]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
21 years agoThe object type can't be OBJT_PHYS in vm_mmap().
Maxime Henrion [Sun, 30 Mar 2003 00:56:20 +0000 (00:56 +0000)]
The object type can't be OBJT_PHYS in vm_mmap().

Reviewed by: peter

21 years agorid is not a pointer to int, but rather just a plain int.
Warner Losh [Sun, 30 Mar 2003 00:30:29 +0000 (00:30 +0000)]
rid is not a pointer to int, but rather just a plain int.

21 years agoDocument what bus_set_resource(9) does. It is used by bus drivers to
Warner Losh [Sun, 30 Mar 2003 00:25:23 +0000 (00:25 +0000)]
Document what bus_set_resource(9) does.  It is used by bus drivers to
bind resources that client drivers later access.

21 years agoFix a bug in the ENOMEM pacing code which probably made it panic systems
Poul-Henning Kamp [Sat, 29 Mar 2003 22:34:37 +0000 (22:34 +0000)]
Fix a bug in the ENOMEM pacing code which probably made it panic systems
after a lot of ENOMEM errors.

21 years ago- Free interrupt handler in ep_free().
Matthew N. Dodd [Sat, 29 Mar 2003 22:27:41 +0000 (22:27 +0000)]
- Free interrupt handler in ep_free().
- Move ep_pccard_detach() to if_ep.c and rename to ep_detach()
- Specify detach methods for all bus frontends.

21 years agoAdd create_geom and destroy_geom methods.
Poul-Henning Kamp [Sat, 29 Mar 2003 22:14:21 +0000 (22:14 +0000)]
Add create_geom and destroy_geom methods.

21 years agoremove a blank line.
Poul-Henning Kamp [Sat, 29 Mar 2003 22:13:32 +0000 (22:13 +0000)]
remove a blank line.

21 years agoUse default geometry in case the device doesn't provide one:
Poul-Henning Kamp [Sat, 29 Mar 2003 22:06:52 +0000 (22:06 +0000)]
Use default geometry in case the device doesn't provide one:
512 for sectorsize.
63 for sectors.
255 for heads.
This will mostly show up on MD(4) devices.

21 years agoChecksum the configuration EEPROM.
Matthew N. Dodd [Sat, 29 Mar 2003 21:57:32 +0000 (21:57 +0000)]
Checksum the configuration EEPROM.

21 years agoBDE'ify
Max Khon [Sat, 29 Mar 2003 21:56:59 +0000 (21:56 +0000)]
BDE'ify

21 years ago- Return error status instead of value in get_e().
Matthew N. Dodd [Sat, 29 Mar 2003 21:44:46 +0000 (21:44 +0000)]
- Return error status instead of value in get_e().
- Modify ep_get_macaddr() to return an error status.
- Reverse the return value logic of eeprom_rdy().

21 years agoSwitch vr_encap over to using m_defrag instead of its own version
Mike Silbersack [Sat, 29 Mar 2003 21:39:49 +0000 (21:39 +0000)]
Switch vr_encap over to using m_defrag instead of its own version
of the function.

No functional change, this driver already handled long mbuf chains
correctly.

MFC after: 2 weeks

21 years agofix truncation check and buffer overflow check
Max Khon [Sat, 29 Mar 2003 21:34:13 +0000 (21:34 +0000)]
fix truncation check and buffer overflow check

21 years agoUpdate if_dc to use m_defrag, removing the semi-duplicate dc_coal
Mike Silbersack [Sat, 29 Mar 2003 21:27:07 +0000 (21:27 +0000)]
Update if_dc to use m_defrag, removing the semi-duplicate dc_coal
function.

Also, use m_defrag where appropriate to defrag long mbuf chains
in the same fashion as was done in if_sis.c.  Before this change,
if_dc would blow up and take down the interface if fed a really long
mbuf chain.

MFC after: 2 weeks

21 years agosendmail 8.12.9 has been imported
Gregory Neil Shapiro [Sat, 29 Mar 2003 19:21:28 +0000 (19:21 +0000)]
sendmail 8.12.9 has been imported

21 years agoFix conflicts from sendmail 8.12.9 import
Gregory Neil Shapiro [Sat, 29 Mar 2003 19:18:07 +0000 (19:18 +0000)]
Fix conflicts from sendmail 8.12.9 import

21 years agoThis commit was generated by cvs2svn to compensate for changes in r112810,
Gregory Neil Shapiro [Sat, 29 Mar 2003 19:12:53 +0000 (19:12 +0000)]
This commit was generated by cvs2svn to compensate for changes in r112810,
which included commits to RCS files with non-trunk default branches.

21 years agoImport sendmail 8.12.9
Gregory Neil Shapiro [Sat, 29 Mar 2003 19:12:53 +0000 (19:12 +0000)]
Import sendmail 8.12.9

21 years agoFix the osf1 abi module on SMP systems by making the size of
Andrew Gallatin [Sat, 29 Mar 2003 18:22:29 +0000 (18:22 +0000)]
Fix the osf1 abi module  on SMP systems by making the size of
a struct pmap be the same on both SMP and UP kernels.

It turns out that the size of a struct pmap is much larger on alpha
SMP systems due to the number of pm_asn's being dependant on MAX_CPU.
Since modules are supposed to be SMP agnostic, this has the affect of
moving around the "interesting bits" of the vmspace (daddr, dsize)
that the osf1 module wants to frob.  So the module ends up scribbling in a
pmap struct, and the user either sees a panic, or an application failure.

While here, I've also shrunk MAXCPU to 8 now that it affects the size
of pmap structs on UP systesm.  This should be plenty, as I'm
unware of any hardware we currently run in which supports more than 8
CPUs.

21 years agoHave sis_encap use m_defrag if:
Mike Silbersack [Sat, 29 Mar 2003 17:50:37 +0000 (17:50 +0000)]
Have sis_encap use m_defrag if:

1.  The chain passed in is > 31 fragments long
or
2.  The chain will not fit in the remaining descriptors without
    defragmentation.

This is slightly less clear than other network drivers because the sis
chips share one descriptor list for all packets, it seems.

Before this change, a > 127 fragment chain would get stuck in the IFQUEUE
permanently, bringing all network traffic to a halt.

MFC after: 2 weeks

21 years agoAlign signal frame placed on stack to 16 bytes so that SSE FPU register
Daniel Eischen [Sat, 29 Mar 2003 16:38:15 +0000 (16:38 +0000)]
Align signal frame placed on stack to 16 bytes so that SSE FPU  register
restores can work.  Also correct allocation for signal frame size.

Reviewed by: mini

21 years agoopt_geom.h is no longer required for this module.
Matthew N. Dodd [Sat, 29 Mar 2003 16:13:11 +0000 (16:13 +0000)]
opt_geom.h is no longer required for this module.

21 years agoAdd 'idt' driver to module build.
Matthew N. Dodd [Sat, 29 Mar 2003 15:56:13 +0000 (15:56 +0000)]
Add 'idt' driver to module build.

21 years ago- Deal with netisr changes.
Matthew N. Dodd [Sat, 29 Mar 2003 15:55:04 +0000 (15:55 +0000)]
- Deal with netisr changes.
- Wrap idt_atm_harpconn() in conditional define T_ATM_BUFQUEUE

21 years agoAdd module infrastructure for 'dpt' driver.
Matthew N. Dodd [Sat, 29 Mar 2003 15:47:13 +0000 (15:47 +0000)]
Add module infrastructure for 'dpt' driver.

21 years agoAdd module infrastructure for 'if_ex'.
Matthew N. Dodd [Sat, 29 Mar 2003 15:41:44 +0000 (15:41 +0000)]
Add module infrastructure for 'if_ex'.

21 years ago- Declare ex_devclass in if_ex.c
Matthew N. Dodd [Sat, 29 Mar 2003 15:38:53 +0000 (15:38 +0000)]
- Declare ex_devclass in if_ex.c
- Rename module datastructures in if_ex_isa.c to be more standard.

21 years ago- Move ex_pccard_detach() to if_ex.c and rename it to ex_detach().
Matthew N. Dodd [Sat, 29 Mar 2003 15:34:26 +0000 (15:34 +0000)]
- Move ex_pccard_detach() to if_ex.c and rename it to ex_detach().
- Add detach method to ISA front end.

21 years agoReformat for readability.
Matthew N. Dodd [Sat, 29 Mar 2003 15:33:04 +0000 (15:33 +0000)]
Reformat for readability.

21 years agoAdd module infrastructure for 'vpd' and 'smbios' and move 'smbios'
Matthew N. Dodd [Sat, 29 Mar 2003 15:20:45 +0000 (15:20 +0000)]
Add module infrastructure for 'vpd' and 'smbios' and move 'smbios'
to new home.

21 years agoDon't dereference flags if NULL (see http.c rev 1.87)
Dag-Erling Smørgrav [Sat, 29 Mar 2003 15:15:38 +0000 (15:15 +0000)]
Don't dereference flags if NULL (see http.c rev 1.87)

21 years agoRemove useless code.
Matthew N. Dodd [Sat, 29 Mar 2003 14:51:50 +0000 (14:51 +0000)]
Remove useless code.

21 years agoDon't compile the identify method yet.
Matthew N. Dodd [Sat, 29 Mar 2003 14:50:14 +0000 (14:50 +0000)]
Don't compile the identify method yet.

21 years agoRegister dependency on the elink module.
Matthew N. Dodd [Sat, 29 Mar 2003 13:45:17 +0000 (13:45 +0000)]
Register dependency on the elink module.

21 years agoAdd module infrastructure for elink.
Matthew N. Dodd [Sat, 29 Mar 2003 13:42:20 +0000 (13:42 +0000)]
Add module infrastructure for elink.

21 years agoAdd module infrastructure for if_ie.
Matthew N. Dodd [Sat, 29 Mar 2003 13:40:41 +0000 (13:40 +0000)]
Add module infrastructure for if_ie.

21 years agoSecond round of updates to the ATA driver.
Søren Schmidt [Sat, 29 Mar 2003 13:37:09 +0000 (13:37 +0000)]
Second round of updates to the ATA driver.

Clean up the DMA interface too much unneeded stuff crept in with
the busdma code back when.

Modify the ATA_IN* / ATA_OUT* macros so that resource and offset
are gotten from a table. That allows for new chipsets that doesn't
nessesarily have things ordered the good old way. This also removes
the need for the wierd PC98 resource functions.

Tested on: i386, PC98, Alpha, Sparc64

21 years ago- Move driver to newbus.
Matthew N. Dodd [Sat, 29 Mar 2003 13:36:41 +0000 (13:36 +0000)]
- Move driver to newbus.
- Provide identify methods for EtherExpress and 3c507 cards; this
  means these cards no longer need wired configs.
- Provide a detach method.

21 years ago- Change ELINK_ID_PORT; the 3c507 likes 0x100 better.
Matthew N. Dodd [Sat, 29 Mar 2003 13:18:20 +0000 (13:18 +0000)]
- Change ELINK_ID_PORT; the 3c507 likes 0x100 better.
- Add module metadata.

21 years agoClear the IFF_OACTIVE flag at the end of ieinit(). This prevents
Matthew N. Dodd [Sat, 29 Mar 2003 12:42:01 +0000 (12:42 +0000)]
Clear the IFF_OACTIVE flag at the end of ieinit().  This prevents
the adapter from becoming wedged when when the interface is
is brought up by ether_ioctl() (when you set an IP address for example.)

Confirmed this "fix" from NetBSD's i82586 backend.  It seems hackish
to me but whatever.

21 years agoFill in the rest of the fields in the resulting struct tm
Mike Makonnen [Sat, 29 Mar 2003 11:55:37 +0000 (11:55 +0000)]
Fill in the rest of the fields in the resulting struct tm
from strptime(3). Previously, they would get filled only
for the %s specifier and as a side effect of using the
the %Z specifier with a GMT time zone.

PR: misc/48993
Approved by: markm (mentor)
Silence on: -standards

21 years agoCorrect a typo in previous commit
Kris Kennaway [Sat, 29 Mar 2003 10:12:15 +0000 (10:12 +0000)]
Correct a typo in previous commit

21 years ago'speaker' not 'atspeaker'
Matthew N. Dodd [Sat, 29 Mar 2003 10:01:03 +0000 (10:01 +0000)]
'speaker' not 'atspeaker'

21 years agoRetire 'atspeaker' and 'pcspeaker'. We'll use 'speaker' which
Matthew N. Dodd [Sat, 29 Mar 2003 09:59:49 +0000 (09:59 +0000)]
Retire 'atspeaker' and 'pcspeaker'.  We'll use 'speaker' which
is the same as the config device directive.

Reminded by:  Takahashi Yoshihiro <nyan@jp.FreeBSD.org>

21 years agoAdjust shared memory allocation to work on cards with only 8 KB of RAM.
Matthew N. Dodd [Sat, 29 Mar 2003 09:49:22 +0000 (09:49 +0000)]
Adjust shared memory allocation to work on cards with only 8 KB of RAM.

21 years agoClean up argument comments for bus_dma_tag_create() calls.
Matthew N. Dodd [Sat, 29 Mar 2003 09:46:10 +0000 (09:46 +0000)]
Clean up argument comments for bus_dma_tag_create() calls.

21 years agofixed a mbuf leak when an IP packet from ESP tunnel is redirected
SUZUKI Shinsuke [Sat, 29 Mar 2003 08:31:28 +0000 (08:31 +0000)]
fixed a mbuf leak when an IP packet from ESP tunnel is redirected

obtained from: KAME

21 years ago- Track resources in our softc.
Matthew N. Dodd [Sat, 29 Mar 2003 08:30:45 +0000 (08:30 +0000)]
- Track resources in our softc.
- Sanitize dpt_alloc().
- Add helper functions for resource alloc/release.
- Add detach method.
- Relocate definition of devclass_t.
- Move some debugging output behind bootverbose.
- Implement an identify method for ISA devices
  but don't use it right now.

21 years agoFix up the long mbuf chain recovery code to use m_defrag; the old
Mike Silbersack [Sat, 29 Mar 2003 06:34:25 +0000 (06:34 +0000)]
Fix up the long mbuf chain recovery code to use m_defrag; the old
code messed up on B & C chipsets because it lost the packet header
and therefore the flag indicating the need for hardware checksums.

MFC after: 2 weeks

21 years agoPass the vm_page's address to sf_buf_alloc(); map the vm_page as part
Alan Cox [Sat, 29 Mar 2003 06:14:14 +0000 (06:14 +0000)]
Pass the vm_page's address to sf_buf_alloc(); map the vm_page as part
of sf_buf_alloc() instead of expecting sf_buf_alloc()'s caller to map it.

The ultimate reason for this change is to enable two optimizations:
(1) that there never be more than one sf_buf mapping a vm_page at a time
and (2) 64-bit architectures can transparently use their 1-1 virtual
to physical mapping (e.g., "K0SEG") avoiding the overhead of pmap_qenter()
and pmap_qremove().

21 years agoAdd the m_defrag routine, as discussed on committers@. This
Mike Silbersack [Sat, 29 Mar 2003 05:48:36 +0000 (05:48 +0000)]
Add the m_defrag routine, as discussed on committers@.  This
incarnation should address the concerns of all in the discussion,
and keeps statistics which show how much it is used.

MFC after: 2 weeks

21 years agoAdd MBUF_FRAG_TEST to NOTES.
Mike Silbersack [Sat, 29 Mar 2003 05:46:34 +0000 (05:46 +0000)]
Add MBUF_FRAG_TEST to NOTES.

Submitted by: Hiten Pandya <hiten@unixdaemons.com>

21 years agoNote: this is OLDCARD, and tweak comment about clpd 6729/30.
Warner Losh [Sat, 29 Mar 2003 05:44:49 +0000 (05:44 +0000)]
Note: this is OLDCARD, and tweak comment about clpd 6729/30.

21 years agoAdd a new man page for bus_activate_resource and
Warner Losh [Sat, 29 Mar 2003 05:16:15 +0000 (05:16 +0000)]
Add a new man page for bus_activate_resource and
bus_deactivate_resource.  It could use some improvement, and others
are welcomed to do so :-)

21 years agoClearify what happens with the cookiep in various cases.
Warner Losh [Sat, 29 Mar 2003 05:11:14 +0000 (05:11 +0000)]
Clearify what happens with the cookiep in various cases.

21 years agoFix some typos:
Warner Losh [Sat, 29 Mar 2003 05:09:15 +0000 (05:09 +0000)]
Fix some typos:
race->rare
rman_activate_resource -> bus_activate_resource

Add xref to bus_activate_resource
Minor markup changes for the start/end/count parameters.

21 years ago320.rdist is OBE as we don't have rdist in the base system any more.
David E. O'Brien [Sat, 29 Mar 2003 03:28:28 +0000 (03:28 +0000)]
320.rdist is OBE as we don't have rdist in the base system any more.

21 years agoThis is OBE as we don't have rdist in the base system any longer.
David E. O'Brien [Sat, 29 Mar 2003 03:27:49 +0000 (03:27 +0000)]
This is OBE as we don't have rdist in the base system any longer.

21 years agoGlobally use -mieee in /usr/src for Alpha.
David E. O'Brien [Sat, 29 Mar 2003 03:14:24 +0000 (03:14 +0000)]
Globally use -mieee in /usr/src for Alpha.

21 years agoReduce "code duplication" for AMD CPU's.
David E. O'Brien [Sat, 29 Mar 2003 03:13:19 +0000 (03:13 +0000)]
Reduce "code duplication" for AMD CPU's.

21 years agoBack out the totally unapproved backout of my commits.
David E. O'Brien [Sat, 29 Mar 2003 03:03:42 +0000 (03:03 +0000)]
Back out the totally unapproved backout of my commits.

21 years agoSmall mdoc fix
Sean Chittenden [Sat, 29 Mar 2003 02:09:02 +0000 (02:09 +0000)]
Small mdoc fix

21 years agoText formatting changes.
Matthew N. Dodd [Sat, 29 Mar 2003 01:53:13 +0000 (01:53 +0000)]
Text formatting changes.

21 years agoDon't panic if we aren't plugged into a hub.
Matthew N. Dodd [Sat, 29 Mar 2003 01:40:42 +0000 (01:40 +0000)]
Don't panic if we aren't plugged into a hub.

21 years agoadd missing copyright notices
Sam Leffler [Fri, 28 Mar 2003 20:28:05 +0000 (20:28 +0000)]
add missing copyright notices

Noticed by: Robert Watson

21 years agoIf IPCP is disabled, susccess of IPV6CP negotiation is sufficient
Hajimu UMEMOTO [Fri, 28 Mar 2003 18:23:43 +0000 (18:23 +0000)]
If IPCP is disabled, susccess of IPV6CP negotiation is sufficient
to communicate by IPv6.  So, the prompt should be `PPP' rather
than `PPp'.

21 years agoTypo fixes. Translators can probably ignore these.
Bruce A. Mah [Fri, 28 Mar 2003 18:22:23 +0000 (18:22 +0000)]
Typo fixes.  Translators can probably ignore these.

Submitted by: ispell(1)

21 years agoNew release notes: adduser(8) -g, gstat(8), omshell(1)/dhcpctl(3).
Bruce A. Mah [Fri, 28 Mar 2003 18:14:54 +0000 (18:14 +0000)]
New release notes:  adduser(8) -g, gstat(8), omshell(1)/dhcpctl(3).

21 years agoCheck for the PS_NEEDSIGCHK flag in the right flags field.
John Baldwin [Fri, 28 Mar 2003 18:08:57 +0000 (18:08 +0000)]
Check for the PS_NEEDSIGCHK flag in the right flags field.

21 years agoAdd an option to print the time in 12-hour format.
Maxime Henrion [Fri, 28 Mar 2003 16:37:45 +0000 (16:37 +0000)]
Add an option to print the time in 12-hour format.

Submitted by: Samy Al Bahra <samy@kerneled.com>

21 years agoPerform warm reset if codec reports ready on attach.
Orion Hodson [Fri, 28 Mar 2003 16:33:15 +0000 (16:33 +0000)]
Perform warm reset if codec reports ready on attach.

Move AC97 link control power up poke so it is always performed on attach.

21 years agoxargs(1) is WARNS=6 clean.
Maxime Henrion [Fri, 28 Mar 2003 16:24:11 +0000 (16:24 +0000)]
xargs(1) is WARNS=6 clean.

Tested on: i386, sparc64

21 years agoWhen -o isn't specified, open /dev/null as stdin.
Maxime Henrion [Fri, 28 Mar 2003 16:22:54 +0000 (16:22 +0000)]
When -o isn't specified, open /dev/null as stdin.

Suggested by: jhb

21 years agoMove details of dev_t (and udev_t) to <sys/_types.h>.
Mike Barcroft [Fri, 28 Mar 2003 15:27:30 +0000 (15:27 +0000)]
Move details of dev_t (and udev_t) to <sys/_types.h>.

21 years agoSysctl treats `long' and `unsigned long' differently - fix a comment.
Robert Drehmel [Fri, 28 Mar 2003 14:17:17 +0000 (14:17 +0000)]
Sysctl treats `long' and `unsigned long' differently - fix a comment.

21 years ago- MAXPATHLEN -> PATH_MAX (pass correct buffer size to readlink as well)
Max Khon [Fri, 28 Mar 2003 12:05:45 +0000 (12:05 +0000)]
- MAXPATHLEN -> PATH_MAX (pass correct buffer size to readlink as well)

Requested by: bde

21 years agoAdd pc-card from ARCHOS CD-224E cdrom.
Maxim Sobolev [Fri, 28 Mar 2003 11:12:52 +0000 (11:12 +0000)]
Add pc-card from ARCHOS CD-224E cdrom.

MFC after: 1 week

21 years agobump __FreeBSD_version for thread-safe realpath(3)
Max Khon [Fri, 28 Mar 2003 09:50:26 +0000 (09:50 +0000)]
bump __FreeBSD_version for thread-safe realpath(3)

21 years agoFix an XXX: and implement LINUX_BLKGETSIZE correctly.
Poul-Henning Kamp [Fri, 28 Mar 2003 08:58:11 +0000 (08:58 +0000)]
Fix an XXX: and implement LINUX_BLKGETSIZE correctly.

21 years agoExpound a little bit about the softc, where its size is found and a
Warner Losh [Fri, 28 Mar 2003 07:28:31 +0000 (07:28 +0000)]
Expound a little bit about the softc, where its size is found and a
caution against emulating legacy softc methods used in the tree.

21 years agoRemove bogus card.4, and replace it with a link to oldcard.4. ru
Warner Losh [Fri, 28 Mar 2003 07:14:12 +0000 (07:14 +0000)]
Remove bogus card.4, and replace it with a link to oldcard.4.  ru
pointed this out a while ago, and I'm just now getting back to it.  It
obsoletes one of ru's changes, but since oldcard.4 is so much better,
that doesn't matter.

Submitted by: ru

21 years agoCreate a man page for bus_child_present(9), a function to see if a
Warner Losh [Fri, 28 Mar 2003 07:09:17 +0000 (07:09 +0000)]
Create a man page for bus_child_present(9), a function to see if a
given device is considered accessible by its parent device bus driver
(or its parent, etc).

21 years agopccarddevs_data.h no longer generated
Warner Losh [Fri, 28 Mar 2003 06:44:57 +0000 (06:44 +0000)]
pccarddevs_data.h no longer generated

21 years agoMFp4:
Warner Losh [Fri, 28 Mar 2003 06:43:50 +0000 (06:43 +0000)]
MFp4:

Don't generate the data file.
Use 0xffffffff rather than -1 for sign safety as these fields are uint32_t.

21 years ago- Clean up function calling conventions.
Matthew N. Dodd [Fri, 28 Mar 2003 06:27:08 +0000 (06:27 +0000)]
- Clean up function calling conventions.
- Be consistent about what we call our softc.
- Minor formatting.
- Add some register definitions gleaned from NetBSD/Linux.

21 years agoAllow m_dup_pkthdr to accept mbufs with attached clusters as
Mike Silbersack [Fri, 28 Mar 2003 05:57:48 +0000 (05:57 +0000)]
Allow m_dup_pkthdr to accept mbufs with attached clusters as
targets.

Submitted by: bmilekic

21 years agoAdd a checksum to the kernel message buffer, and update it every
Ian Dowse [Fri, 28 Mar 2003 02:50:10 +0000 (02:50 +0000)]
Add a checksum to the kernel message buffer, and update it every
time a character is written. Use this at boot time to reject the
existing buffer contents if they are corrupt. This fixes a problem
seen on some hardware (especially laptops) where the message buffer
gets partially corrupted during a short power cycle or reset, but
the msgbuf structure is left intact so it gets reused, resulting
in random junk and control characters appearing in dmesg and
/var/log/messages.

PR: kern/28497

21 years agoMulticast support, gleaned from NetBSD's if_iy driver.
Matthew N. Dodd [Fri, 28 Mar 2003 02:15:01 +0000 (02:15 +0000)]
Multicast support, gleaned from NetBSD's if_iy driver.

Submitted by:  Craig Boston <craig@xfoil.gank.org>

21 years agoClose the disk file descriptor that is RO before trying to open the
Juli Mallett [Fri, 28 Mar 2003 01:50:11 +0000 (01:50 +0000)]
Close the disk file descriptor that is RO before trying to open the
new one, and do not fall back to the RO fd.  There was a bug here
in that the RO fd was never closed, if the RDRW open succeeded, but
this code is bogus anyway, and it breaks newfs of floppies, at least
for me, due to "Device busy."  Anything that wants to fall back is
doing something significantly odd that it should have some more complex
code on its end.

21 years agoDefault data length should be 56, not 64.
Matthew N. Dodd [Fri, 28 Mar 2003 00:31:51 +0000 (00:31 +0000)]
Default data length should be 56, not 64.

21 years agoMake realpath() thread-safe. New implementation does not use chdir(2) at all.
Max Khon [Thu, 27 Mar 2003 20:48:53 +0000 (20:48 +0000)]
Make realpath() thread-safe. New implementation does not use chdir(2) at all.

Submitted by: Constantin S. Svintsoff <kostik (at) iclub.nsu.ru>

21 years agoCheck return value from vmapbuf instead of the function address.
Tor Egge [Thu, 27 Mar 2003 20:48:34 +0000 (20:48 +0000)]
Check return value from vmapbuf instead of the function address.

21 years agoDon't pass NULL in as a flags argument to libfetch's fetchGetURL()
Robert Watson [Thu, 27 Mar 2003 20:44:09 +0000 (20:44 +0000)]
Don't pass NULL in as a flags argument to libfetch's fetchGetURL()
as this can result in a NULL pointer deference when parsing the
flags later.  This change fixes "pkg_add -r" on 5.0-CURRENT for
me; not quite clear how the problem was introduced.

21 years agoCorrect LDBL_* constants based on values from i386.
David Schultz [Thu, 27 Mar 2003 20:38:22 +0000 (20:38 +0000)]
Correct LDBL_* constants based on values from i386.

21 years ago- Rely on ether_ifattach() to initialize part of our struct ifnet
Matthew N. Dodd [Thu, 27 Mar 2003 20:38:00 +0000 (20:38 +0000)]
- Rely on ether_ifattach() to initialize part of our struct ifnet
  rather than doing it ourself.
- Set ifq_maxlen.
- Include the size of the ethernet header when we allocate
  our mbuf chain in ieget().

21 years agoAdd an information and fix typo.
Takeshi Shibagaki [Thu, 27 Mar 2003 19:33:22 +0000 (19:33 +0000)]
Add an information and fix typo.

21 years agoEliminate a buffer sleep/wakeup race.
Tor Egge [Thu, 27 Mar 2003 19:28:11 +0000 (19:28 +0000)]
Eliminate a buffer sleep/wakeup race.

21 years agoTrim "trustedbsd_" from the front of the policy module "short names";
Robert Watson [Thu, 27 Mar 2003 19:26:39 +0000 (19:26 +0000)]
Trim "trustedbsd_" from the front of the policy module "short names";
the vendor is only included in the long name currently, reducing
verbosity when modules are registered and unregistered.

Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories

21 years agoAdd missing includes from previous commit.
John Baldwin [Thu, 27 Mar 2003 18:18:35 +0000 (18:18 +0000)]
Add missing includes from previous commit.

Reported by: des

21 years agoFix indentation.
Maxim Konovalov [Thu, 27 Mar 2003 15:00:10 +0000 (15:00 +0000)]
Fix indentation.