summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Expand)AuthorAgeFilesLines
* Use NG_HOOKSIZ instead of the deprecated (NG_HOOKLEN + 1)julian2004-05-241-1/+1
* Use PCI_BAR() rather than PCIR_MAPS.jhb2004-05-241-5/+5
* Wrap the code to save/restore PCI config registers on suspend/resume injhb2004-05-243-0/+18
* Change a if (...) panic() to a KASSERT().mux2004-05-241-2/+2
* Use PCI_BAR() in preference to PCI_MAPS + x * 4.imp2004-05-241-2/+2
* Grammar.maxim2004-05-241-1/+1
* Do not write to those config registers that are unambiguously definedimp2004-05-241-19/+0
* Revert part of rev 1.230 and assume that all EISA IRQs use active highjhb2004-05-241-8/+1
* Oops, ".align 4" for the data section in the previous commit shouldbde2004-05-241-3/+3
* Fixed profiling of trap, syscall and interrupt handlers and somebde2004-05-244-9/+43
* Don't repeat the definition of IDTVEC(). It is in asmacros.h.bde2004-05-241-3/+0
* Remove another redundant if_output initialization.mux2004-05-241-1/+0
* Fixed insertion sort error in previous commit (prof_machdep.c).bde2004-05-241-4/+4
* Store the target handles in a separate list from normal commands. Add anjl2004-05-246-15/+125
* Add support for BCM5705Kps2004-05-242-0/+3
* The socket code upcalls into the NFS server using the so_upcallrwatson2004-05-247-84/+642
* Consult parent's if_capenable for active VLAN-related capabilities.yar2004-05-231-1/+8
* Only initialize the if_sis callout as MPSAFE if debug.mpsafenet is setrwatson2004-05-231-1/+4
* A handler for ioctl(SIOCSIFCAP) should not alter a bit inyar2004-05-238-8/+17
* According to the design of checksum offloading framework,yar2004-05-231-2/+5
* Mark the VLAN_MTU capability as initially enabled since it'syar2004-05-232-0/+2
* Added profiling support for Xint0x80_syscall.bde2004-05-231-0/+2
* Don't try to copy out the result payload if there isn't one. This ioctldfr2004-05-231-6/+14
* Build prof_machdep.c if profiling.bde2004-05-231-0/+1
* Adjusted for amd64 after repo-copy. The adjustments are routine, except:bde2004-05-231-23/+53
* MFamd64 (1.117: made the FAKE_MCOUNT() in doreti work non-accidentally,bde2004-05-231-2/+1
* Restored FAKE_MCOUNT() and MEXITCOUNT invocations and adjusted them forbde2004-05-231-0/+8
* Update URL to HID spec.le2004-05-231-1/+1
* Update URL to HID spec.le2004-05-231-1/+1
* MFamd64 (put TF_EIP in assym.s and use it instead of a magic offset inbde2004-05-234-7/+8
* Adjusted FAKE_MCOUNT()s for amd64. This is needed for both ordinarybde2004-05-232-4/+4
* We don't need to initialize if_output, ether_ifattach() does itmux2004-05-2337-37/+0
* MFamd64 (1.111: fixed missing call to .mexitcount in lgdt()).bde2004-05-231-0/+1
* Spell "retq" as "ret" in pagezero() like it is everywhere, else sobde2004-05-231-1/+2
* Initialise OHCI_CROMHDR and OHCI_BUS_OPT in fwohci_ibr to make sure thatdfr2004-05-231-0/+7
* MFNetBSD 1.80; author: wizle2004-05-231-2/+3
* Replace the static "qdat" structure with a per-instance softc fieldiedowse2004-05-2312-32/+24
* Diff reduction to NetBSD.le2004-05-231-3/+5
* Use __FBSDID.mux2004-05-232-4/+6
* Apply fix for long timeouts on driver initialisation.dmlb2004-05-231-2/+7
* In agp_generic_bind_memory(), grab the needed pages before acquiringmux2004-05-232-60/+84
* Implement the new KERN_PROC_RGID option, and also implement thegad2004-05-221-0/+28
* Define a KERN_PROC_RGID option for kvm_getprocs().gad2004-05-221-0/+1
* Revert the previous commit. The bus accessor macros do not check thenjl2004-05-221-3/+56
* - Rearrange a comment to fit in 80 chars per line, like the rest of thismarius2004-05-221-4/+4
* - Move OFW_NAME_MAX, used as a limit for OFW property names and devicemarius2004-05-222-13/+24
* Change u_intXX_t to uintXX_t. Change a couple of 'unsigned long's todfr2004-05-2218-573/+573
* Add Intel PCI vendor ID.le2004-05-221-0/+4
* Don't forget to reset if_hwassist back to 0 when hardware checksummingyar2004-05-221-1/+3
* Remove two debugging printf().mux2004-05-221-2/+0
OpenPOWER on IntegriCloud