made the first write command fail in very rare cases.
This has been fixed by checking the region returned by malloc() for
being contiguous, but there should be a physical memory allocate
and later mapping to virtual memory instead.
A few assertions and error messages are improved.
This change forces the controller drivers to allocate a scsibus_data struct
via a call to scsi_alloc_bus(), fill in the adapter_link field, and optionally
modify any other fields of the struct. Scsi_alloc_bus() initializes all fields
to the default, so the changes in most drivers are very minimal. For drivers
that support Wide controllers, the maxtarg field will have to be updated to
allow probing of all targets (for an example, look at the aic7xxx driver).
Scsi_attachdevs() now takes a scsibus_data* as its argument instead of an
sc_link*. This allows us to expand the role of the scsibus_data struct for
other bus level configuration setings (max number of transactions, current
transaction opennings, etc for better tagged queuing support).
Reviewed by: Rodney Grimes <rgrimes>, Peter Dufault <dufault>, Julian Elischer <julian>
ISA GAT mode and hidden refresh seem to cause reliability problems
on Saturn based systems and are now reported when booting with '-v'.
Submitted by: Danny J. Zerkel <dzerkel@feephi.phofarm.com>
series of hard disk drives, which don't accept any SCSI message
within an REQUEST SENSE command (i.e. even not an IDENTIFY to set
the LUN).
This patch obviates the need for QUIRK_NOMSG and thus all of the
device_tab[] entries in the NCR driver.
people tend to assume their devices won't work if they see this
message, though it may indicate that those devices just don't
need any PCI driver (e.g. devices that emulate an ISA card, or
that have been initialised by the BIOS and need no further care).
etc.). The tulip_start routine was rewritten to use less stack space (I've
been having problems with wcarchive overflowing the stack and this should
help a little). This version also has preliminary NetBSD support.
Rod Grimes helped in testing this version of the driver. Thanks Rod. It's
additionally been extensively tested here and on wcarchive.
Submitted by: Matt Thomas
to most users (the wrong length is passed to ether_input). The
second is more serious. The multicast hash algorithm uses the wrong
(low) bits instead of the right (high) bits. This is only an issue
if you use >12 multicast addresses but if you are using IP multicast
then it might affect you...
Submitted by: Matt Thomas
thrown out if bpfilter support and no BPF listener. (submitted by Bill
Fenner)
Removed unused variable and changed another from a stack variable to a
static - the variable was a rather large array of structs that consumed
a lot of stack space. (me)
fragmented.
Added support for Cogent em100 boards.
Fixed bug that caused BPF to toggle the card to UTP.
Various other improvements.
Submitted by: Matt Thomas and David Greenman
got a 2.2 version DC21040 chip in my SMC ethernet card! He suggests bumping
the check all the way down to 2.0 since it's pre-2.0 we're actually guarding
against.
Submitted by: Matt Thomas <matt@lkg.dec.com>
to poll succesfully even if we are sharing the interrupt.
Register the interrupt handler before the attach.
This commit makes the 294x PCI shared interrupt compliant. This has
been tested with an aic7870 motherboard controller and a 294x in the
same machine shareing an irq.
include <pci/pcivar.h> without including <sys/devconf.h> and other
drivers include <pci/pcivar.h> before including <sys/devconf.h> if
certain identifiers are defined.
The devconf headers have convoluted interdependencies. <sys/devconf.h>
includes <machine/devconf.h> which includes <pci/pcivar.h>. Most
drivers include <sys/devconf.h> so even isa drivers depend on
<pci/pcivar.h>. For similar reasons, most drivers depend on another
pci header, on an isa header and on two scsi headers.
1) Supports PCI to PCI bridge devices (and tries to initialise them,
even if the BIOS is brain dead).
2) Supports shared PCI interrupts. Interrupt handlers now MUST return
'0' if they found nothing to do, '1' otherwise.
New features tested with i486 systems based on the Intel Saturn and
a DEC 4channel Ethernet card only, but expected to work on most systems.
The option PCI_REMAP has been removed !
Submitted by: Wolfgang Stanglmeier <wolf@kintaro.cologne.de>
a device specific shutdown routine for devconf. Assign the value of this
to the kern_devconf struct. Implement a device shutdown routine for if_de
that disables the device. This will stop the device from corrupting memory
after a reboot.
be 486 chip sets that can't tolerate bursts > cache line size.
This should really made dependent on the particular buggy
chip sets, but for now we'll play safe ...
Try to deduce maximum number of PCI buses in system (working around
chip set bugs).
Better check for devices at multiple addresses (aliases).
Reviewed by: se
Submitted by: <wolf@kintaro.cologne.de> Wolfgang Stanglmeier
DEC 21050 chip in particular, don't have specs of other such chips).
This should add support for Multiple-Ethernet PCI cards (e.g. Znyx 314).
Reviewed by: se
Submitted by: <wolf@kintaro.cologne.de> Wolfgang Stanglmeier
PCI BIOS mappings are retained, except if option PCI_REMAP
is specified in the kernel config file.
There is now a list of attach addresses, and the first
address that seems to make some device registers appear
is chosen.
Reviewed by: se
Submitted by: wolf
Test for correct execution of cache test script by NCR,
and give meaningful error description if it fails.
(A cache problem was reported before.)
Don't wait forever for cache test to complete (to protect
against faulty hardware).
Submitted by: wolf
New config option "NCR_IOMAPPED" makes the driver use port I/O.
Put back in 53c815 defines, submitted by Mikael Hybsch <micke@dynas.se>.
These had got lost between cvs rev. 1.14 and now ...
pci.c:
Really write config space register.
Assign ports starting at 0xbc00.
Submitted by: wolf
Reviewed by: se
so that the interface won't have the effect of blocking other senders
during bulk transfers (i.e. hogging the ethernet). It improves performance
in all of my tests by reducing collisions and I believe it to be a Good
Thing.
definitions taken from the PCI specs. Part of them were typed
in by Wolfgang Stanglmeier, the (at that time unneeded) rest
by Charles Hannum (thanks !).
may be the result of reselect following too fast for
the driver to notice. Not the final solution, but the
problem has been seen only with very few devices.
Reviewed by: se
Submitted by: wolf (Wolfgang Stanglmeier)
worked in the past only because of good fortune. Anyway, use the contig alloc
routine I wrote awhile ago (vm_page_alloc_contig) for the sound code to do
this allocation. Also, specify read+write on the permissions to pmap_enter().
Specifying just read can have unexpected consquences.
This should have been disabled for some time, but I had screwed up ...
This made spurious values appear for fd0 in systat, when there was
NCR SCSI activity.
and into ether_input(). It was silly to have bpf want this one way and
ether_input want it another way. Ripped out trailer support from the few
remaining drivers that still had it.
1) make #includes correct
2) fix bugs in address check macros
3) fixed bugs in, and enabled, recopy if heavily fragmented code
4) moved call to bpf tap to be before enqueing packet (probably gratuitous)
5) fixed bug that caused "abnormal interrupt" at boot time/first use
6) added support for reading Zynx address ROM
7) fixed bug that caused broadcasts to not work shortly after booting (only
manifested if not using multicast - e.g. not in FreeBSD 2.0)
8) fixed spelling errors in comments
Submitted by: Matt Thomas
