This is a partial list of the major machine-independent changes
(i.e., these are the changes people ask about most often). Machine
specific changes have also been made, and are sometimes mentioned
in the pages for the specific platforms.
Note: Problems for which patches exist are marked in red.
For changes in other releases, click below:
2.0,
2.1,
2.2,
2.3,
2.4,
2.5,
2.6,
2.7,
2.8,
2.9,
3.0,
3.1,
3.2,
3.3,
3.4,
3.5,
3.6,
3.7,
3.8,
3.9,
4.1,
4.2,
4.3,
4.4,
4.5,
4.6,
4.7,
4.8,
4.9,
5.0,
5.1,
5.2,
5.3,
5.4,
5.5,
5.6,
5.7,
5.8,
5.9,
6.0,
6.1,
6.2,
6.3,
6.4,
6.5,
6.6,
6.7,
6.8,
6.9,
7.0,
7.1,
7.2,
7.3,
7.4,
current.
Changes made between OpenBSD 3.9 and 4.0
- Plug memory leak in top(1).
- Limit maximum work done in ssh(1) performed by the CRC compensator attack detector.
- Fix for error message parsing in pkg_add(1).
- Initial import of uath(4), an Atheros USB2.0 WLAN device. Added to i386 GENERIC.
- Avoid a write to NULL on mac68k.
- Bring pgt(4) to a working state and enable it on i386 GENERIC.
- Rework mpi(4) error handling from hardware so the scsi midlayer can handle them correctly.
- Do not ask for HWERR interrupts on msk(4) and not acknowledging them later. Leads to interrupt storms on several machines.
- Also add Intel 6321ESB SATA support to pciide(4).
- Add VIA VT8237A SATA to supported devices. Untested but should work.
- Reserve some command slots in ciss(4) for bioctl/sensorsd so ciss does not crash under load.
- Make sure re(4) devices preserve mac address on thecus if the hardware mac is a specific number.
- Reset the armish sysfail LED on thecus, so it does not blink.
- Fixes for integer overflows in CID encoded fonts parsing for X.org.
- Fix RSA signature padding vulnerability in OpenSSL.
- Security update to BIND 9.3.2-P1.
- Teach sasyncd(8) to set isakmpd(8) up in passive mode, based on current carp(4) state.
- Add command to isakmpd(8) to force it into passive mode. Will be used by sasyncd(8) in HA setups.
- Fix for UltraSparcIII sparc64 machines with more than one memory bank filled.
- Add IPv6 link-local addresses to carp(4) interfaces when virtual MAC is set. Makes route6d work on systems with carp(4) interfaces.
- Even more dhclient link detection adjustments.
- Fix for getting mac address for bge(4) on sparc64, should fix Blade 1500 problems.
- Fixes for the VAX ze(4) driver, and various other small VAX fixes. Allows VXT2000+ machine to (net)boot multiuser.
- Have applications that write to pf_key socket understand EAGAIN errors and retry later.
- Make sure atw(4) does not panic when channel tuning code returns IEEE_80211_CHAN_ANY.
- In netstat(1), fix KVM snooping code so netstat -A work again.
- Enable arc(4) on amd64 GENERIC and RAMDISK_CD kernels.
- Fixed handling of 802.11 beacon frames by stations while associated. Should fix BSS mode with both 802.11b and 802.11g stations.
- Added support for UFQDNs in ipsec.conf(5).
- Fix for the 802.11 rate handling code to prevent devices being forced into protection mode and degrade performance.
- Added support for IKE AH rules to ipsecctl(8).
- Add support on UltraSparc machines for the onboard bge(4) devices.
- Allow arbitrary values when changing the carp demotion counter, as long as the result still is in valid range.
- In ssh(1) work around a problem in Heimdal when KRB5CCNAME file is missing.
- Fix a possible kernel panic in the ucom(4) code caused by using already freed data in ttyclose().
- Have dhclient(1) take net devices up in order for them to do media negotiation, since some devices requires this before the dhclient(1) 10 second link test.
- Fixes for the installation scripts when parsing site-version-hostname.tgz files.
- Make sure sparc and sparc64 platforms create more tty device files.
- Allow building of procfs on arches that do not define PT_STEP, like sparc64.
- Rework iwi(4) RX buffer allocation to make sure it never does DMA outside the mapped buffers.
- Enable arc(4) on i386 RAMDISK kernels.
- Make sure the sppp(4) code check for illegal options in LCP packets and drop such packets.
- For sparc64, make sure onboard bge(4) gets the MAC address from the OpenFirmWare local-mac-address.
- Fix for disc size check in cdio(1).
- Fixes for LED handling on bge(4) chipsets found in some Dell machines.
- For sparc64, enable HORRID_III_HACK, but make sure USI & II CPUs get to keep their data caches enabled. GENERIC now runs on USIII too.
- Add some support for bio(4) on ciss(4) controllers.
- PCI interrupt fixes for sparc64, makes it work on v210 machines.
- Work in progress of VXT2000 support for vax.
- Endian fix in bgpd(8) regarding sending of route refreshes.
- Rework arc(4) code for detecting degraded volumes, since it would report degraded disks as online.
- For armish platforms with fintek(4), have the fans blow at full speed.
- Disable interrupts during the attach of UHCI and EHCI controllers. Avoids possible NULL pointer reference.
- Make sure AMD PowerNow code does not propagate low level errors to userland on i386 and amd64.
- Add support for VIA C3/C7 crypto driver HMACs.
- Make tcpdump(8) recognize ipsec xauth vendor payload.
- For ral(4) and rum(4), fallback to 1Mbps if tx rate is returned as zero. Avoids division by zero seen on amd64 sometimes.
- Make ld.so(1) on ARM platform support libraries with DT_TEXTREL.
- Add Nova Tech NV-902W to the list of supported ural(4) devices.
- For i386, provide some smp locking around protection fault trap from user too.
- Initial import of code to support the PrismGT pgt(4) chipsets.
- Change bgpd(8) back to not send empty UPDATE messages, some routers do not handle it very well.
- Enable acx(4) on amd64 GENERIC kernels too.
- Have carp(4) interfaces answer ARP queries if the interface is in a bridge with the physical interface that received it.
- Add support for Jumbo frames in bnx(4).
- Enable RX checksum offload in bnx(4).
- Add Intel 6321ESB and 82801H SMBus controllers to ichiic(4).
- Add Intel 6300ESB and 6321ESB AC97 controllers to auich(4).
- Add Intel 82801H/ICH8 SATA support to pciide(4). Untested but should work.
- Add lm75a to the list of supported lmtemp(4) devices.
- Lots of small fixes for iwi(4).
- Add ucycom(4) to all USB capable arches.
- Fix some file descriptor leaks in XFree error handlers.
- for i386, add a new Core Xeon CPU and an older Pentium Pro model to the bus clock detection.
- Add support for RT5225 in rum(4) and various other rum(4) fixes.
- Disable SPNEGO as per RFC4462 in ssh(1).
- Add bio(4) sensor support to arc(4) devices.
- Enable msk(4) for alpha, amd64, macppc and sparc64 too.
- Fixes in the VFS layer for proper mountpoint selection when mounting softdep file systems in a dir of a async mounted file system.
- On hp300 with 68040, setup the transparent translation with writeback cache instead of writethrough.
- Major rework of the hp300 boot code, removes lots of old junk and adds sgc frame buffer console support.
- Jumpscroll code for vt100 and sun wscons emulations. Speeds up macppc and zaurus console a lot.
- Add msk(4) to i386 GENERIC and RAMDISK kernels.
- Initial driver msk(4), for Marvell Yukon-2 Gigabit ethernet controllers.
- Add BelkinRT2601 USB device to list of supported rum(4) devices.
- Have carp(4) send out a second delayed gratuitious ARP request after the MASTER election is won, so wacky L3 switches picks it up correctly.
- Do not misreport some VS4k/90 machines as uVAXens.
- LCSPX probe for vax now more reliable.
- Add Belkin F5D705C to list of supported zyd(4) devices.
- Initial code for a ucycom(4) driver.
- Make bgpd(8) send an "empty" update message after a bulk transfer, used as End-of-RIB marker.
- Add bnx(4) to i386 and amd64 GENERIC and RAMDISK kernels.
- Make vacation(1) also add headers as per RFC 3834.
- For ssh(1), add ability to match groups to Match keyword in sshd_config(5).
- Add uslcom(4) and uark(4) on all USB capable arches.
- Initial driver for uark(4), ARK3116 USB UART devices.
- Fix for systrace symlink handling when calling rename(2).
- Fixes for USB2.0 error handling.
- Enable arc(4) on sparc64 GENERIC kernels.
- Make cron(8) add Auto-Submitted header in generated mail as per RFC 3834.
- Fix a panic when trying to fdisk vnd(4) devices which are backed by weird images.
- Fixes for domain keyboards on hp300 models 362 and 382 aswell.
- Fix for hp300 HIL keyboards during boot.
- Update to sendmail 8.13.8.
- Enable hostap mode for acx(4).
- Preliminary import of Xwsfb for vax, still has endian issues on smg.
- Make signal strength for acx(4) rssi depend on the radio type, so percentage value works.
- Enable rum(4) on all arches where ural(4) is enabled.
- In OpenRCS, correctly handle -e and -E flags for merge(1) and rcsmerge(1).
- For hp300, try and determine keyboard layout from its identification string.
- Fix a panic regarding SysV semaphores when overallocating more than the default number.
- Fixes in xge(4) for DMA maps and RX descriptor maps.
- Fix a use-after-free in mpi(4).
- Fix for dhcpd(8) where a DHCPDISCOVER packet with a 32-bit client identifier would cause dhcpd(8) to mistakenly exit.
- On hp300 platforms, use the blitter again for the cursor on non-mono frame buffers.
- Lots of fixes for the rum(4) driver.
- Endianess fix in ral(4) in the rate adaptation code.
- Sync em(4) to Intels latest FreeBSD em driver (6.1.4). Adds support for the PCIe quad port copper adapter, improvements for media support with fiber adapters and some fixes for the ICH8 support.
- Avoid div-by-zero in visual effects in fdformat(1).
- Fix use-after-free bug in sendmail(8), triggered with too long header lines.
- Enable WEP on acx(4).
- ld.so(1) fix for Alpha in regard of gcc3.
- Endian fixes for both wpi(4) and arc(4). arc(4) works on sparc64 now.
- Switch luna88k platform to rasops from rcons.
- Add BPF hooks to acx(4).
- Enable arc(4) in i386 GENERIC kernel.
- Fix for re(4) on PCIe busses where a second packet might get stuck if sent in short succession of an earlier packet.
- Add Intel Raid Controller SRCSAS18E, and SRCSAS144E to mfi(4).
- Add Intel 6321ESB to the pciide(4) driver.
- In the re(4) driver, restructure TX/RX descriptor handling, and bump the number of TX descriptors for rtl8169-based cards.
- Various fixes to the vax bootloader code.
- Let pkg_add(1) check for errors on all reads/writes.
- Make sure ssh(1) can build without kerberosV support, and with SKEY defined too.
- Make the scsi layer report ENOMEDIUM and EMEDIUMTYPE to userland, so tar(1), dd(1) and friends can report useful error messages when there is no media in the device.
- Add "restart" to max-prefix in bgpd(8), allows restart of suspended sessions after a given number of minutes.
- Enable acx(4) on i386 GENERIC kernels.
- Have bgpd(8) clean the routing table of all PROTO1 routes, since they might be leftovers from an earlier crash.
- Re-enable IPMI on i386/amd64 GENERIC kernels.
- Initial import of the acx(4) wireless driver.
- Disable jumbo frames on xge(4) until reception of jumbo frames is fixed.
- Add NVIDIA MCP51 AC97 to the auich(4) driver.
- Have vax support font widths from 9 to 16 bits, and use 12x22 font by default.
- Make ssh(1) allow fallback to known_hosts entries without port qualifiers for non-standard ports too.
- Make pkg_delete(1) be better in reporting dependency problems when removing multiple packages at the same time.
- Added puc(4) at cardbus for amd64, i386 and macppc, but commented out.
- For amd64 GENERIC kernels, enable an(4) at pci. Untested, but should work.
- Armish platform now allows argument passing from bootblock to kernel.
- Add more strict receive path validation code to sk(4).
- Fixed race condition in bgpd(8) when neighbor reached max-prefix, leading to a crash.
- For dz(4), allow keyboard attachment on VS 4k/90 aswell.
- Various vax fixes for console detection.
- Improved armish bootloader into an interactive one like i386. Defaults to boot from wd0.
- Make sure never to attach a dz(4) console to a keyboard port on VAXstations so machines with unsupported frame buffers can work with serial console.
- Fixes for wscons(4) code on non-i386 platforms so an older X server runs.
- First version of arc(4), the Areca RAID controller driver. Not usable yet.
- Fix for scsi_inquiry_data version number parsing. Fixes SCSI5 and SCSI6 in dmesg.
- Initial driver for color frame buffers found on VAXstation 3100 models 3x/4x. Rough but usable.
- Make ipmi code only read one sensor at a time, also lower the polling frequency to every 5 seconds instead of every 10.
- Make sure mount_nfs(8) allows synchronous mounts.
- Initial code for an armish bootloader.
- Fix in httpd(8) for mod_rewrite which potentially could be exploited.
- Add some new twe(4) pci ids.
- Compile all kernels with the new stack usage warning code in gcc(1).
- Add cmpci(4) to i386, alpha and amd64.
- Fixes for rcs binary file handling.
- Add checks to all fdisk(8) write attempts, and try to keep it in a consistent state upon write failure.
- Malloc fixes for rcs(1) and cvs(1).
- Newer cmcpci(4) sound driver for i386.
- Make sure kernels with VFSDEBUG option set can boot again.
- Fix several timeout quirks in tftp(1) and tftpd(8).
- For vax, replace kvtope() and kvtophys() with assembler versions.
- Kill option GPL_MATH_EMULATE from the i386 kernel.
- Add a wsdisplay driver for VS4000/60 and VLC vax machines.
- Add TFTP Option Extension to tftp(1) according to RFC2347.
- In ssh(1), disable tunnel forwarding if no strict host key check and key has changed.
- Make brconfig(8) without arguments behave like "brconfig -a", similar to the ifconfig(8) change.
- Lots of fixes for smg(4) on vax.
- Various LUN-related fixes to the scsi code.
- Improved Yukon init routine for sk(4).
- Remove some scsi ioctls, most importantly those that can reprobe the bus.
- Add ADMtek ADM9511/9513 to dc(4).
- Add ATI SB450 and SiS 966 HD Audio to the azalia(4) driver.
- Prevent segfault when "cpio -o" called without input.
- Make "PermitOpen" in sshd_config(5) take a list of permitted ports.
- Lower the number of TX descriptors in xge(4) from 8k to 2k.
- Introduced new warning to gcc(1) to check for greedy functions using too much local stack storage.
- Initial support for LEDs on vax machines.
- Fix latvian keyboard map for wscons.
- Add blksize option to tftpd(8) according to RFC2348.
- On armish platforms, always assume root is on wd0a.
- Add 'target kvm' support in gdb(1) for alpha.
- Various fixes for the rum(4) devices.
- Dont reject FAT filesystems with more than 255 heads since some USB keychains exist that use 256.
- Add ForceCommand support to sshd_config(5) too. Equivalent to "command=" key option.
- Add support for X11Forwarding, X11DisplayOffset and X11UseLocalhost to the Match keyword in sshd_config(5).
- Add cd_powerhook to the scsi code. Drives may forget they were locked otherwise on PWR_RESUME.
- Add support for ADMTek 983C chipsets.
- Swap inode's associated device number on big-endian architectures when calling checkalias() for ext2fs.
- Various other fixes for rum(4).
- Fix channel 11 RF R4 setting for RF2528 based rum(4) devices.
- On certain TI controllers, clear the SD/MMC bit in the flash media so the SD host takes over.
- Remove arc network support.
- Fix for obvious race in SD/MMC code and simplify the implementation of future host controllers.
- Slight increase in number of jumbo frame slots for ti(4) and bge(4) on sparc64.
- Correct the way SD/MMC hosts are allocated, fixes crashes on controllers with multiple slots.
- Fixes for NVIDIA MCP04 SATA interrupts.
- Initial import of csplit(1).
- Import texinfo 4.8.
- Change in udp receive code which allows IP_RECVIF to get the incoming interface, similar to INET6 code.
- Add "PermitOpen" directive to sshd_config(5). Useful in combination with "Match".
- Fix false positives in vax splasserts for older VAXstations.
- Fix for LX164 Alpha systems where ISA video boards are reported as pci boards with invalid bus/slot numbers. Check for this when deciding which vga attachment to use.
- Add Airprime PC5220 to the list of supported umsm(4) devices.
- Fix for bus_dma on arm platforms, makes em(4) work much better.
- Correctly NULL-terminate strings while parsing fonts and avoid a buffer overflow later in X11 Freetype code.
- Add support for Genesys Logic GL523SM and Global Mixed-mod Technology G781.
- Make mpi(4) use the lowest possible speed while doing inquiry and attach. Some devices like tapes and enclosures dont like being probed at high speeds.
- Have the scsi code catch "Media Removal Prevented" and return EBUSY to userland instead of spewing scsi sense info to the console.
- Add major line for wd* on armish, to allow kernels with root on wd0a to work.
- Allow vax kernel stack tracebacks even when the kernel did not panic.
- Initial PHY driver for IC Plus IP1000A.
- Make sure little endian platforms without alignment requirements do not split multibyte access on usb.
- Add support in sshd_config(5) for "Match" keyword, similar to the "Host" directive in ssh_config(5).
- Add round(3) and roundf(3) from C99 to libm.
- Increase the maximum jumbo frame size on stge(4).
- Make sure pppoe(8) code handle remote error messages not being null terminated.
- Ensure virtual and profiling interval timers are reset in child process after fork().
- Add support for ExitOnForwardFailure to ssh(1) to terminate if we cannot set up all requested dynamic, local and remote port forwardings.
- Make the UDF mounting code hint the kernel where to find the VAT.
- Fix for gcc(1) false positives in the -Wbounded warnings for sscanf(3).
- Only copy the part of environment variable in ssh(1) that we actually use. Prevents ssh(1) from bailing out later on long variables.
- Hook up hw.product sysctl on macppc.
- Remove SpamHaus entry from the spamd.conf(5).
- Make sure the ELF loader can handle executables with all sections marked read-only.
- Disable Speedstep and p4tcc setperf mechanisms on SMP systems.
- Fix potential crash in SOCKS 5 code for ssh(1).
- Fix for zero sized remote files in OpenCVS.
- Make sure fdisk(8) doesn't try to read /usr/mdec/mbr unless the platform actually provides it.
- Support for 68020-based hp300 machines removed.
- Some fixes for ne(4) detection, allowing for unplug of a card when the probe gets stuck without panicing.
- Implement firmware upload to mpi(4). This is the last feature mpt(4) had.
- Make sure modification flags are unset if you undo all the way from a loaded buffer in mg(1).
- Various fixes regarding line counting in mg(1).
- Add support for ATI IXP300 SATA.
- First version of remote code for OpenCVS. Fairly usable, but not recommended for real development yet.
- Fix for in-kernel ddb hangman so it counts correct guesses right.
- Re-sync the em(4) driver to Intel's 6.0.5 FreeBSD driver. Previous version included an unrelated change in the I/O which broke it last time.
- Let mac68k kernels map an empty page at the beginning of the kernel so it can be mapped invalid to catch null pointer references.
- In pf(4), allow rules to point to alternate routing tables and tag packets matching that rule.
- Add options to sftp(1) to enable logging of transactions.
- Add support for arguments to Subsystem commands in sshd_config(5).
- Add VAT support in the UDF file system according to UDF 1.50.
- Revert em(4) back to previous state.
- Make sure the libc crypt(3) routine allocates enough space for gsalt.
- Disable alipm on sparc64 until a spurious bug is found.
- Initial version of nmeaattach(8). Attaches a serial line as a source of a timedelta sensor.
- Sync em(4) driver with Intel's latest FreeBSD driver (6.0.5). Adds support for new chipset revisions found on ESB2 and ICH8 core logic chipsets.
- Back out premature 68060 stuff from mvme68k locore.s.
- Fix for OpenOSPF where ospfd(8) would redistribute bgpd(8) routes into ospf.
- Various fixes for re(4).
- Let isakmpd(8) send out vendor ID announcing isakmpds release version.
- Sync systrace(1) to 1.6d.
- Enable zyd(4) on i386.
- Fix channel frequency/flags in radiotap structs. Channel probes now actually cycle 1-14 instead of just sitting at channel 1.
- Make sure ifconfig(8) also updates the 802.11 address when the user changes the mac address. Replaces old wicontrol(8) -m option.
- Add support for 5750 C2-based bge(4) cards.
- Improvements for the sparc and sparc64 frame buffers.
- Add -K option to vnconfig(8) which uses a salt file and pkcs5 pbkdf2 to create a more secure key.
- Make sure ntpd(8) doesn't write anything to the log until it is daemonized.
- Remove the mpt(4) now that mpi(4) has replaced it.
- Change the iostat output in systat(1) to print kilobytes instead of bytes.
- Make sure the EEPROM reading code on re(4) devices work as expected.
- Make sure amd64 kernels compile cleanly even if DEBUG is defined.
- Fixes for the softupdates queue handling code to avoid endless recursion when forcing worker threads to process worklist items before adding more to the worklist.
- In OpenOSPF, track the uptime of the ospfd(8) itself.
- Add the "nwflag hiddenwid" for hidden SSID mode and "nwflag nobridge" to prevent inter-station communications to ifconfig(8) for 802.11-based devices.
- In hostapd(8), add new event rules to match optional elements like signal percentage, transmit rate and channel frequency.
- Remove jumbo frame capabilities from the 5714-based bge(4) devices for now.
- In systat(1), replace the iostat view to something that actually looks like the ifstat view.
- Add code to ifconfig(8) to support scanning and node listing for 802.11 devices. Also allow it to print BSSID.
- Fix in re(4) for RX path on architectures which require strict alignments and workarounds for reception of jumbo frames.
- Add support for RealTek 8169SC and 8101E based network cards.
- Upgrade OpenSSL to 0.9.7j.
- Add support for RSSI max reporting in RFMD 2958 based atu(4) devices.
- Pass process credentials to VOP_READ making it possible to load firmwares over NFS.
- Enable mpi(4) on SGI.
- On sparc64, always skip the offending instruction on a fpu fault.
- Enable wide targets on FAS366 esp(4) controllers.
- Add Sun type 7 keyboards to detected usb devices and try to see what country code they have selected..
- Have ntpd(8) reset adjtime on startup, otherwise it will overcompensate and produce confusing log entries.
- Initial import of bcx(4), driver for Broadcom NetXtreme II gigabit cards for PCI-X and PCIe. Still in development.
- wicontrol(8) goes away, ifconfig(8) will replace the lost functionality.
- Enable ne(4) on amd64 platforms.
- Allow Xorg to run without privileges when using the wsfb driver.
- Import support for Promise PDC205xx based SATA controllers.
- Set the RSSI max value in ath(4) and add code to allow hostapd(8) and tcpdump(8) to print those values.
- Let ifconfig print out signal quality for 802.11 devices if those have reported the RSSI max value.
- Add an optional max_rssi attribute to the 80211com structure so it can be exported to userland. Used to control signal strength.
- Ext2fs removed from RAMDISKC due to space constraints.
- Add Promise PDC40518 SATA-II to the list of supported devices.
- Endian fixes for sk(4), allows those cards to work on macppc and sparc64.
- Let ntpd(8) save computed clock drift and use it on next startup.
- Add CX4-based hardware to list of supported ixgb(4) devices.
- Add BCM5703 B0 to the list of supported bge(4) devices.
- Enable mpi(4) on alpha too.
- Allow init(8) to accept SIGUSR2 for mandatory powerdowns.
- Improvements in the UDF file system so it can find virtual partition maps.
- Add support on sparc64 to power down nicely using the power button on the front panel.
- Initial code to handle the zyd(4) driver for Zydas ZD1211 802.11 devices.
- Avoid a race in ntpd(8) by installing SIGCHLD handler before fork().
- Add media types for 10GE.
- Make sure the nmea(4) driver stores the timestamp events early so other events don't overwrite it.
- Fix for zaurus suspend problem introduced earlier.
- Fix memory reporting on 2+ GB alphas.
- Added kvm routines for vax platform so "target kvm" works in gdb for kernel crash dumps.
- Add support for hotspares on mfi(4).
- pcn(4) removed from i386 RAMDISKC install image due to size constraints.
- Improve 802.11b/g interoperability and move towards better compliance with IEEE Std 802.11g-2003.
- Fix a crash in ext2fs code.
- On rt2661 and rt2560 based ral(4) cards, enable packet bursting when operating as as STA.
- Enable udcf(4) on sparc64.
- Add support for equal-cost multipath IP, enable with sysctls.
- Import frequency correction code to ntpd(8).
- Reenable AMRR on wpi(4).
- The FFS code will now refuse to mount a filesystem where the number of frags isn't 1,2,4 or 8, instead of panicing later.
- Make sure mouse and keyboard devices on recent Powerbooks doesn't register twice.
- Implement carp demotion control for OpenBGPD. Sessions can be configured to modify carp demotion counter for an interface group.
- Allow udcf(4) to build without clock_subr.c, which means any usb-enabled arch can attach it.
- Fix for dvmrpd(8) on sparc64 which lets it fetch the interfaces correctly.
- Fix file descriptor leak in isakmpd(8).
- Driver for Philips PCA9532 LCD dimmer on ARM-based Thecus.
- Better logging of refusals and transfers in tftpd(8).
- Initial import of rum(4), a driver for the next generation Ralink Technology USB 802.11 a/b/g wireless network device.
- Add rt2573 microcode to ral(4) driver.
- Add skeleton driver for sparc64 Ultra5 power button. Not useful yet.
- On wpi(4), load firmware block by block since contiguous memory may not be possible, and fix reassociations.
- Make sure the privileged child process of dhclient has "/" for working directory so it wont hold locks on directories like /mnt if started from there.
- Support for multichannel (>3) recording on azalia(4).
- Fix for mpi(4) when running under VMWare since the VMWare mpi code doesn't fill in all the fields when probed.
- Enable mpi(4) on hppa.
- Switch i386 and amd64 to use mpi(4) in favor of mpt(4).
- Add support for the fan controller on Thecus armish machines.
- Add Ricoh rtc support for the armish platform.
- Fix watchdog timeout errors in sk(4).
- Rework Alpha interrupt code.
- Put support for multiple routing tables back in again.
- Add adjfreq(2) call to adjust the clock frequency.
- Make sure iwi(4) ipw(4) and wpi(4) clear the IFF_UP interface flag before shutting down the interface.
- Fix bug in sendmail(8) for deeply nested mime which could crash sendmail queue processing.
- Make sure KerberosV doesn't look for a krb4 keytab.
- Avoid potential deadlock in UDF filesystem code during hash lookup.
- In ssh(1), limit the number of pre-banner chars we will accept.
- Enable NMEA 0183 line discipline by default.
- Initial support for iic buses on sparc64 for UltrasparcIII machines. Still in development.
- Allow SHA2 and AESCTR in IKE for ipsecctl(8).
- Kernel Virtual Mode 8086 support added to i386.
- Disable jumbo frames in vge(4) for now.
- Make sure we return EIO and not something else if cpu fails to alter performance mode for both powernow and p4tcc code paths.
- Some more Enhanced SpeedStep fixes for i386 in regard to unknown EST CPUs.
- Make kernel crash dumps work on vaxes too.
- Move default keying in ipsecctl(8) from AESCTR to AESCBC.
- wdt(4) device removed. watchdog(4) is there instead.
- Small fix for link detection code in bge(4).
- In isakmpd(8), make deletion of SAs on shutdown optional, needed for ipsec failover. Default is to not delete SAs now.
- Fix checking of WEP flags in ral(4).
- In isakmpd(8), support SHA2 for main mode and AESCTR for quick mode encryption.
- Allow setsockopts succeed for any value that is not an increase in size when under mbuf pressure rather than when setting the value to the 4k minimum.
- Simplify carp(4) device detection code in OpenBGPD.
- Add X11 support to the armish platform.
- Add EFI partition types to fdisk(8).
- Let ami(4) report percentages for BGI and rebuild when probed from bioctl(8).
- Make sure ipsecctl(8) allocates enouigh storage for sockaddr_in6.
- On ral(4) devices based on RT2661 chipsets, keep track of the average RSSI to tune radio receive sensitivity.
- Fix for potential NULL pointer dereference in ipsecctl(8) when testing for quick mode lifetimes.
- Fix for fxp(4) on cardbus. Should make it work on macppc.
- On hp300, when killing a process due to parity error, notify syslog. Also be careful with caches when checking for the parity error itself.
- azalia(4) audio driver enabled on i386 and amd64.
- Fixes in FFS softdep code relating to removal of large directories.
- ntpd(8) now compensates old offsets with the amount of adjustment done, avoiding overcompensation.
- Updated a couple of CAs that has been changed in the openssl(1) codebase.
- Teach tip(1) set line disciplines.
- On mips64, when deciding to flush the icache for a page, invoke the TLB-friendly cache function.
- In the re(4) driver, write the words of the mac address in reverse order to fix a MAC problem on some Realtek 8169s.
- In the wpi(4) driver, discard bad Rx frames early, dont start AMRR if we have a fixed rate and a fix to TSF synchronization.
- Allow any user to request the current time adjustment with adjtime(2).
- Fix the issue when ifconfig would create an address after a delete.
- Make aic6360 devices detachable.
- Add com* at cardbus? for macppc.
- Add cloneable device implementation for the filesystem code.
- Make sasyncd(8) fall back correctly with carp preemption enabled. Slight change in syntax for carp interfaces.
- Better logic to find leaf le(4) device on sparc.
- Move the carp(4) demotion counter into the interface group. This adds possibility to define which carp interfaces are demoted together.
- splassert() implemented on vax.
- Fix issue where bge(4) cards claim all interrupts for it's own.
- ppm(4) driver added to sparc64 in order to be able to blink the power switch LED.
- Add powerhooks to ral(4) cards with RT2560 chipsets.
- Reenable the I-cache on UltraSparcIIIs even if HORRID_III_HACK is defined.
- Add xge(4) to macppc and alpha builds.
- Speedup in kernel build with debug enabled while stripping the bsd file.
- Add basic NMEA0183 support as a line discipline. This allows for normal serial communication with the device, but adds a timedelta sensor suitable for ntpd(8).
- Fix for ifconfig(8) where a delete of the last address of an interface could unintentionally re-add 0.0.0.0/0 to it afterwards.
- Add support for flows with port modifiers in ipsecctl(8).
- Various fixes in X11 for several integer overflows and null pointer dereferences in freetype2.
- Add powerhooks for ral(4) devices based on RT2561 chipsets.
- Initial import of dvmprd(8), still in development.
- Enable RX hardware checksum on xge(4) cards and enable xge(4) on i386 and amd64 RAMDISK CD kernels.
- Have cdio(1) test disc size before trying to burn 4Gb onto a CD-R.
- Make transpose in mg(1) undoable.
- Endian fixes in mpi(4) for the ioc_status field.
- Make sure ACPI powerdown hooks are not called unless the init went ok.
- Initial support for Blade 2000 sparc64s, though without cache.
- Have vic(4) stop counting packets that already are accounted for in the general ethernet code.
- In isakmpd(8), make sure phase 1 SAs of active connections stay alive. Fixes DPD breakage.
- Enable mpi(4) on amd64, i386 and macppc.
- In ch(4), add support for reading volume tags, usually barcodes.
- Allow dhcpd(8) to manipulate pf tables on certain lease events.
- Enable wdt(4) on i386 GENERIC.
- Add rewritable blanking and track-at-once burning support to cdio(1).
- Implemented chained scatter gather lists in mpi(4).
- Implemented monitor mode for ipsecctl(8).
- Widen the aperture used for legacy vga space on macppc, needed for mac minis ATI graphics cards.
- Fix crash in netstat/route when ipv6 SA flows are present.
- Handle zero length bulk and interrupt transfers on USB 2.0.
- Bump minimum stack size for pthreads.
- Make sure the UFS filesystem code does not dereference a pointer before NULL check.
- Enable mpi(4) on sparc64.
- Added support for import in OpenCVS.
- Add SAS1064 as a supported mpi(4) device.
- When ntpd(8) backs off polling, tell the user how long until next poll.
- Crash prevention in ip printing tcpdump(8).
- Add support for audio volume keys found on many laptops builtin keyboard.
- Enable lists in the ipsecctl(8) parser.
- Limit the number of scatter gather entries sent to mpi(4) so it fits in the maximum request frame size.
- Implement \$ and \# expansion for PS1 in ksh(1).
- Make sure tcpdump(8) uses snapend when printing captured ipx packets, could possibly print behind the captured packet otherwise. Also fixes for netbios parsing.
- Add proper bounds checks for the CDP and SSDP protocol parsers in tcpdump(8).
- Add support for i80321 "armish" systems.
- Revert the recent NFS change regarding reserved ports.
- Add gssapi-with-mic to PreferredAuthentications on ssh_config(5).
- Implement reliable microtime on i386 smp systems.
- Add support for NVIDIA nForce MCP04 to auich(4).
- Bump mpi(4) to use 64 bit for all dva.
- Add support for remove in OpenCVS.
- Fixes for interrupt handling in the re(4) driver and flush the tx data buffer before handing it over to the hardware. Fixes tx on arm.
- Improved error checking in the MBR read and write routines for fdisk(8).
- Correction in the USB 1.0 status value check code.
- Initial version of sparc64 beeper support.
- Enable SD/MMC on amd64 too.
- Fixed a memory overrun in the ACPI aml parser code.
- Misc fixes to the st(4) subsystem.
- Add support for NFS mounts from non-reserved ports, controlled by sysctl. Default is still to require reserved ports.
- Improved nexthop delete behaviour in OpenBGPD.
- Fix bounds checks in tcpdump(8) when printing tcp to prevent crashes.
- Have OpenBGPD preload and pin nexthop used in filtersets so they are validated when used.
- In tcpdump(8), make sure complete llc frame is captured, rather than the first three bytes.
- Initial ipv6 support for ipsecctl. Not complete yet.
- Add weight to ntpd(8) regarding sensors or servers so one can weigh sensors higher than peers.
- Fix open(2) semantics for ch(4) devices.
- Added support for SD host controllers like Ricoh 5C822, and scsi emulation for SD/MMC memory cards.
- In systrace, allow len ==1 for systrace_fname() so systraced processes can get ENOENT instead of forced EINVAL.
- Fix multicast and broadcast over gre(4) tunnels.
- In pfctl(8), enable adaptive timeouts by default. Explicitly setting the timeouts will override the default.
- hotplugd(8) now includes device id in the event logging.
- In trunk(4) code, check if interfaces are active and UP. Some interfaces report link even if DOWN.
- On st(4) devices, always allow ioctl's to work, even if there is no media loaded.
- Fix potential crash in tcpdump(8) when printing ipv6 frags using -v.
- Fix for National variants of the sis(4) cards when switching to allmulti mode.
- Let dhcpd(8) continue even if an interface lacks a subnet configured in the dhcpd.conf(5), instead that interface will be ignored.
- Add mpi(4), an alternative driver for LSI Logic Fusion MPT controllers, currently supported by mpt(4).
- Rework udcf(4) attach code to create a timedelta sensor immediately, so ntpd(8) doesn't have to check for hotplug events.
- Make libc getcwd(3) use the new __getcwd() system call.
- Add pruning support to both update and checkout commands in OpenCVS.
- Large updates to the OpenCVS code. Basic checkout, update, status, diff and commit works in local mode.
- Make pkg_add(1) exit gracefully if it cannot read from a file handle, which happens when the ssh connection could not be established.
- Add support for Option UMTS 3G+.
- Add support for timedelta sensors in ntpd(8). Only udcf(4) at the moment.
- Enable optional handling of the u-area in pmap if defined. Currently used only on xscale arm.
- In ntpd(8), figure out the refid to send to NTPv3 clients early and store it.
- Add automatic rate control (AMRR) to wpi(4).
- Switch diff3(1) to use mkstemp(), avoiding possible race conditions.
- Fix to allow alpha kernels to compile with gcc3.
- Add support in X11 for Intel 945GM chipset video.
- Fix for incomplete logic in the ipv6 input code.
- On i386, simplify the SpeedStep detection.
- Make sure the kernel compiles without SYSVMSG support.
- Have less(1) remove trailing space in filename completion list.
- Add code to dissect Vlan Query Protocol in tcpdump(8).
- Allow OpenBGPD to request a route refresh from a neighbor if that neighbor has announced route refresh capabilities.
- In pfctl(8), handle the case when member interface groups have no IP address.
- Added support for Myrcicom Z8E.
- Attach TTTech MC322 using the re(4) driver instead of rl(4).
- Add support for RAID 10 and 50 in the mfi(4) bio.
- On USB1.0 shutdowns, make sure we only call the disestablish function if there actually is a shutdown function established.
- On wdc(4), directly invoke the wdc_do_reset() instead of using the reset function pointer, this could cause NULL pointer dereference on some systems.
- Fix a possible endless loop situation in vi(1).
- Remove static buffer usage in ksh(1) and make sure item 0 in an array is set since it might be used in a loop even if unset before.
- On ami(4), allocate ccbs in the attach code instead of preallocating 126 of them in the softc.
- In the tcpdump(8) bootp printout code, fix size checks for bootp packets.
- Recognize Perc4/DC firmware.
- Add BCM5706C and 5708C support.
- Make a copy of the sensor structs in the sysctl code so we don't get a use after free if the sensor is detached while it is being used.
- If packets arrive on inactive failover ports in a trunk(4) setup, do not increase the error count, just silently drop them.
- In the routing code, make sure the kernel doesn't panic if the address family is not found in the rt_tables.
- Add show all pools command to inkernel ddb, listing all pools like vmstat(8) does.
- In FFS code, make sure we dont mark a mount-point Read-Only if we are about to sync it.
- Reimplement Enhanced SpeedStep CPU detection code. Makes a number of VIA C7-M and Pentium-M models work.
- Initial version of thermal zone support in ACPI.
- Enable mfi(4) on i386 and amd64 GENERIC.
- Add a shutdown hook for aue(4) devices to reset and stop the device since they can do weird things on a warm reboot.
- Remove spppcontrol(8) now that ifconfig handles sppp(4).
- Added a duplicate check to carp(4) for our own advertisements, necessary for dumb non simplex interfaces.
- Fix a SEGV in bc(1) when reading erroneous syntax from the command line.
- Enable azalia(4) on i386 and amd64 to get feedback. Still in development.
- Make sure the scsi disk probing tests the state of the disk after spinning it up, and not before in order to handle stuff like empty USB card reader slots.
- Various fixes for wpi(4).
- Implemented support for sppp(4) in ifconfig(8).
- Add network bootloader for AV400 Aviion systems.
- Add https URL support to ftp(1).
- Attach 8139s capable of C+ mode to the re(4) instead of rl(4).
- In the re(4) driver, only allow Jumbo MTU frames if the chipset is 8169.
- Initial implementation of IP Roaming in hostapd(8).
- Add -h option to spamd(8) to override the hostname that is reported in the SMTP banner.
- Permit proto 0 in ipsecctl(8).
- On i386, add support for Intel 945G/GM video chipsets.
- Add xge(4) and clcs(4) to amd64 GENERIC.
- Initial import of wpi(4), a blob-free driver for Intel PRO/Wireless 3945ABG, found in Centrino Duo laptops.
- Let Aviion deal with arbitrary load addresses needed for later boot code.
- Fix for corrupted graphics when using X.Org on some Intel PCI bridges, which misreport themselves as normal bridges when they are transparent.
- Fix for unwanted int to unsigned short conversions in libcurses.
- Add -hex option to the openssl(1) rand command to produce hexadecimal output.
- Add support for the usb-attached MetaGeek Wi-Spy.
- Enable twe(4) on amd64.
- Add ALTQ support to san(4) devices too.
- Make sure hostapd(8) sets the correct timezone before writing to syslog.
- Allow hotspare disk to be used on ami(4) even if the firmware sometimes misreports what SCSI type this device is.
- Fix for pim register corruption in the multicast routing code.
- Handle malloc() failure in df(1).
- Add -x and -X options to kdump(1) to print output in hex.
- Add support for listing interfaces as arguments to dhcpd(8) and traverse the list, ignoring all other interfaces for which no entry was given.
- Make sure zzz(8) fails with a reasonable error message if apmd(8) is not running.
- Added support for the Aviion platform, a Data General m88k based machine.
- Enable alipm(4) on sparc64 again, and sprinkle some bus_space_barriers around the driver to help debugging it.
- Import global ACPI interrupt code to i386 from the amd64 platform.
- Do not allow delay times of 0 in systat(1).
- On i386, avoid estimating pentium mhz too low, resulting in non-monotonic time.
- Add 'link-timeout n' support to dhclient(8). Default remains at 10 seconds.
- Removed unsafe setjmp/longjmp calls in tftpd(8) that were used in signal handlers.
- azalia(4) audio driver added to i386 and amd64 GENERIC, though commented out.
- Endian fix for getservbyport(), which wants network byte order.
- Add bio(4) framework to gdt(4).
- Remove the pool(9) draining hooks. Probably never used in a meaningful way anyhow.
- Fixes for sgi debug code, unbreaks builds with DEBUG defined.
- Move the userland stack on m88k to the top of userland address space.
- Make sure MVME188 machines dont panic if the abort switch is pressed while at ipl 7.
- On m88k, decide whether we are running a 88100 or 88110 by reading the processor identification register instead of basing it on the MVME board number.
- Endian fix for pfsync code.
- Fix for pfkey, allowing sasyncd(8) to reliably set up pfkey promiscuous mode.
- Some pmap fixes for the m88k platform regarding obio on 88100-based system.
- Added various ethernet devices to Alpha GENERIC and RAMDISKBIG kernel images.
- Restore ALTQ support which was lost on de(4).
- Fix two SEGVs in lint(1).
- Make sure ssh(1) and isakmpd(8) also check for such degenerate Diffie-Hellman public exponents.
- Backport checks for degenerate Diffie-Hellman public exponents from OpenSSL 0.9.8a.
- Remove potential overflow in mg(1) when doing file I/O and len == 0.
- Add 88110 instruction support to the m88k in-kernel debugger.
- Initial code in ld.so(1) for prebinding binaries without compromising the security of randomized library linking. Still under development.
- Make sure OpenRCS diff3 respects the TMPDIR environment variable.
- Add Neterion Xframe-II 10GbE id to xge(4) pcidevs.
- In mg(1), have each buffer store their own working directory.
- Plug possible memory leak if AMD K7 powernow init fails.
- Repair creation of subanchors with pfctl(8), and loading of subanchors from file.
- Add Rx/Tx FIFO flush support to Yukon based sk(4) cards.
- Fix for broken mpbioses on VT8237 and nForce4 chipsets. This fixes interrupt routing for several integrated devices on those chipsets in GENERIC.MP.
- Increase Tx ring count on nfe(4) from 64 to 256.
- In pf.conf(5), add support for "tagged {}" lists.
- Make sure iwi(4) get initialized correctly, fixes a firmware panic if the same WEP key is entered multiple times.
- Initial import of the Neterion Xframe-I 10GigE xge(4) driver.
- On macppc, add support for fxp(4) and rl(4) on Cardbus on GENERIC and RAMDISK kernels, and add dc(4) and wdc(4) to RAMDISK too.
- Added a driver for the Silicon Laboratories CP2101/CP2102 usb serial adapters.
- On macppc, allow the nv X driver to mmap() as on i386 when running in mapped mode and allowaperture is non-zero.
- Added support for the National Instruments PCI-GPIB card.
- Cleanup of pdisk(8) for mac68k.
- Implemented separate pmap for PAE i386 machines, allows for support for machines with more than 4G RAM. Not enabled by default.
- Support for HBG and DCF77 time sources in the udcf(4) device.
- Added support for the ICH7-M DH controller in RAID mode.
- Fixed NULL pointer access in error path for kernel routing code.
- Initial framework for the azalia(4) audio driver added.
- Added support for ATI IXP 600 IDE/SATA controllers.
- Rewrite setjmp()/alarm() code in tftpd(8) with poll() to avoid signal races.
- Remove virtual tunnel support from the mrouting code. It breaks multicast on gif(4) interfaces and a real gif(4)-tunnel is more manageable than a multicast tunnel.
- Make sure ssh(1) tries only once to open private keys, and not ask for passphrases if the keys cant be opened later anyhow.
- Fix endian issue in netatalk code.
- Fixes in libc regarding malloc() when allocation of the region succeeds but the allocation of a required page dir fail. Could leave holes in the directory list.
- Enable asm profiling for i386 if either PROF or GPROF is enabled.
- Add pci ID for 82801GBM SATA in AHCI mode, found on X60/T60 Thinkpads.
- Endian-safe codepath for AMDtek variants, fixes dc(4) on macppc.
- Add support for Dell Bluetooth 350 USB module.
- Allow Intel SATA to map compatibility interrupts, makes a lot of ICH6M/ICH7M systems work.
- Add support for puc(4) and com on macppc.
- Rework the signal handler code in the privsep part of tcpdump(8).
- Repair MVME147.
- Add support for Adaptec 2130S and 2230SLP raid adapters.
- Make udcf(4) devices work as a SENSOR_TIMEDELTA type sensor.
- Implement outgoing interrupt pipes in uhci USB, makes Creative Voip Blaster work.
- Add support for Encore ENL832-TX-ICNT device in the ste(4) driver.
- Add support for BCM5752 A2 devices in the bge(4) driver.
- Switch ssh(1) pipe usage to socketpairs. Saves a file descriptor and makes userland ppp over ssh work.
- Add a few more integrated devices on nForce4 amd64s.
- Add support for Belkin F5D6020 802.11b Wlan card.
- Fix off-by-one read in yacc(1).
- Fix some signal races in fsck_ext2fs(8) too.
- Initial driver to decode DCF77 time codes.
- Add hostname resolving and support for interface groups to ipsec.conf(5) parser.
- Make sure we dont leak other processes execution history when restoring fpu state on the i386 and amd64 platforms.
- Backport a gcc(1) bug fix for code generation which appeared when compiling htons() with -march=686 -O2 flags.
- Added some more Intel GigE pci IDs.
- In the kernel watchdog, make sure the parameters are reset to defaults when it is shut down.
- Move bignum functions in ssh(1) to a file of its own, now sftp(1) and the sftp server don't have to link with libcrypto.
- Display logical disk status in sysctl for ami(4). Now you can use sensorsd to monitor disks on ami.
- Better detection between K7 powernow and Cool-n-Quiet for amd cpus.
- Several signal race issues in fsck_ffs(8).
- Fix two NULL dereferences in the sk(4) driver.
- Support for ICH7R SATA in pciide(4).
- Preliminary support for Colorgraphic VoyagerVGA pcmcia frame buffer.
- Add support for Asus MyPal A730 PocketPC.
- Make sure awk(1) handles / inside []'s correctly.
- Fix problem in ssh(1) when input buffers becomes too large.
- Produce missing caps lock key event on hp300.
- Add support for the VT8363 PCI-PCI bridge found in Sony Vaio FS201s.
- Fix for Kerberos5 code on static arches.
- On i386 and amd64, enable 3DNow! and SSE support in X11, and enable the via driver for amd64.
- Added another nForce4 ISA bridge PCI device ID to amd64.
- Various fixes for AMD K7-powernow code on i386.
- Add support for non-RAID part of the ITExpress IT8212F.
- Remove all of the i386 DUMMY_NOPS code for good.
- Initial support for HP98705 "Tigershark" framebuffer for hp300. Text console and X11 are supported.
- Update Heimdal Kerberos 5 to release 0.7.2.
- Show cpu time spent in interrupts on the vmstat page in systat(1).
- File descriptor leak in httpd/mod_ssl code fixed.
- Various fixes and updates for both OpenCVS and OpenRCS.
- Free up memory in axe(4) if attach fails.
- In ipsecctl(8), add support for "local" to ike rules for multihomed machines.
- Fix signal races in ping(8) and ping6(8).
- Rewrite attribute changing code to remove+add instead. Fixes "equal attributes encountered" errors in bgpd(8).
- Make rthreads code compile again after header change.
- Fixes for spl and interrupt counting in em(4) when dealing with half-completed work queues.
- Fix some memory leaks in ldconfig(8).
- In the vlan interface code, make sure we copy the baudrate (link speed) on link state changes from the parent device.
- Switch to OpenRCS, even though some features are still lacking.
- Add Intrepid 2 GMAC support to the gem(4) driver.
- Added some wi(4) reset code since some cards freeze up sometimes.
- On macppc and sparc64, if allowaperture=0 only allow mmap()-ing of the framebuffer memory in WSDISPLAYIO_MODE_DUMBFB mode.
- Do not honor the address split scheme on MVME188 systems with more than two CMMUs per CPU when operating on caches, since it doesn't work.
- Fix for sparc64 bus attachment code to better detect which driver to attach. Allows gem(4) netbooted machines to detect their root device.
- Don't panic in apic code for i386 if an interrupt isn't shareable and is already taken.
- Plug simple memory leaks in pfctl(8) and lpr(1) too.
- Fix memory leak on fork failure in mount(8).
- Improvements for the umidi(4) USB MIDI interface.
- Add workaround for nfe(4) mbuf leak.
- Superblock sanity checking code added to ffs vfsops.
- Update xterm(1) to version 211.
- Revert pci access method for xfree code on amd64 since the /dev/pci could cause problems on some machines.
- Allow lists inside lists in pf.conf(5).
- When running emulated binaries, release kernel lock after the exit hook to protect possible free()s.
- Fix a process datasize leak in UVM when doing mmap with MAP_FIXED.
- Add skeleton driver for mfi(4) MegaRAID SAS cards.
- Fixes for floating point state saving in threads code for i386.
- Fix for bgpctl(8) where it would show only the first announced prefix when doing "show rib det source-as #nr".
- Fix a bug in mg(1) when writing buffers sharing the same basename.
- Update to BIND 9.3.2.
- Endian fix for an(4), now also enabled on hppa, hppa64 and macppc.
- Reap child processes asynchronously in hotplugd(8).
- Add power hook to bge(4).
- Enable tqphy(4) for macppc found on xl(4) cards.
- Add "set nexthop self" support to bgpd(8).
- DMA fixes for xscale ARM when changing DMA completion handler.
- Remove bundled dbm and regex lib from httpd, it picks up our dbm anyhow.
- Installation fixes for Zaurus C3200, avoids a known problem in zbsdmod.
- When parsing file urls in ftp(1), ignore http_proxy.
- Don't nul-terminate GSSAPI buffers in ssh(1).
- Fix interactivity issues in ami(4) and remove a busywait.
- Prevent panic in iwi(4) if pre-3.0 firmware is being loaded.
- Fix for timestamp setting problem over nfs.
- Fix for ctrl-W in mg(1).
- Lots of small non-invasive changes in preparation for FFS2.
- Allow 32-bit session-ids in ssh(1) now that sha256 kex is used.
- Make som channels in ICH IDE/SATA controllers native to be able to use the maximum number of disks.
- Fix for sensorsd(8), it will now probe all sensors even if gaps are found in the list.
- Plug huge memleak in libc regarding how the semantics around telldir() works.
- Fix memory leak in sasyncd(8).
- Add sysctl to retrieve routing table statistics.
- Small steps taken in preparation for FFS2.
- Add support for GSM modem SIEMENS ES75 on usb.
- Add iop(4) to amd64 GENERIC.
- In ssh(1), make sure we handle truncated files while converting keys, and also prevent duplicate log messages when privsep=yes.
- Add percent escape chars to the IdentityFile option in ssh_config(5).
- When fsck(8) asks y/n, it will now also accept "F" to force yes to all further questions.
- In ftp(1), allow fetching zero-length files, and also fix progress-meter for >2G files.
- Fix it(4) detection on newer ASUS boards for i386 and amd64.
- Fix ami(4) sync_cache command so it can run asynchronously.
- Enable MPI-401 MIDI UART on amd64 GENERIC.
- Imported Perl 5.8.8.
- Add simple printer for IEEE 802.1AB LLDP to tcpdump(8).
- Add Intel 82801GB SATA support to pciide(4).
- Add more space to memory columns in the vm page for systat(1).
- Sync em(4) with Intel's FreeBSD em (5.1.5) driver, adds support for 82563 PCI Express chipsets and a few fixes.
- Added some HiFn, Safenet and Bluesteel crypto cards to GENERIC on alpha.
- Rework of iwi(4), enabling s/w antenna diversity, upgrade firmware to v3.0 layout and reworked rings allocation.
- Sync ixgb(4) with Intel's FreeBSD ixgb (5.0.1) driver.
- Add support for Realtek RT8111B PCI Express Gbit MAC.
- In pstat(8), check for syncer vnodes before calling fs specific routines, so NFS and ext2fs gets correctly handled.
- Split serial on macppc so zs devices are ttya and tty00 supports com(4) devices.
- Sync libpcap with tcpdump.org's 0.9 APIs.
- Added support for accessing 8-bit ISA I/O through the gpio(4) framework. Makes Acrosser AR-B1662 boards work.
- Added support for Falcom Samba GPRS modem in the uftdi(4) driver.
- Don't allow gpio ping change if /dev/gpio was opened in read-only mode.
- Per-file I/O accounting is now possible to show in fstat(1).
- Allow bpf(4) to ignore packets based on their direction.
- Fixup IP checksum when modifying IP header fields in pf(4).
- Fixes for the receive filter in sis(4) when enabling/disabling promiscuous mode.
- On dc(1), show error message if the argument is a directory.
- Added support for umsm(4) serial interface for Qualcomm MSM EVDO modems. Enabled on amd64, hppa, hppa64, i386 and macppc.
- Turn on PCI parity checking on sparc64, and set the PCI latency timer for bus master devices.
- Fix for adb keyboard caps key problem, if remapped to a dead key.
- Fix for broken amd64 mpbios'es on VT8237 and nForce4 chipsets. Fixes interrupt routing for SATA on those when running GENERIC.MP.
- Various fixes from the NetBSD coverity runs.
- Add support for macros in ipsec.conf(5).
- Change the way bgpd(8) selects nexthops. Required for complex setups e.g. where an additional IGP is running.
- Add a new SHOW imsg to send optional attributes in bgpd(8) - used to show communities in bgpctl output.
- Rework signal handling in ftp-proxy(8).
- Add a helper pci_find_device() function to pci(4) to enumerate the PCI bus.
- Plug a memory leak and fix a possible crash in usbhid(3).
- In the /etc/security script, deal with the fact that some shell scripts may contain binary characters.
- In pkg_add(1), fix an infinite loop in case of an error in the archive header.
- Implement the Michael MIC as defined in IEEE 802.11i for TKIP. This generates a weak 64-bit digest protected by an additional key - required for future IEEE 802.11i/WPA support.
- Better error checking in ifconfig(8).
- Make ural(4) read and show real MAC/BBP rev as determined, instead of the hardcoded RT2570.
- Fix panic on iMac G5 (iSight) in smu(4) due to a missing kernel unlock.
- Fix an out-of-bounds memory access in ksh(1).
- In m4(1), add limited support for format builtin in gnu-m4 mode. This is used by some autoconf scripts.
- Allow the wscons(4) wsfb X server to run without the need for xf86(4) aperture to be set.
- In the X server, use /dev/pci to access the PCI configuration space on i386 and amd64, instead of directly accessing the addresses from userland.
- A large amount of memory leak plugging in various system utilities inspired by Coverity reports, as well as ruling out of hypothetical NULL dereferences.
- Many changes in ssh(1) to be more careful with types.
- In isakmpd(8), make sure the command fifo is ready before returning - fixes a startup race when interacting with ipsecctl(8).
- Improve lint(1) to reduce more false positives and thus make the output more useful.
- Fix some memory leaks in ssh's gssapi(3) code.
- Change powerpc's BLKDEV_IOSIZE down to 2KB like on all other platforms.
- Recognize more Intel pci(4) devices as found on the Acer Travelmate 3012WTMi laptop.
- Plug memory leak and potential NULL dereference in hash(3).
- Make asa(1) use getopt(3).
- In init(8), plug potential memory leaks in error path.
- Teach pcn(4) to recognize Am79c976 and Am79c978.
- On sparc64, use OFW nodes to enumerate the PCI bus. This makes detection of devices hiding behind bridges more reliable.
- Extend PCI bus enumeration to allow architecture-specific hooks.
- Plug potential memory and descriptor leaks in ssh(1) and sftp(1).
- Provide new amd64-specific implementation of exp(3).
- Plug memory leak in revnetgroup(8).
- Teach it(4) to probe IT8712F-a and SiS950 devices.
- Rework tty handling in bc(1) so that things get cleaned up reliably on abort.
- Place reasonable bounds on memory usage of bufferevents in libevent.
- Provide implementations of the C99 math functions trunc(3) and truncf(3).
- Plug memory leaks and a possible NULL dereference in systrace(1).
- Make isadma happier on alpha by honoring the 'high' parameter in _bus_dmamem_alloc_range.
- Change implementation of delay(9) on alpha to use rpcc instruction.
- Plug memory leak in usbhid(3).
- Clean up fts(3) leak in cp(1).
- Clean up file descriptor leak in authpf(8).
- Make ptrace(2) single-stepping work on alpha.
- Fix an off-by-one error in csh(1) error handling.
- Simplify descriptor cleanup in ksh(1), helping us to avoid any descriptor leaks that may have existed.
- Fix an off-by-one error in ping6(8).
- Drastically simplify handling of SCSI vs. AMI commands in the ami(4) driver. This is the first step to supporting more RAID features.
- Correctly deal with -I (interactive) option in rcs(1) co command.
- Improve clarity of messages in SGI installer.
- Fix some off-by-one errors in talkd(8).
- Delint various size_t/ssize_t/int mismatches in cvs(1) and rcs(1).
- In rcs(1), set exit code when checking out a specific revisions fails.
- Fix value of hw.cpuspeed sysctl on certain Pentium III models (early stepping 1 Coppermines).
- Turn EXTENT_ALIGN into an inline function, shaves 100 bytes from RISC kernels.
- Remove unused code from siop(4), shaves half a KB from kernel.
- Fix mincore(2) to pass byte count to UVM functions, rather than page count.
- In lmc(4) driver, consolidate error message array to shrink kernel by about 1KB.
- Clean up and ANSIfy tip(1) code and fix some latent bugs.
- Make rcs(1) co command honor the -q option.
- In scp(1), display error messages even when remote stdout is not initialized.
- Optimize tun(4) by using mbuf clusters instead of mbuf chains for large packets.
- Fix some "reload" bugs in ifstated(8).
- Make AMD PowerNow speed control stuff print nice dmesg lines.
- Fix a quirk of ^C in vi(1).
- Add sysctl_int_lower() API, consequence of which is that root can now lower the machdep.allowaperture variable without rebooting.
- Some i386 laptops need more physmem segments (consequence of the kernel moving up by 1MB).
- A few small fixes to bgpd(8).
- If poll(2) is interrupted, remember to clear the pollfds array.
- The zaurus no longer needs to use the xf86(4) aperture.
- Fix %A printing in spamd(8).
- Move parsing of wsconctl.conf(5) to earlier in the /etc/rc sequence, so that people can use their remapped control key earlier.
- Install apm(4) utilities on amd64 so that the hw.setperf sysctl for CPU speed can be conveniently manipulated through them.
- Fix BN_is_prime(3) to realize that 2 is prime. It took the OpenSSL people what... 8 years to fix this?
- Fix amd64's exp(3) implementation which was broken in certain cases by switching to the C version.
- Repair anonymous logins without passwords in ftpd(8).
- Some churn in ami(4), heading towards improvements...
- In pf(4), implement Unicast Reverse Path Forwarding (uRPF) which optionally verifies that a packet is received on the interface that holds the route back to the packet's source address. This makes it an automatic ingress filter, but only when routing is fully symmetric.
- Add support for Sony CLIE PEG-T625 (and other new) devices in uvisor(4).
- Move the i386 kernel up 1MB in virtual address space to provide more space for isadma(4).
- Fix pkg_add(1) to work with stdin input (pkg_add -).
- On alpha, protect sgmap extents with splvm().
- On i386, use pause instructions during IPI lock spinning.
- Slightly improve the situation with multi-function pcmcia(4) cards.
- Make compress(1)/gzip(1) work without /dev/stdin and /dev/stdout.
- In ssh(1), allow configuration directives to contain spaces, if wrapped in double quotes.
- In ssh(1), set TCP_NODELAY flag for all connections not just interactive ones - fixes poor network performance and protocol stalls under some network conditions.
- On the i386, fix userldt handling (consequence of recent i386 W^X changes) which causes a crash if a LDT-using process does an execve().
- In mount(8), add -s flag which skips a mount if the file system is already mounted, handy for MFS partitions.
- Switch from using PCI le(4) driver to pcn(4).
- Handle emc6d100 and emc6w201 in the i2c adt(4) driver.
- Fix local arbitrary command execution vulnerability on local/local and remote/remote copies in rcp(1).
- Change the install script question 'Do you expect to run the X Window System?' to have 'no' as a default answer, to prevent people accidentially hitting return on the prompt on machines that don't need it.
- Fix a few subtle bugs in vi(1).
- Support IAPP pre-standard port 2313 in hostapd(8).
- A output formatting bug (missing space in command line printing) was recently introduced into ps(1).
- Grok %z in the format string warning code for the older gcc we have in the tree.
- Foil potential timing attacks in login by using bcrypt(3) on non-existent users as well, instead of DES.
- On amd64, properly route ISA interrupts through ioapics.
- Some obscure fixes to ospfd(8).
- Support the iic(4) controller found in the amdpm(4).
- Now AMD PowerNow K8 support works on i386 and amd64, so that hw.setperf can adjust the processor speed.
- Serial port support on the Zaurus was busted by the recent clock changes because of an interrupt mask mistake.
- Improve the iic(4) scanning for ADM1030 and ADM1031 devices (supported by admtemp(4)).
- In grep(1) do proper handling for patterns which include newlines.
- Using pcibios(4) information, properly initialize the VIA VT8233, and support SiS 962.
- Fix the long delay probe for adb(4) on some machines (in particular, the powerpc Mac Mini).
- In bgpd(8), group block parameters were reset to default (duh, that is wrong).
- Repair a really stupid (and old) bug in spanning-tree bridge(4) support.
- Make pkg_add(1) work with scp (see PKG_PATH).
- In ssh(1), implement diffie-hellman-group-exchange-sha256 key exchange method (which is now default when client/server support it).
- In the com(4) driver, add support for the cardbus(4) Sony Ericsson GC89 UMTS/GPRS/GSM modem.
- For all you Road Warriors in need of more dynamic IPSEC, in ipsecctl(8), add dynamic option ("ike dynamic esp") which uses the system hostname as source identifier instead of the IP address. See ipsec.conf(5).
- Repair UDMA for the Via VT8237 in the pciide(4) driver.
- Fix a mmap(2) related panic.
- Fit mt(1) and eject(1) onto all the install media.
- Add some acpi(4) debugging support to ddb(4).
- Move remaining bits of the network stack to use splnet instead of splimp, see spl(9).
- OpenCVS development continues with new energy.
- Stop trying to allocate rotationally optimal blocks in FFS, since modern disks don't care.
- Repair stdin handling in sdiff(1).
- tun(4) is not a point-to-point interface, but a tunnel.
- Fix a buffer cache race in vnd(4).
- Fix a memory leak in cap_mkdb(1).
- Fix a bug in tar(1) that cut short our filename lengths by one miserable character.
- Initial 1-Wire onewire(4) bus support, including drivers for gpioow(4), owid(4), owtemp(4).
- More churn in pkg_add(1).
- In X11, support 10-bit color DAC video cards on Alpha TGA2 cards.
- Do underlining in the zx(4) sparc/sparc64 video driver.
- Fix an old bug on ARM-based machines, that would very rarely and very randomly deliver a SIGBUS that was not expected or wanted by anyone.
- Display cpu states right away (not delayed) in top(1).
- art(4) now works on sparc64 as well.
- In ami(4), support a few more re-licensed parts by ACER, FSC, Intel and NEC.
- Support Compaq Flash Socket CF 10/100 card in ne(4).
- Support National Semiconductor DP83847 DsPHYTER PHY in nsphyter(4).
- Some repair to vic(4) so that it might soon work for VMWARE.
- Handle National Semicondutor lm96000 in the i2c adt(4) driver.