Commit Graph

87 Commits

Author SHA1 Message Date
John Baldwin
e240255ffc Add a bus_null_rescan() method that always fails with an error.
Use this in place of kobj_error_method to disable BUS_RESCAN() on
PCI drivers that do not use the "standard" scanning algorithm.
2016-04-27 17:49:42 +00:00
John Baldwin
3d0338a092 Implement a PCI bus rescan method.
Rescanning a PCI bus uses the following steps:
- Fetch the current set of child devices and save it in the 'devlist'
  array.
- Allocate a parallel array 'unchanged' initalized with NULL pointers.
- Scan the bus checking each slot (and each function on slots with a
  multifunction device).
- If a valid function is found, look for a matching device in the 'devlist'
  array.  If a device is found, save the pointer in the 'unchanged' array.
  If a device is not found, add a new device.
- After the scan has finished, walk the 'devlist' array deleting any
  devices that do not have a matching pointer in the 'unchanged' array.
- Finally, fetch an updated set of child devices and explicitly attach any
  devices that are not present in the 'unchanged' array.

This builds on the previous changes to move subclass data management into
pci_alloc_devinfo(), pci_child_added(), and bus_child_deleted().

Subclasses of the PCI bus use custom rescan logic explicitly override the
rescan method to disable rescans.

Differential Revision:	https://reviews.freebsd.org/D6018
2016-04-27 16:31:12 +00:00
John Baldwin
6cd99ae86d Add a new PCI bus interface method to alloc the ivars (dinfo) for a device.
The ACPI and OFW PCI bus drivers as well as CardBus override this to
allocate the larger ivars to hold additional info beyond the stock PCI ivars.

This removes the need to pass the size to functions like pci_add_iov_child()
and pci_read_device() simplifying IOV and bus rescanning implementations.

As a result of this and earlier changes, the ACPI PCI bus driver no longer
needs its own device_attach and pci_create_iov_child methods but can use
the methods in the stock PCI bus driver instead.

Differential Revision:	https://reviews.freebsd.org/D5891
2016-04-15 03:42:12 +00:00
John Baldwin
496dfa89a6 Convert pci_delete_child() to a bus_child_deleted() method.
Instead of providing a wrapper around device_delete_child() that the PCI
bus and child bus drivers must call explicitly, move the bulk of the logic
from pci_delete_child() into a bus_child_deleted() method
(pci_child_deleted()).  This allows PCI devices to be safely deleted via
device_delete_child().
- Add a bus_child_deleted method to the ACPI PCI bus which clears the
  device_t associated with the corresponding ACPI handle in addition to
  the normal PCI bus cleanup.
- Change cardbus_detach_card to call device_delete_children() and move
  CardBus-specific delete logic into a new cardbus_child_deleted() method.
- Use device_delete_child() instead of pci_delete_child() in the SRIOV code.
- Add a bus_child_deleted method to the OpenFirmware PCI bus drivers which
  frees the OpenFirmware device info for each PCI device.

Reviewed by:	imp
Tested on:	amd64 (CardBus and PCI-e hotplug)
Differential Revision:	https://reviews.freebsd.org/D5831
2016-04-06 04:10:22 +00:00
Warner Losh
8e5f761422 On my Lenovo T400, a Atheros 2413 has a problem powering up
sometimes. It will power up wrong and identify itself badly:

cardbus0: <network, ethernet> at device 0.0 (no driver attached)
cardbus0: <simple comms, UART> at device 0.1 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.2 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.3 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.4 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.5 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.6 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.7 (no driver attached)

All the higher numbered functions (.2 and above) have a config space
of all 0's. This smells a bit like a special debug mode, but the
current atheros driver doesn't cope. It is unclear if this card is
just a flake, or if we're doing something wrong in the power-up
sequence.

Put a work around into the code that tests for this rather unusual
condition. If we power a CardBus device up, and the device says it is
multi-function, and any of the functions have a 0 device ID, try the
power-up sequence again.
2015-02-18 05:53:04 +00:00
Warner Losh
8bf007e1e6 Add a rather obnoxious warning if you don't have NEW_PCIB defined
since it's a total crap shoot if things will work.
2015-01-14 05:41:31 +00:00
Hans Petter Selasky
af3b2549c4 Pull in r267961 and r267973 again. Fix for issues reported will follow. 2014-06-28 03:56:17 +00:00
Glen Barber
37a107a407 Revert r267961, r267973:
These changes prevent sysctl(8) from returning proper output,
such as:

 1) no output from sysctl(8)
 2) erroneously returning ENOMEM with tools like truss(1)
    or uname(1)
 truss: can not get etype: Cannot allocate memory
2014-06-27 22:05:21 +00:00
Hans Petter Selasky
3da1cf1e88 Extend the meaning of the CTLFLAG_TUN flag to automatically check if
there is an environment variable which shall initialize the SYSCTL
during early boot. This works for all SYSCTL types both statically and
dynamically created ones, except for the SYSCTL NODE type and SYSCTLs
which belong to VNETs. A new flag, CTLFLAG_NOFETCH, has been added to
be used in the case a tunable sysctl has a custom initialisation
function allowing the sysctl to still be marked as a tunable. The
kernel SYSCTL API is mostly the same, with a few exceptions for some
special operations like iterating childrens of a static/extern SYSCTL
node. This operation should probably be made into a factored out
common macro, hence some device drivers use this. The reason for
changing the SYSCTL API was the need for a SYSCTL parent OID pointer
and not only the SYSCTL parent OID list pointer in order to quickly
generate the sysctl path. The motivation behind this patch is to avoid
parameter loading cludges inside the OFED driver subsystem. Instead of
adding special code to the OFED driver subsystem to post-load tunables
into dynamically created sysctls, we generalize this in the kernel.

Other changes:
- Corrected a possibly incorrect sysctl name from "hw.cbb.intr_mask"
to "hw.pcic.intr_mask".
- Removed redundant TUNABLE statements throughout the kernel.
- Some minor code rewrites in connection to removing not needed
TUNABLE statements.
- Added a missing SYSCTL_DECL().
- Wrapped two very long lines.
- Avoid malloc()/free() inside sysctl string handling, in case it is
called to initialize a sysctl from a tunable, hence malloc()/free() is
not ready when sysctls from the sysctl dataset are registered.
- Bumped FreeBSD version to indicate SYSCTL API change.

MFC after:	2 weeks
Sponsored by:	Mellanox Technologies
2014-06-27 16:33:43 +00:00
John Baldwin
4edef187b8 Add support for managing PCI bus numbers. As with BARs and PCI-PCI bridge
I/O windows, the default is to preserve the firmware-assigned resources.
PCI bus numbers are only managed if NEW_PCIB is enabled and the architecture
defines a PCI_RES_BUS resource type.
- Add a helper API to create top-level PCI bus resource managers for each
  PCI domain/segment.  Host-PCI bridge drivers use this API to allocate
  bus numbers from their associated domain.
- Change the PCI bus and CardBus drivers to allocate a bus resource for
  their bus number from the parent PCI bridge device.
- Change the PCI-PCI and PCI-CardBus bridge drivers to allocate the
  full range of bus numbers from secbus to subbus from their parent bridge.
  The drivers also always program their primary bus register.  The bridge
  drivers also support growing their bus range by extending the bus resource
  and updating subbus to match the larger range.
- Add support for managing PCI bus resources to the Host-PCI bridge drivers
  used for amd64 and i386 (acpi_pcib, mptable_pcib, legacy_pcib, and qpi_pcib).
- Define a PCI_RES_BUS resource type for amd64 and i386.

Reviewed by:	imp
MFC after:	1 month
2014-02-12 04:30:37 +00:00
John Baldwin
1b1596a3b3 - Add a bus_dma tag to each PCI bus that is a child of a Host-PCI bridge.
The tag enforces a single restriction that all DMA transactions must not
  cross a 4GB boundary.  Note that while this restriction technically only
  applies to PCI-express, this change applies it to all PCI devices as it
  is simpler to implement that way and errs on the side of caution.
- Add a softc structure for PCI bus devices to hold the bus_dma tag and
  a new pci_attach_common() routine that performs actions common to the
  attach phase of all PCI bus drivers.  Right now this only consists of
  a bootverbose printf and the allocate of a bus_dma tag if necessary.
- Adjust all PCI bus drivers to allocate a PCI bus softc and to call
  pci_attach_common() from their attach routines.

MFC after:	2 weeks
2012-03-02 20:38:04 +00:00
Ed Schouten
6472ac3d8a Mark all SYSCTL_NODEs static that have no corresponding SYSCTL_DECLs.
The SYSCTL_NODE macro defines a list that stores all child-elements of
that node. If there's no SYSCTL_DECL macro anywhere else, there's no
reason why it shouldn't be static.
2011-11-07 15:43:11 +00:00
John Baldwin
8d43fea45f Don't explicitly list pci_write_ivar() for bus_write_ivar, the method is
already inherited from the PCI bus driver.
2011-05-02 15:01:28 +00:00
John Baldwin
af827f9642 Move the PCI-specific logic of removing a cardbus device into a
pci_delete_child() function called by the cardbus driver.  The new function
uses resource_list_unreserve() to release the BARs decoded by the device
being removed.

Reviewed by:	imp
Tested by:	brooks
2010-01-05 20:42:25 +00:00
Warner Losh
afb1ec6d56 Move the deactivation of the device's BAR to before the loop where we
turn deactivate the resources.  While this likely doesn't matter, it
is likely to be safer.
2009-03-12 08:42:27 +00:00
Warner Losh
41f0c91517 Restore blank line removed when fixing my earlier botch. Never do
just one last change before bed...

Pointy had to: imp
2009-03-10 22:04:52 +00:00
George V. Neville-Neil
7b77311219 Complete removal of cardbus_write_ivar which was left hanging. 2009-03-10 17:54:04 +00:00
Warner Losh
161832a1cf When freeing all the resources of the card, it is better to turn off
the PORTEN and MEMEN bits in the command register than to zero the
bars.

Use pci_write_ivar directly instead of a one-line wrapper that adds no
value.

Track verbosity changes in pci.

Remove a stray blank line.
2009-03-10 12:10:50 +00:00
John Baldwin
ba8d936dd2 Goof, catch up to constant rename (I renamed it to match the overall PCI
style of having register offsets start with PCIR_* rather than PCI_*).

Submitted by:	rss
2009-02-02 22:06:20 +00:00
Warner Losh
a84736bdf6 When no driver attaches to a card, don't power down the card. We can
now read config registers of cardbus cards that are inserted, but
aren't attached to a driver.
Also, add a power related comment...
2008-12-31 07:41:42 +00:00
Warner Losh
a7de0b74a7 Overhaul of CIS parsing, next step: keep a cached copy of the CIS,
read before we configure the card, so we can implement
/dev/cardbus*.cis.  Also, do this on a per-child basis, so we now have
a different name than before.  I think i'll have to fix that for some
legacy tools to keep working.

I can now do a dumpcis on my running atheros card and have it still work!
2008-11-17 01:32:29 +00:00
Marius Strobl
55aaf894e8 Make the PCI code aware of PCI domains (aka PCI segments) so we can
support machines having multiple independently numbered PCI domains
and don't support reenumeration without ambiguity amongst the
devices as seen by the OS and represented by PCI location strings.
This includes introducing a function pci_find_dbsf(9) which works
like pci_find_bsf(9) but additionally takes a domain number argument
and limiting pci_find_bsf(9) to only search devices in domain 0 (the
only domain in single-domain systems). Bge(4) and ofw_pcibus(4) are
changed to use pci_find_dbsf(9) instead of pci_find_bsf(9) in order
to no longer report false positives when searching for siblings and
dupe devices in the same domain respectively.
Along with this change the sole host-PCI bridge driver converted to
actually make use of PCI domain support is uninorth(4), the others
continue to use domain 0 only for now and need to be converted as
appropriate later on.
Note that this means that the format of the location strings as used
by pciconf(8) has been changed and that consumers of <sys/pciio.h>
potentially need to be recompiled.

Suggested by:	jhb
Reviewed by:	grehan, jhb, marcel
Approved by:	re (kensmith), jhb (PCI maintainer hat)
2007-09-30 11:05:18 +00:00
John Baldwin
d193b870a5 Fix interrupt routing for some cards after the previous fix to clear BARs
in cardbus cards: update the saved copy of the PCIR_INTLINE register in
the PCI ivars in addition to the actual register.

Reviewed by:	imp
2007-05-16 23:40:08 +00:00
Warner Losh
cfd1374177 Write 0 to all the bars in cardbus_cardbus_setup_regs. This is needed
because on at least my dc based cards there's garbage in there.  The
recent changes in the resource code appears to have unmasked this
problem...  At least dc now probes/attaches better than it did before.

Also, we no longer need to write to the cfg for the other registers.
2007-05-16 20:00:36 +00:00
Warner Losh
7e2ff8bbff Cope gracefully with device_get_children returning an error.
Obtained from: Hans Petter Selasky
P4: http://perforce.freebsd.org/chv.cgi?CH=112957
2007-01-19 08:49:28 +00:00
Warner Losh
a63eba960f When we can't parse the CIS, note with a warning that the bogus CIS
was ignored, rather than freaking out.  In the past, it wasn't possible
to not parse the CIS, so this changes no behavior.
2006-06-12 03:17:24 +00:00
John Baldwin
5aa58b3e8f Make the 'pci_devclass' pointer variable private (drivers really shouldn't
share devclass pointers, a mistake I've encouraged in the past) and
move the declaration of the pci_driver kobj class from cardbus.c to
pci_private.h so that other drivers can inherit from pci_driver.
2006-01-20 22:00:50 +00:00
Warner Losh
388395ccaa Remove now-obsolete printf warning. 2005-12-30 19:38:47 +00:00
Warner Losh
ef5cc9accd Move all of the resource allocation into the pci layer. The resource
allocation here just duplicated it (badly).
2005-12-30 19:25:04 +00:00
Gleb Smirnoff
6afb2250fb - Retire BARBIT in favor of new PCI_RID2BAR.
- Fix build.
2005-12-30 11:31:47 +00:00
Gleb Smirnoff
408b85a1e2 Help Warner with merge from p4. 2005-12-29 10:38:42 +00:00
Warner Losh
47147ce799 Implement /dev/cardbus%d.cis, same thing as /dev/pccard%d.cis. There
are some rough edges with this still, but it seems to work well enough
to commit.
2005-12-29 01:43:47 +00:00
Gleb Smirnoff
086745614c When in rev. 1.47 cardbus_alloc_resources() function was moved from
cardbus_cis.c to this file, some code was not merged and thus resource
list entries were invalid. They didn't have a resources attached to
them.
  However, the problem was masked for some time later, because newer
resources list entries were added to the head of the list, and
resource_list_find() always returned the first matching resource list
entry. Usually the underlying driver allocated a valid resource and
added it to the head of the list, and invalid one wasn't used.
  In rev. 1.174 of subr_bus.c the sorting of resource list entries was
reversed demasking the problem in cardbus_alloc_resources().
  This commit fixes the problem returning back some code from
cardbus_cis.c, pre-1.49 revisions.

PR:			kern/87114
PR:			kern/90441
Hardware provided by:	Vasily Olekhov <olekhov yandex.ru>
Reviewed by:		imp
2005-12-28 10:15:01 +00:00
Warner Losh
6811d2bc72 Cardbus has only 1 slot, so simplify a little. 2005-10-28 06:03:53 +00:00
Ruslan Ermilov
360d00a0b2 Calling rman_get_start() after bus_release_resource() is evil.
It became fatal after a recent "struct resource" split change.
2005-09-27 13:33:46 +00:00
Warner Losh
36fed96550 Use STAILQ in preference to SLIST for the resources. Insert new resources
last in the list rather than first.

This makes the resouces print in the 4.x order rather than the 5.x order
(eg fdc0 at 0x3f0-0x3f5,0x3f7 is 4.x, but 0x3f7,0x3f0-0x3f5 is 5.x).  This
also means that the pci code will once again print the resources in BAR
ascending order.
2005-03-18 05:19:50 +00:00
Warner Losh
2219bbc553 Doh! silly typo precludes compiling 2005-03-11 05:30:59 +00:00
Warner Losh
0f675788ff Revert changes of 1.49. Lots-a-people broke with it, for reasons
unknown (since my sony vaio didn't :-(.

Instead, fix the problem described by 1.49 in a different way: just
add the two calls I'd hoped I'd avoid in 1.49 by doing the (wrong)
gymnastics there.  While 1.49 is a good direction to go in, each step
of the way should work :-(.
2005-03-11 05:27:05 +00:00
Warner Losh
af5e97c122 There were two calls to cardbus_do_cis when cardbus_do_cis changed,
yet I only changed one of them.  So when we loaded drivers, we'd fail
to allocate resources correct.

This pointed out that we were doing the wrong thing when we failed to
attach a child.  We released all the resources and almost deleted the
child.  Instead, we should keep the resources allocated so when/if a
driver is loaded, we can go w/o having to allocate them.  We use
pci_cfg_save/restore to restore the BARs with these resources.

This seems to fix the problems that we were seeing that I thought
might have magically gone away in the last revision of cardbus.c (but
really didn't).

Noticed by: avatar (nicely done!)
2005-02-28 01:27:24 +00:00
Warner Losh
164b196506 style(9) nit 2005-02-20 20:32:44 +00:00
Warner Losh
440b5ade31 Move resource allocation routines from cardbus_cis.c to cardbus.c.
They have nothing at all to do with CIS parsing.

Remove some unused funce parsing: nothing used the results.

Use more of pccard_cis.h's deifnitions for the cardbus specific cis
parsing we do.  More work is needed in this area.

This reduces the size of the cardbus module by 380 bytes or so...
2005-02-06 21:03:13 +00:00
Warner Losh
2dd5c91ebb Use the standard FreeBSD license
Approved by: imp, jon
2005-01-13 19:12:10 +00:00
Warner Losh
098ca2bda9 Start each of the license/copyright comments with /*-, minor shuffle of lines 2005-01-06 01:43:34 +00:00
Warner Losh
5d11e83faa MFp4:
The hack for setting the bus has been moved down into the cbb driver.
I've been running without this hack in my tree for so long I had
forgotten that I'd removed it :-).  Please let me know if this causes
difficulty for your laptop.
2004-06-27 13:07:02 +00:00
Poul-Henning Kamp
41ee9f1c69 Add some missing <sys/module.h> includes which are masked by the
one on death-row in <sys/kernel.h>
2004-05-30 17:57:46 +00:00
Doug Rabson
aec21b56e8 Make the cardbus driver a derived class of the pci driver. In theory, this
should allow many of the pci methods to be re-staticised.
2003-11-01 12:45:03 +00:00
Warner Losh
f6e01094a4 remove the cardbus cis reading code. nobody ever used it and it has
locking issues down to the api level.
2003-10-06 15:56:29 +00:00
David E. O'Brien
aad970f1fe Use __FBSDID().
Also some minor style cleanups.
2003-08-24 17:55:58 +00:00
Thomas Moestl
3920999db7 Add a new PCI interface method, assign_interrupt, to determine the
interrupt to be used for a device. This is intended solely for internal
use of PCI bus implementations, and exists so that PCI bus drivers
implementing special interrupt assignment methods which require
additional work at the bus level to work right can be easily derived
from the generic driver (or any other one) without resorting to hacks.

It will be used in the sparc64 ofw_pcibus driver, which will be
committed shortly.

Make use of this method in the generic implementation, and add it to
the method table of bus drivers derived from the PCI one.

Reviewed by:	imp, -hackers
2003-07-01 14:08:33 +00:00
Warner Losh
51715fe79d Move the resource handling down into the pci bus as well.
Minor CIS resource allocation code cleanup
Remove some fairly useless debug writes.

This finishes the work to move as much cardbus code as possible into
pci.  We wind up removing 800-odd lines from cardbus.c: we go from
1285 to 400 lines.

Reviewed by: mdodd
2003-02-18 21:24:00 +00:00