summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
* More issues with pre-blocking:rrs2008-10-271-4/+10
| | | | | | | | | | a) Need for EEOR mode to take the min of the socket buffer size and the add more threshold, otherwise if you are so silly as to set a send buf size less than the add-more you could block forever in eeor mode. b) We were incorrectly using the sysctl vs the calculated value. This causes us to block forever if the addmore theshold is larger than then the socket buffer size.
* Two inter-related bugs.rrs2008-10-271-2/+8
| | | | | | | | | | | | | | - If we send EXACTLY the size left in the send buffer and then send again, we end up with exactly 0 bytes and don't hit the pre-block code to wait for more space. - If we fall into the loop with our max_len == 0 (the bug above) we then call in to copy out the data, setup the length of the waiting to transmit data to 0 and call the mbuf copy routine which 0 indicates copy all the data to the mbuf chain.. which it does. This then leaves a "stuck" message on the stream queue with its size exactly 0 bytes but all the data there and thus nothing left in the uio structure. We then reach a stuck forever state never being able to send data.
* Get rid of ifdef for vimage on version 8 comparison. Now therrs2008-10-271-2/+0
| | | | scrubbing program properly takes care of this.
* Invariants changes that make more sense.rrs2008-10-272-5/+6
|
* Add TrustedBSD credit to new ugidfw_internal.h file.rwatson2008-10-271-0/+2
|
* Break mac_bsdextended.c out into multiple .c files, with the base accessrwatson2008-10-275-1452/+153
| | | | | | | control logic and policy registration remaining in that file, and access control checks broken out into other files by class of check. Obtained from: TrustedBSD Project
* Copy mac_bsdextended.c to two object-specific files as a prototype for howrwatson2008-10-272-0/+1536
| | | | | | modularize MAC policy layout. Obtained from: TrustedBSD Project
* Remove unused consdev structure fields.ed2008-10-274-8/+2
| | | | | | The cn_unit and cn_tp fields don't seem to be used anywhere. Some drivers set them, while others don't. Just remove them, in an attempt to make our consdev code a little easier to understand.
* Only call dma.unload() if initialized.sos2008-10-271-1/+2
| | | | Reported by: Stanislav Sedov
* Improved IDE HDD geometry adjustment. Previous code didn't work withkato2008-10-272-7/+112
| | | | | | | | | | | | | | | certain ATA-6 drives including CF cards. The IDE geometry of the PC98 is calculated from the drive capacity. In addition to the algorithm in NEC BIOS, a variety of algorithms are provided by 3'rd party boards and BIOS hacks. This change has implemented the three algorithms: IDE BIOS compatible mode, SCSI BIOS compatible mode and same way as the previous version. The tunable machdep.ad_geom_method selects the algorithm. I have been using this change for a year with CF cards. Reminded by: nyan
* vm_pnames should be "const char *const[]".sobomax2008-10-271-1/+1
| | | | Submitted by: Christoph Mallon
* vm_pnames has no reason to be global.sobomax2008-10-271-1/+1
| | | | MFC after: 2 weeks
* Default HZ value (1,000) on i386/amd64 is not very virtual machine friendly.sobomax2008-10-271-1/+39
| | | | | | | | | | | | | | | | Due to the nature of the beast it causes lot of unproductive overhead. This is especially bad when running SMP kernel on VMWare with several virtual processors - idle FreeBSD guest with SMP kernel takes 150% host CPU time on my dual-core MacBook Pro when I am enabling two virtual CPUs, making even host not very usable. Detect when we are running in the sandbox and reduce HZ to 10 (can be adjusted via VM_HZ in the kernel config) in such cases. This brings host CPU usage of idle FreeBSD/SMP on two virtual processors down to 10%. Detect most popular VM platforms out there - VMWare, Parallels, VirtualBox and VirtualPC. MFC after: 2 weeks
* Add support for kernel profiling for both AIM and BookE.marcel2008-10-273-51/+83
| | | | Obtained from: Juniper Networks, Inc (BookE support).
* Remove unused declarations (interrupt_vector_{base|top}).marcel2008-10-271-3/+0
|
* Declare btext and etext. Needed by sys/kern/subr_prof.c formarcel2008-10-271-0/+3
| | | | for kernel profiling.
* Bring Kauai ATA driver in line with Macio ATA by reading the PIO config regnwhitehorn2008-10-271-6/+2
| | | | | | | | | | | | to set the initial PIO mode instead of assuming PIO4. There are still a few nagging issues: - There are some problems with 64 K DMA transfers waiting on lower level changes. - ATAPI DMA is broken on Marcel's Mac Mini because we need an ATA SELECT hook propagated up to individual drivers for hardware without timing registers for each ATA channel.
* - Add stream bus_space operations.stas2008-10-261-0/+36
| | | | | | Reviewed by: cognet Approved by: kib (mentor, implicit) MFC after: 1 week
* - Add a missing NULL-pointer check.stas2008-10-261-2/+4
| | | | | | Reviewed by: cognet Approved by: kib (mentor, implicit) MFC after: 3 days
* Implement MAC policy support for IPv6 fragment reassembly queues,rwatson2008-10-265-5/+275
| | | | | | modeled on IPv4 fragment reassembly queue support. Obtained from: TrustedBSD Project
* Add a MAC label, MAC Framework, and MAC policy entry points for IPv6rwatson2008-10-265-5/+160
| | | | | | | | | | | fragment reassembly queues. This allows policies to label reassembly queues, perform access control checks when matching fragments to a queue, update a queue label when fragments are matched, and label the resulting reassembled datagram. Obtained from: TrustedBSD Project
* In both dropwithreset paths in tcp_input.c, drop the tcbinfo lockrwatson2008-10-261-18/+4
| | | | | | | | sooner to decomplicate locking and eliminate the need for a rather chatty comment about why we have to handle the global lock in a special way for the benefit of ipfw and pf cred rules. MFC after: 3 days
* o re-enable a lock assertsam2008-10-261-5/+5
| | | | o while here reformat a comment to sync w/ other repot's
* Fix joining an 11b BSS: scanning is normally done using 11g channelssam2008-10-261-17/+72
| | | | | | | | | | | | | | (unless explicitly locked to mode 11b) so when we join the bss the channel attached to the scan cache entry may need to be demoted. o demote to 11b if the ap is advertising 11b rates o skip the ap if it's 11b but we're locked to 11g (could consider this advisory but for now treat it as mandatory) o handle an odd edge case, if there is a fixed transmit rate for 11g then the rate check against the 11b ap will fail, try to demote to 11b and retry the rate check Reviewed by: sephe, thompsa
* Add ADB support. This provides support for the external ADB bus on the PowerMacnwhitehorn2008-10-2617-0/+3830
| | | | | | | | G3 as well as the internal ADB keyboard and mice in PowerBooks and iBooks. This also brings in Mac GPIO support, for which we should eventually have a better interface. Obtained from: NetBSD (CUDA and PMU drivers)
* Remove endearing but syntactically unnecessary "return;" statementsrwatson2008-10-263-6/+0
| | | | | | directly before the final closeing brackets of some TCP functions. MFC after: 3 days
* Style changes only:bz2008-10-261-44/+45
| | | | | | | | | | - Consistently add parentheses to return statements. - Use NULL instead of 0 when comparing pointers, also avoiding unnecessary casts. - Do not use pointers as booleans. Reviewed by: rwatson (earlier version) MFC after: 2 months
* Fix division by zero panic if kern.hz less than 32.sobomax2008-10-262-2/+10
| | | | MFC after: 1 day
* - Import macros used in gmirror for printing gvinum debug messages and makinglulf2008-10-2610-72/+94
| | | | | | the output more standardized. - Add a sysctl to set the verbosity of the debug messages. - While there, fixup typos and wording in the messages.
* New ap-side power save implementation; the main change is to allow driverssam2008-10-268-113/+285
| | | | | to queue frames previously encapsulated on a separate high priority list that is dispatched before the unencapsulated frames (to preserve order).
* move decl up before first usesam2008-10-261-1/+1
|
* introduce M_ENCAP flag to mark packets encapsulated w/ an 802.11 headersam2008-10-262-1/+8
|
* fix comment that belonged w/ previous commitsam2008-10-261-3/+2
|
* change ieee80211_send_probereq to supply raw xmit parameters so it cansam2008-10-261-2/+19
| | | | control how frames are handled by the driver
* change ieee80211_send_nulldata to send a QoS Null Data frame to asam2008-10-261-14/+51
| | | | | | QoS-enabled station; this makes inactivity handling follow the spec as previously it would probe inactive stations w/ a Null Data (no QoS) data frame
* o change ieee80211_mgmt_output to take a raw xmit parameters block sosam2008-10-264-30/+45
| | | | | | | | | the net80211 layer has complete control over the handling of mgt frames (in particular, the ac, tx rate, and retry count); this also allows us to purge the M_LINK0 flag that was attached to mbufs to mark them as needing encryption for shared key auth o change ieee80211_send_setup to take a tid parameter so it can be used to setup QoS frames
* fix old merge botch that causes gaps in the tx seq# space for QoS framessam2008-10-261-1/+0
|
* Sync BAR frame handling with out of tree work:sam2008-10-253-56/+190
| | | | | | | | | | | | | o correct BAR frame construction for AMPDU o retransmit BAR frames until ACK'd or timeout (use tunables to control behaviour, default is very aggressive) o defer seq# update until BAR frame is ACK'd o add BAR response handling callback for driver to interpose and push new state to device or push pending aggregates While here also: o add backpointer to node in the per-tid tx aggregation data structure o move ampdu tx state setup/teardown work to separate functions
* set erp flag and tx parameter state when joining a bss; this allowssam2008-10-251-0/+3
| | | | drivers to use ni_txparms
* honor IEEE80211_CLONE_NOBEACONS for type of vap, not just a sta mode vapsam2008-10-251-5/+3
|
* improve inactivity handling debug supportsam2008-10-251-5/+29
|
* use a private mgt frame recv handler for ahdemo mode instead of an inlinesam2008-10-251-4/+24
| | | | | test in the adhoc mode rx path so classes derived from ahdemo mode can override the default behaviour
* expand captured ie's before calling match_bss so they can potentiallysam2008-10-251-1/+1
| | | | be used in the process
* change ieee80211_sta_join to take an explicit channel instead ofsam2008-10-254-7/+18
| | | | | using the value in the scan parameters; this will be used to fix issues with 11b operation
* o add support for ifconfig wlanX mode foosam2008-10-251-62/+8
| | | | | | o yank useless code for setting fixed rate through media opts: this mechanism didn't scale to HT rates and couldn't handle multiple bands; fixed tx rates are set with the IEEE80211_IOC_TXPARAMS ioctl
* add debug msgssam2008-10-251-5/+28
|
* fix associd check for adhoc modesam2008-10-251-18/+13
| | | | Reviewed by: jhay
* ignore IEEE80211_ELEMID_PWRCNSTR when parsing beacon/probe response framessam2008-10-251-0/+1
| | | | | to avoid counting it as an unknown elemid (we often beacon this so it shows up as a "false positive")
* add 1/2 and 1/4 channel flags to IEEE80211_CHAN_ALL so that operationssam2008-10-251-0/+1
| | | | like ieee80211_find_channel are more useful
* add/improve debug msgssam2008-10-253-4/+18
|
OpenPOWER on IntegriCloud