]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
19 years agoTurn on the FreeBSD login user capabilities database support.
obrien [Wed, 18 Aug 2004 06:46:25 +0000 (06:46 +0000)]
Turn on the FreeBSD login user capabilities database support.

19 years agoNetBSD has updated their groff to a version that handles .Nm the same
obrien [Wed, 18 Aug 2004 06:41:13 +0000 (06:41 +0000)]
NetBSD has updated their groff to a version that handles .Nm the same
way ours does.  So use the vendor files now.

19 years agoMerge rev 1.2 (OPIE, login user capabilities database, PAM) support
obrien [Wed, 18 Aug 2004 06:34:20 +0000 (06:34 +0000)]
Merge rev 1.2 (OPIE, login user capabilities database, PAM) support
into 'nbsd_20040809'.

19 years agoThis commit was generated by cvs2svn to compensate for changes in r133936,
obrien [Wed, 18 Aug 2004 06:29:48 +0000 (06:29 +0000)]
This commit was generated by cvs2svn to compensate for changes in r133936,
which included commits to RCS files with non-trunk default branches.

19 years agoImport of LukeM's ftpd taken from the NetBSD CVS repo on 9-Aug-2004.
obrien [Wed, 18 Aug 2004 06:29:48 +0000 (06:29 +0000)]
Import of LukeM's ftpd taken from the NetBSD CVS repo on 9-Aug-2004.
This closes the remotely exploitable vulnerability documented at
ftp://ftp.netbsd.org/pub/NetBSD/security/advisories/NetBSD-SA2004-009.txt.asc
http://www.vuxml.org/freebsd/c4b025bb-f05d-11d8-9837-000c41e2cdad.html

19 years agoWhen one entry in the RSDT is corrupted, just skip it instead of bailing out.
njl [Wed, 18 Aug 2004 05:56:07 +0000 (05:56 +0000)]
When one entry in the RSDT is corrupted, just skip it instead of bailing out.
This gets us the info we need on systems which have proprietary tables that
don't match the standard.  For instance, an AMI system has a table of type
"OEMB" with an invalid checksum.

Tested by: Maxim Maximov <mcsi_at_mcsi.pp.ru>
MFC after: 1 day

19 years agoUpdate man page for supported table types.
njl [Wed, 18 Aug 2004 05:50:32 +0000 (05:50 +0000)]
Update man page for supported table types.

MFC after: 1 day

19 years agoRemove the ACPIIO_ENABLE and ACPIIO_DISABLE ioctls as well as all
njl [Wed, 18 Aug 2004 05:48:24 +0000 (05:48 +0000)]
Remove the ACPIIO_ENABLE and ACPIIO_DISABLE ioctls as well as all
callers.  These ioctls attempted to enable and disable the ACPI
interpreter at runtime.  In practice, it is not possible to boot with
ACPI and then disable it on many systems and trying to do so can cause
crashes, interrupt storms, etc.  Binary compatibility with userland is
retained.

MFC after: 2 days

19 years agoThis commit was generated by cvs2svn to compensate for changes in r133931,
njl [Wed, 18 Aug 2004 05:41:21 +0000 (05:41 +0000)]
This commit was generated by cvs2svn to compensate for changes in r133931,
which included commits to RCS files with non-trunk default branches.

19 years agoAdd a padding member to the header overlaid on data allocated in the
njl [Wed, 18 Aug 2004 05:41:21 +0000 (05:41 +0000)]
Add a padding member to the header overlaid on data allocated in the
ACPI_DEBUG case.  Without this, use of allocated memory is unaligned and
causes a trap on ia64.  Intel may fix this differently in a subsequent
release but this is adequate for now.

Submitted by: marcel
MFC after: 2 days

19 years agoAssert Giant in fwe_start(), as it is not yet MPSAFE.
rwatson [Wed, 18 Aug 2004 04:54:54 +0000 (04:54 +0000)]
Assert Giant in fwe_start(), as it is not yet MPSAFE.

19 years agoSince pc98 shares the AGP driver with the i386, also define the
rwatson [Wed, 18 Aug 2004 03:46:39 +0000 (03:46 +0000)]
Since pc98 shares the AGP driver with the i386, also define the
amd64 agp option here in order to let the pc98 kernel build
complete.  This doesn't seem right, since there probably aren't
plans to build a pc98 amd64 box; however, it's not clear to me
how to get config to generate an opt_agp.h without an option
defined.

19 years agoWe're not yet ready for BURN_BRIDGES to break the build, as
rwatson [Wed, 18 Aug 2004 03:13:16 +0000 (03:13 +0000)]
We're not yet ready for BURN_BRIDGES to break the build, as
BURN_BRIDGES won't be removed for a bit yet.  Fix more of the
build in HEAD.

19 years agoFix build of ip_input.c with "options IPSEC" -- the "pass:" label
rwatson [Wed, 18 Aug 2004 03:11:04 +0000 (03:11 +0000)]
Fix build of ip_input.c with "options IPSEC" -- the "pass:" label
is used with both FAST_IPSEC and IPSEC, but was defined for only
FAST_IPSEC.

19 years agoMake the kernel compile again if you are not using PFIL_HOOKS
peter [Wed, 18 Aug 2004 00:37:46 +0000 (00:37 +0000)]
Make the kernel compile again if you are not using PFIL_HOOKS

19 years agoHEAD is now 6-CURRENT
scottl [Wed, 18 Aug 2004 00:21:12 +0000 (00:21 +0000)]
HEAD is now 6-CURRENT

19 years agoConvert ipfw to use PFIL_HOOKS. This is change is transparent to userland
andre [Tue, 17 Aug 2004 22:05:54 +0000 (22:05 +0000)]
Convert ipfw to use PFIL_HOOKS.  This is change is transparent to userland
and preserves the ipfw ABI.  The ipfw core packet inspection and filtering
functions have not been changed, only how ipfw is invoked is different.

However there are many changes how ipfw is and its add-on's are handled:

 In general ipfw is now called through the PFIL_HOOKS and most associated
 magic, that was in ip_input() or ip_output() previously, is now done in
 ipfw_check_[in|out]() in the ipfw PFIL handler.

 IPDIVERT is entirely handled within the ipfw PFIL handlers.  A packet to
 be diverted is checked if it is fragmented, if yes, ip_reass() gets in for
 reassembly.  If not, or all fragments arrived and the packet is complete,
 divert_packet is called directly.  For 'tee' no reassembly attempt is made
 and a copy of the packet is sent to the divert socket unmodified.  The
 original packet continues its way through ip_input/output().

 ipfw 'forward' is done via m_tag's.  The ipfw PFIL handlers tag the packet
 with the new destination sockaddr_in.  A check if the new destination is a
 local IP address is made and the m_flags are set appropriately.  ip_input()
 and ip_output() have some more work to do here.  For ip_input() the m_flags
 are checked and a packet for us is directly sent to the 'ours' section for
 further processing.  Destination changes on the input path are only tagged
 and the 'srcrt' flag to ip_forward() is set to disable destination checks
 and ICMP replies at this stage.  The tag is going to be handled on output.
 ip_output() again checks for m_flags and the 'ours' tag.  If found, the
 packet will be dropped back to the IP netisr where it is going to be picked
 up by ip_input() again and the directly sent to the 'ours' section.  When
 only the destination changes, the route's 'dst' is overwritten with the
 new destination from the forward m_tag.  Then it jumps back at the route
 lookup again and skips the firewall check because it has been marked with
 M_SKIP_FIREWALL.  ipfw 'forward' has to be compiled into the kernel with
 'option IPFIREWALL_FORWARD' to enable it.

 DUMMYNET is entirely handled within the ipfw PFIL handlers.  A packet for
 a dummynet pipe or queue is directly sent to dummynet_io().  Dummynet will
 then inject it back into ip_input/ip_output() after it has served its time.
 Dummynet packets are tagged and will continue from the next rule when they
 hit the ipfw PFIL handlers again after re-injection.

 BRIDGING and IPFW_ETHER are not changed yet and use ipfw_chk() directly as
 they did before.  Later this will be changed to dedicated ETHER PFIL_HOOKS.

More detailed changes to the code:

 conf/files
Add netinet/ip_fw_pfil.c.

 conf/options
Add IPFIREWALL_FORWARD option.

 modules/ipfw/Makefile
Add ip_fw_pfil.c.

 net/bridge.c
Disable PFIL_HOOKS if ipfw for bridging is active.  Bridging ipfw
is still directly invoked to handle layer2 headers and packets would
get a double ipfw when run through PFIL_HOOKS as well.

 netinet/ip_divert.c
Removed divert_clone() function.  It is no longer used.

 netinet/ip_dummynet.[ch]
Neither the route 'ro' nor the destination 'dst' need to be stored
while in dummynet transit.  Structure members and associated macros
are removed.

 netinet/ip_fastfwd.c
Removed all direct ipfw handling code and replace it with the new
'ipfw forward' handling code.

 netinet/ip_fw.h
Removed 'ro' and 'dst' from struct ip_fw_args.

 netinet/ip_fw2.c
(Re)moved some global variables and the module handling.

 netinet/ip_fw_pfil.c
New file containing the ipfw PFIL handlers and module initialization.

 netinet/ip_input.c
Removed all direct ipfw handling code and replace it with the new
'ipfw forward' handling code.  ip_forward() does not longer require
the 'next_hop' struct sockaddr_in argument.  Disable early checks
if 'srcrt' is set.

 netinet/ip_output.c
Removed all direct ipfw handling code and replace it with the new
'ipfw forward' handling code.

 netinet/ip_var.h
Add ip_reass() as general function.  (Used from ipfw PFIL handlers
for IPDIVERT.)

 netinet/raw_ip.c
Directly check if ipfw and dummynet control pointers are active.

 netinet/tcp_input.c
Rework the 'ipfw forward' to local code to work with the new way of
forward tags.

 netinet/tcp_sack.c
Remove include 'opt_ipfw.h' which is not needed here.

 sys/mbuf.h
Remove m_claim_next() macro which was exclusively for ipfw 'forward'
and is no longer needed.

Approved by: re (scottl)

19 years agoFix a deadlock on boot for some systems where reading the battery status
njl [Tue, 17 Aug 2004 18:36:07 +0000 (18:36 +0000)]
Fix a deadlock on boot for some systems where reading the battery status
also generates a notify.  Since we held the lock over this call, the
notify never got to run and the battery status read never returned.
Document this also.

Tested by: Maxim Maximov <mcsi_at_mcsi.pp.ru>
Approved by: re (scottl)

19 years agoDefer the capture of the "expected sync bits" until the first "normal"
gibbs [Tue, 17 Aug 2004 18:12:37 +0000 (18:12 +0000)]
Defer the capture of the "expected sync bits" until the first "normal"
data packet is received from the mouse.  In the case of many KVM's,
this avoids a bug in their mouse emulation that sends back incorrect
sync when you explicitly request a data packet from the mouse.  Without
this change, you must force the driver into stock PS/2 mode or be flooded
with a never ending stream of "out of sync" messages on these KVMs.

Approved by: re

19 years agoUpdate Document Date to match last change.
blackend [Tue, 17 Aug 2004 07:44:25 +0000 (07:44 +0000)]
Update Document Date to match last change.

Approved by: re

19 years ago/me kicks cvs update
trhodes [Tue, 17 Aug 2004 04:56:03 +0000 (04:56 +0000)]
/me kicks cvs update

Revert previous commit, tjr already fixed it and I was too stupid to
notice this fact.

Approved by: re (to avoid failing cvs ci)

19 years agoFix incorrect code in an example. The previous example would produce
trhodes [Tue, 17 Aug 2004 04:45:52 +0000 (04:45 +0000)]
Fix incorrect code in an example.  The previous example would produce
19 column positions wide in the first line and 20 in the rest of the lines.
This fixes the example to provide the correct output.

PR: 53454
Noticed by: Kuang-che Wu <kcwu@kcwu.homeip.net>
Submitted by: Marc Silver <marcs@draenor.org>
Approved by: re (scottl)

19 years agoAnnotate call to DELAY() in interrupt storm mitigation as being
rwatson [Tue, 17 Aug 2004 04:09:09 +0000 (04:09 +0000)]
Annotate call to DELAY() in interrupt storm mitigation as being
something to revisit.

Approved by: re (scottl)

19 years agoAdd ISA attachement files for the aic7xxx driver.
gibbs [Tue, 17 Aug 2004 02:32:30 +0000 (02:32 +0000)]
Add ISA attachement files for the aic7xxx driver.

Approved by: re
Reminded by: obrien

19 years agoAdd an ISA attachement to the aic7xxx driver to handle 284X controllers.
gibbs [Tue, 17 Aug 2004 00:14:31 +0000 (00:14 +0000)]
Add an ISA attachement to the aic7xxx driver to handle 284X controllers.
The ISA probe uses an identify routine to probe all slot locations from
1 to 14 that do not conflict with other allocated resources.  This required
making aic7770.c part of the driver core when compiled as a module.

aic7xxx.c:
aic79xx.c:
aic_osm_lib.c:
Use aic_scb_timer_start() consistently to start the watchdog timer.
This removes a few places that verbatum copied the code in
aic_scb_timer_start().

During recovery processing, allow commands to still be queued to
the controller.  The only requirement we have is that our recovery
command be queued first - something the code already guaranteed.
The only other change required to make this work is to prevent
timers from being started for these newly queued commands.

Approved by: re

19 years ago- Sync whitespace and comments with i386 GENERIC.
marius [Mon, 16 Aug 2004 23:23:21 +0000 (23:23 +0000)]
- Sync whitespace and comments with i386 GENERIC.
- Add some commented out NICs from i386 GENERIC. Most of them look like they
  would work but I'm not sure if they are endian-clean and can't test. There
  was a report that sk(4) works on sparc64 but it doesn't look like it would
  because it doesn't use busdma.
- Improve some of the descriptions of sparc64 specific devices.

There's no functional change, i.e. no added or deleted uncommented devices or
options, in this commit.

19 years agoTry to catch up with reality:
marius [Mon, 16 Aug 2004 23:21:06 +0000 (23:21 +0000)]
Try to catch up with reality:
- Chase the split of pcm(4). This unbreaks LINT compiles.
- sc(4) basically works and a lot of its options should be supported.
- Add the creator and ofw_console drivers.
- vinum(4) should work, at least its module was turned on for sparc64 a while
  ago.
- Don't build sio(4). Its EBus front-end was removed a while ago and the ISA
  one hardly works. Use uart(4) instead, it's not perfect yet but works much
  better.

19 years agociss's interrupt handler was missing the INTR_ENTROPY flag.
peter [Mon, 16 Aug 2004 23:13:16 +0000 (23:13 +0000)]
ciss's interrupt handler was missing the INTR_ENTROPY flag.

19 years agoSync with i386 - Optimize intr_execute_handlers a bit etc.
peter [Mon, 16 Aug 2004 23:12:30 +0000 (23:12 +0000)]
Sync with i386 - Optimize intr_execute_handlers a bit etc.

19 years agoSync with i386 - remove unused includes
peter [Mon, 16 Aug 2004 23:10:46 +0000 (23:10 +0000)]
Sync with i386 - remove unused includes

19 years agoSync with i386 - get the softc via the devclass rather than caching the dev
peter [Mon, 16 Aug 2004 23:10:18 +0000 (23:10 +0000)]
Sync with i386 - get the softc via the devclass rather than caching the dev

19 years agoSync with i386 - add ADAPTIVE_GIANT, remove pcic
peter [Mon, 16 Aug 2004 22:59:24 +0000 (22:59 +0000)]
Sync with i386 - add ADAPTIVE_GIANT, remove pcic

19 years agoSync with i386 - add foot shooting protection for the DDB/KDB thing.
peter [Mon, 16 Aug 2004 22:57:47 +0000 (22:57 +0000)]
Sync with i386 - add foot shooting protection for the DDB/KDB thing.

19 years agoSync with i386 - set rbp reg to 0 for upcalls as a frame marker, not that
peter [Mon, 16 Aug 2004 22:57:13 +0000 (22:57 +0000)]
Sync with i386 - set rbp reg to 0 for upcalls as a frame marker, not that
it is guaranteed to be used in userland though.

19 years agoSync with i386 - trace syscall entry/exit times, and a cosmetic fix.
peter [Mon, 16 Aug 2004 22:56:20 +0000 (22:56 +0000)]
Sync with i386 - trace syscall entry/exit times, and a cosmetic fix.

19 years agoExample configuration files for customize.sh.
thomas [Mon, 16 Aug 2004 22:55:51 +0000 (22:55 +0000)]
Example configuration files for customize.sh.

Approved by: re (scottl)

19 years agoSync with i386 - fix bounds check in lapic_create()
peter [Mon, 16 Aug 2004 22:55:32 +0000 (22:55 +0000)]
Sync with i386 - fix bounds check in lapic_create()

19 years agoSync with i386 - pass resource requests up to parent
peter [Mon, 16 Aug 2004 22:54:50 +0000 (22:54 +0000)]
Sync with i386 - pass resource requests up to parent

19 years agoSync with i386 - s/cpu_swtch/cpu_switch/
peter [Mon, 16 Aug 2004 22:53:29 +0000 (22:53 +0000)]
Sync with i386 - s/cpu_swtch/cpu_switch/

19 years agoSync with i386 - dont count needed bounce pages if loading a buffer that
peter [Mon, 16 Aug 2004 22:53:03 +0000 (22:53 +0000)]
Sync with i386 - dont count needed bounce pages if loading a buffer that
was created with bud_dmamem_alloc()

19 years agoAdd documentation of the newly-added user-configurable Makefile knobs.
thomas [Mon, 16 Aug 2004 22:52:40 +0000 (22:52 +0000)]
Add documentation of the newly-added user-configurable Makefile knobs.

Approved by: re (scottl)

19 years agoSync with i386 - cosmetic fixes
peter [Mon, 16 Aug 2004 22:52:02 +0000 (22:52 +0000)]
Sync with i386 - cosmetic fixes

19 years agoCatch up with i386 - remove lots of no longer used symbolic constants
peter [Mon, 16 Aug 2004 22:51:44 +0000 (22:51 +0000)]
Catch up with i386 - remove lots of no longer used symbolic constants

19 years agoSync with i386
peter [Mon, 16 Aug 2004 22:51:13 +0000 (22:51 +0000)]
Sync with i386

19 years agoAdd example customization script, to be used through CUSTOMIZE Makefile knob.
thomas [Mon, 16 Aug 2004 22:46:58 +0000 (22:46 +0000)]
Add example customization script, to be used through CUSTOMIZE Makefile knob.

Approved by: re (scottl)

19 years agoProperly identify the root filesystem to be used in /etc/fstab in each
thomas [Mon, 16 Aug 2004 22:41:58 +0000 (22:41 +0000)]
Properly identify the root filesystem to be used in /etc/fstab in each
slice of the flash card, ensuring that the loader will mount the root fs
from the booted slice by default.

Allow usage of FFS volume labels instead of hardcoded device names through
WITH_GEOM_VOL Makefile knob.

Approved by: re (scottl)

19 years agoMake 'client DNS forward update' working again which got broken in rev.
mbr [Mon, 16 Aug 2004 22:35:56 +0000 (22:35 +0000)]
Make 'client DNS forward update' working again which got broken in rev.
RC9 of dhclient 3.0.1. This fix will be part of dhclient 3.0.2. It is
also part of the official redhat package.

Submitted by: Jason Vas Dias <jvdias@redhat.com>
Discussed on: dhcp-hackers@isc.org

19 years agoThe existing code fails some corner cases. Replace it with
arun [Mon, 16 Aug 2004 22:09:58 +0000 (22:09 +0000)]
The existing code fails some corner cases. Replace it with
ia64_bsp_adjust() which has been tested to work in all cases for
arbitrary (bsp, nslots) combinations.

reviewed by: marcel@

19 years agoRemove outb to "prime" the EISA ID registers of each slot. This was
gibbs [Mon, 16 Aug 2004 22:05:53 +0000 (22:05 +0000)]
Remove outb to "prime" the EISA ID registers of each slot.  This was
only required to support probing of the Adaptec 284X VLB SCSI controller
which becomes visible in EISA space if you perform these writes.  284X
probing is moving to an ISA attachment.

19 years agoModify the "legacy bus" to pass all resource allocations through to its
gibbs [Mon, 16 Aug 2004 21:55:29 +0000 (21:55 +0000)]
Modify the "legacy bus" to pass all resource allocations through to its
parent rather than track resources locally.  The original code
was incomplete in that it would only honor requests for resources
that already exist in its resource list.  This prevented many ISA
identify routines from allocating temporary resources.  Passing
the requests up to legacy's parent losing no functionality and
allows these requests to succeed.

Reviewed by: imp, jhb
Approved by: RE

19 years agoDocument local changes r1.41 and r1.42 of dhclient.c
mbr [Mon, 16 Aug 2004 21:31:34 +0000 (21:31 +0000)]
Document local changes r1.41 and r1.42 of dhclient.c

19 years agoRecommit removed Rev. 1.40. This fix does solve a FPE with negative lease
mbr [Mon, 16 Aug 2004 21:26:04 +0000 (21:26 +0000)]
Recommit removed Rev. 1.40. This fix does solve a FPE with negative lease
time as described in the PR below.

It seems that this patch should have been part of the vendor tree but got
accidently missed in the 3.0.1 final version. It will definitly be
part of 3.0.2 but until then it's a long way to go.

Submitted by:   ISC (Vendor)
PR:             bin/54517

19 years agoWe now handle SSDT tables. Remove a reference from the BUGS section
marcel [Mon, 16 Aug 2004 20:33:20 +0000 (20:33 +0000)]
We now handle SSDT tables. Remove a reference from the BUGS section
and explicitly mention SSDT when we talk about the DSDT so that people
don't have to guess whether it includes the SSDT.
While here, touch date.

Pointed out by: le@

19 years agoDon't initialize static variables to 0 (C should just take care of that).
philip [Mon, 16 Aug 2004 20:19:09 +0000 (20:19 +0000)]
Don't initialize static variables to 0 (C should just take care of that).

Spotted by: njl

19 years ago- Add my birthday. 24, yay.
pav [Mon, 16 Aug 2004 20:00:08 +0000 (20:00 +0000)]
- Add my birthday. 24, yay.

19 years agoI'm a dumbass: remember to initialize fh->nf_map to NULL in
wpaul [Mon, 16 Aug 2004 19:25:27 +0000 (19:25 +0000)]
I'm a dumbass: remember to initialize fh->nf_map to NULL in
ndis_open_file() in the module loading case.

19 years agoAs I said: the previous commit was untested... Remove an #endif which
marcel [Mon, 16 Aug 2004 19:05:08 +0000 (19:05 +0000)]
As I said: the previous commit was untested... Remove an #endif which
should have ceased to exist when its corresponding #if was removed.

19 years agoCatch up with the drive-by renaming of IA32 to COMPAT_IA32. It must
marcel [Mon, 16 Aug 2004 18:54:23 +0000 (18:54 +0000)]
Catch up with the drive-by renaming of IA32 to COMPAT_IA32. It must
have been rush hour...

While here, move COMPAT_IA32 from opt_global.h to opt_compat.h like on
amd64. Consequently, it's unsafe to use the option in pcb.h. We now
unconditionally have the ia32 specific registers in the PCB.

This commit is untested.

19 years agoThe Texas Instruments ACX111 driver wants srand(), so provide it.
wpaul [Mon, 16 Aug 2004 18:52:37 +0000 (18:52 +0000)]
The Texas Instruments ACX111 driver wants srand(), so provide it.

19 years agoMake the Texas Instruments 802.11g chipset work with the NDISulator.
wpaul [Mon, 16 Aug 2004 18:50:20 +0000 (18:50 +0000)]
Make the Texas Instruments 802.11g chipset work with the NDISulator.
This was tested with a Netgear WG311v2 802.11b/g PCI card. Things
that were fixed:

- This chip has two memory mapped regions, one at PCIR_BAR(0) and the
  other at PCIR_BAR(1). This is a little different from the other
  chips I've seen with two PCI shared memory regions, since they tend
  to have the second BAR ad PCIR_BAR(2). if_ndis_pci.c tests explicitly
  for PCIR_BAR(2). This has been changed to simply fill in ndis_res_mem
  first and ndis_res_altmem second, if a second shared memory range
  exists. Given that NDIS drivers seem to scan for BARs in ascending
  order, I think this should be ok.

- Fixed the code that tries to process firmware images that have been
  loaded as .ko files. To save a step, I was setting up the address
  mapping in ndis_open_file(), but ndis_map_file() flags pre-existing
  mappings as an error (to avoid duplicate mappings). Changed this so
  that the mapping is now donw in ndis_map_file() as expected.

- Made the typedef for 'driver_entry' explicitly include __stdcall
  to silence gcc warning in ndis_load_driver().

NOTE: the Texas Instruments ACX111 driver needs firmware. With my
card, there were 3 .bin files shipped with the driver. You must
either put these files in /compat/ndis or convert them with
ndiscvt -f and kldload them so the driver can use them. Without
the firmware image, the NIC won't work.

19 years agoITC.{i,d} instructions use format M41 not M42.
arun [Mon, 16 Aug 2004 18:41:24 +0000 (18:41 +0000)]
ITC.{i,d} instructions use format M41 not M42.

reviewed by: marcel@

19 years agoWhite space cleanup for netinet before branch:
rwatson [Mon, 16 Aug 2004 18:32:07 +0000 (18:32 +0000)]
White space cleanup for netinet before branch:

- Trailing tab/space cleanup
- Remove spurious spaces between or before tabs

This change avoids touching files that Andre likely has in his working
set for PFIL hooks changes for IPFW/DUMMYNET.

Approved by: re (scottl)
Submitted by: Xin LI <delphij@frontfree.net>

19 years agosys/boot seems to compile fine on Alpha; unbreak ``make release''.
ru [Mon, 16 Aug 2004 18:17:14 +0000 (18:17 +0000)]
sys/boot seems to compile fine on Alpha; unbreak ``make release''.

Prodded by: re-alpha (wilko)

19 years agoComment out rev. 1.4 after problems w/ IPv6, a better solution must be
mlaier [Mon, 16 Aug 2004 17:58:12 +0000 (17:58 +0000)]
Comment out rev. 1.4 after problems w/ IPv6, a better solution must be
found.

19 years agoFix low level code for Tau-PCI/2E1 and Tau-PCI/4E1 that was broken by previous
rik [Mon, 16 Aug 2004 17:28:04 +0000 (17:28 +0000)]
Fix low level code for Tau-PCI/2E1 and Tau-PCI/4E1 that was broken by previous
commit.
Pointy hat: rik

19 years agoAllow i386 binaries to do amr ioctls such as LSI's megamgr on amd64 and
ambrisko [Mon, 16 Aug 2004 17:23:09 +0000 (17:23 +0000)]
Allow i386 binaries to do amr ioctls such as LSI's megamgr on amd64 and
ia64.

PR: 63155
Submitted by: Mikhail Teterin
Tested on: i386, amd64 (via 64bit Xeon system)

19 years agoCreate temporary files safely.
nectar [Mon, 16 Aug 2004 16:37:06 +0000 (16:37 +0000)]
Create temporary files safely.

Submitted by: Jon Passki <cykyc@yahoo.com>

19 years agoUpdate support for Synaptics Touchpads (Volume V)
philip [Mon, 16 Aug 2004 16:28:27 +0000 (16:28 +0000)]
Update support for Synaptics Touchpads (Volume V)

 o Add (long awaited) support for guest devices

Submitted by: Arne Schwabe <arne@rfc2549.org>
Approved by: njl (in a former revision)

19 years agoIt could happen that the renew/rebind times in the leasefile were
mbr [Mon, 16 Aug 2004 16:09:50 +0000 (16:09 +0000)]
It could happen that the renew/rebind times in the leasefile were
expired but the expire time itself was still valid and in the future.
If this happened we ended up using the state S_RENEW with an
INADDR_BROADCAST address set in the request and dhclient never got
any answer back. The only workaround was to delete the lease file.

Fix this case with a check for the S_RENEW and a fallback to the
S_INIT state.

PR: bin/69361
Submitted by: Andrew Gallatin <gallatin@cs.duke.edu>

19 years agoagp_amd64.c is not needed on pc98.
nyan [Mon, 16 Aug 2004 16:01:40 +0000 (16:01 +0000)]
agp_amd64.c is not needed on pc98.

19 years agoSome cards don't have the info entries in the CIS, so vendorstr and/or
imp [Mon, 16 Aug 2004 15:57:18 +0000 (15:57 +0000)]
Some cards don't have the info entries in the CIS, so vendorstr and/or
prodstr may be NULL when fetched.  For the default device description,
guard against this and return the numeric IDs instead when this
happens.  For the matching routines, and consider NULL to not match
those entries that aren't NULL w/o calling strcmp.

Early patches by: Anders Hanssen

19 years agoSync a reverted part of dhclient.c with the vendor source. It seems
mbr [Mon, 16 Aug 2004 15:48:02 +0000 (15:48 +0000)]
Sync a reverted part of dhclient.c with the vendor source. It seems
that a buxfix from rc14 got backed out again. We do the same.

19 years agoInstead of "OpenFirmware", "openfirmware", etc. use the official spelling
marius [Mon, 16 Aug 2004 15:45:27 +0000 (15:45 +0000)]
Instead of "OpenFirmware", "openfirmware", etc. use the official spelling
"Open Firmware" from IEEE 1275 and OpenFirmware.org (no pun intended).

Ok'ed by: tmm

19 years agoAnd isa_if.h
imp [Mon, 16 Aug 2004 15:24:03 +0000 (15:24 +0000)]
And isa_if.h

19 years agoAdd pccbb_isa.c and pccbb_pci.c
imp [Mon, 16 Aug 2004 15:23:10 +0000 (15:23 +0000)]
Add pccbb_isa.c and pccbb_pci.c

19 years agoUpgrading a lock does not play well together with acquiring an exclusive lock
kan [Mon, 16 Aug 2004 15:01:22 +0000 (15:01 +0000)]
Upgrading a lock does not play well together with acquiring an exclusive lock
and can lead to two threads being granted exclusive access. Check that no one
has the same lock in exclusive  mode before proceeding to acquire it.

The LK_WANT_EXCL and LK_WANT_UPGRADE bits act as mini-locks and can block
other threads.  Normally this is not a problem since the mini locks are
upgraded to full locks and the release of the locks will unblock the other
threads.  However if a thread reset the bits without obtaining a full lock
other threads are not awoken. Add missing wakeups for these cases.

PR: kern/69964
Submitted by: Stephan Uphoff <ups at tree dot com>
Very good catch by: Stephan Uphoff <ups at tree dot com>

19 years agoStore a pointer to "null" in struct ndblock's defn member instead of a
tjr [Mon, 16 Aug 2004 14:18:22 +0000 (14:18 +0000)]
Store a pointer to "null" in struct ndblock's defn member instead of a
duplicate allocated on the heap; the address defn points to is significant,
and is checked against the address of "null" in certain conditionals.

PR: 59883
MFC after: 1 week

19 years agoFix compile, s/tp_dtv/tp_tdv/g.
davidxu [Mon, 16 Aug 2004 14:07:38 +0000 (14:07 +0000)]
Fix compile, s/tp_dtv/tp_tdv/g.

19 years agoAdd MP_WATCHDOG option information to pc98, as it uses the i386
rwatson [Mon, 16 Aug 2004 13:59:01 +0000 (13:59 +0000)]
Add MP_WATCHDOG option information to pc98, as it uses the i386
mp_machdep.c, which relies on the option's include file and
defines.

Constancy of: tinderbox

19 years agoAdd /dev/mem and /dev/kmem to powerpc.
ssouhlal [Mon, 16 Aug 2004 13:07:40 +0000 (13:07 +0000)]
Add /dev/mem and /dev/kmem to powerpc.

Approved by: grehan (mentor)

19 years agoComplete 'IA32' -> 'COMPAT_IA32' change for the Linuxulator32.
obrien [Mon, 16 Aug 2004 12:51:33 +0000 (12:51 +0000)]
Complete 'IA32' -> 'COMPAT_IA32' change for the Linuxulator32.

19 years agoUn-comment LINPROCFS.
tjr [Mon, 16 Aug 2004 12:39:27 +0000 (12:39 +0000)]
Un-comment LINPROCFS.

19 years agoAMD64 on-CPU GART support.
obrien [Mon, 16 Aug 2004 12:25:48 +0000 (12:25 +0000)]
AMD64 on-CPU GART support.
This also applies to AMD64 HW running 'i386' OS.

Submitted by: Jung-uk Kim <jkim@niksun.com>
Integration by: obrien

19 years agostyle.9.
obrien [Mon, 16 Aug 2004 12:23:53 +0000 (12:23 +0000)]
style.9.

19 years agoFix the 'DEBUG' argument code to unbreak the amd64 LINT build.
obrien [Mon, 16 Aug 2004 12:15:07 +0000 (12:15 +0000)]
Fix the 'DEBUG' argument code to unbreak the amd64 LINT build.

19 years agoPut the 'antispoof' opcode in the proper place in the opcode list such
obrien [Mon, 16 Aug 2004 12:05:19 +0000 (12:05 +0000)]
Put the 'antispoof' opcode in the proper place in the opcode list such
that it doesn't break the ipfw2 ABI.

19 years agoUpdate low level code for Cronyx Tau-PCI:
rik [Mon, 16 Aug 2004 12:00:07 +0000 (12:00 +0000)]
Update low level code for Cronyx Tau-PCI:
1. Add support for Cronyx Tau-PCI/E3 rev.B.
2. Improve model identification and firmware load procedure.

19 years agoAdd 3653/tcp and 3653/udp, since they were specifically requested.
cperciva [Mon, 16 Aug 2004 11:52:22 +0000 (11:52 +0000)]
Add 3653/tcp and 3653/udp, since they were specifically requested.

PR: conf/63907
Submitted by: Marc Blanchet
MFC after: 3 days

19 years agoI missed an 'IA32' in the documentation.
obrien [Mon, 16 Aug 2004 11:15:46 +0000 (11:15 +0000)]
I missed an 'IA32' in the documentation.

19 years agoFix the 'DEBUG' argument code to unbreak the amd64 LINT build.
obrien [Mon, 16 Aug 2004 11:12:57 +0000 (11:12 +0000)]
Fix the 'DEBUG' argument code to unbreak the amd64 LINT build.

19 years agoI'm not sure what tjr envisioned for turning on FreeBSD/i386 rt support,
obrien [Mon, 16 Aug 2004 11:09:59 +0000 (11:09 +0000)]
I'm not sure what tjr envisioned for turning on FreeBSD/i386 rt support,
but make it COMPAT_IA32 for now.
Fix the 'DEBUG' argument code to unbreak the amd64 LINT build.

19 years agoFix the 'DEBUG' argument code to unbreak the amd64 LINT build.
obrien [Mon, 16 Aug 2004 10:54:25 +0000 (10:54 +0000)]
Fix the 'DEBUG' argument code to unbreak the amd64 LINT build.

19 years agoRemove Wiretek UBRJ4 from the list of supported devices. While it is
simon [Mon, 16 Aug 2004 10:49:45 +0000 (10:49 +0000)]
Remove Wiretek UBRJ4 from the list of supported devices.  While it is
detected by the driver, it doesn't really work as soon as it really used
for network traffic.  Perhaps it can be re-added later when the issues
are resolved.

Reported by: Soeren Vrist <seet@seet.dk>

19 years agoAdd myself as geom_raid3 maintainer.
pjd [Mon, 16 Aug 2004 10:43:41 +0000 (10:43 +0000)]
Add myself as geom_raid3 maintainer.

19 years agoFix the 'DEBUG' argument code to unbreak the LINT build.
obrien [Mon, 16 Aug 2004 10:36:12 +0000 (10:36 +0000)]
Fix the 'DEBUG' argument code to unbreak the LINT build.

19 years agoMinor style.9 cleanup.
obrien [Mon, 16 Aug 2004 10:33:35 +0000 (10:33 +0000)]
Minor style.9 cleanup.

19 years agoBring back 'NO_FLOPPIES', but as a user external knob.
obrien [Mon, 16 Aug 2004 10:25:43 +0000 (10:25 +0000)]
Bring back 'NO_FLOPPIES', but as a user external knob.
The internal logic knob remains MAKE_FLOPPIES.

Suggested by: ru

19 years agoWhen looking for some extra data to include in the hash, use the
dwmalone [Mon, 16 Aug 2004 10:00:44 +0000 (10:00 +0000)]
When looking for some extra data to include in the hash, use the
address of the dirhash, rather than the first sizeof(struct dirhash
*) bytes of the structure (which, thankfully, seem to be constant).

Submitted by: Ted Unangst <tedu@zeitbombe.org>
MFC after: 2 weeks

19 years agoImprove MIME handling. This patch is based on Eugene's patch, but
dwmalone [Mon, 16 Aug 2004 09:38:34 +0000 (09:38 +0000)]
Improve MIME handling. This patch is based on Eugene's patch, but
with the following changes:

1) Don't make a mime_types.h 'cos we should avoid creating variables
   in header files,
2) Use strrchr to find the extension, rather than strchr,
3) Slightly simplify the mime-type matching loop.

any goof are likely to be mine. Note that there are links to more
improvements by Eugene in the PR.

PR: 29725
Submitted by: Eugene Grosbein <eugen@kuzbass.ru>

19 years agoRemove some debug cruft from the Makefiles and make them ${.OBJDIR}
dfr [Mon, 16 Aug 2004 09:35:49 +0000 (09:35 +0000)]
Remove some debug cruft from the Makefiles and make them ${.OBJDIR}
friendly.