summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2009-08-20fix indentationAlexander Bluhm
no binary change; ok grunk@
2009-08-19syncMatthieu Herrb
2009-08-19Trident Cyberblade XP2 as found in VTBook PCCard. ok marco@.Matthieu Herrb
2009-08-19<machine/macros.h> would provide inline version of a few of the functionsMiod Vallat
traditionnaly found in libkern. However, the memcmp() flavour would behave as bcmp() with only two possible return values: zero and positive non-zero. This broke the name cache RB trees which now rely upon proper memcmp() semantics(negative value, zero, or positive value). Just give up on these macros and provide the same code as libc, in libkern. As a side effect, this no longer uses the cmpc3 instruction, which is not implemented and requires (slow) kernel emulation, on the original uVax.
2009-08-19- missing \nJasper Lievisse Adriaanse
ok miod@
2009-08-18Disable address translation. Necessary on the RB600 where the firmware startsMark Kettenis
running us with address translation enabled. Lifted from a diff from dms@, tested on the Thecus N1200 by me.
2009-08-18Blind partial support for IOC4 chip, found on IO8 and IO9 base I/O boards onMiod Vallat
Origin 350 and Tezro systems. While this chip provides serial ports, an ATAPI interface and a PS/2 keyboard and mouse interface, this code currently only attempts to support the serial ports.
2009-08-18Replace a few hardcoded numbers from the interrupt register with properMiod Vallat
commented symbolic constants.
2009-08-18Blind support for SGI PIC PCI-X controller found on Origin 350 and TezroMiod Vallat
systems. PIC was supposed to be mostly XBridge compatible, but a silicon bug prevents it from working correctly if 32 bit register writes are used, so the xbridge(4) code now needs to issue 64 bit writes. In order to make this a bit more transparent, rebase all widget registers to a 8 byte boundary, and provide a few inline accessors which will do the right thing if you want 32 bit writes. Tested to not cause regressions on Octane, Origin 200 and Fuel; covering Bridge < 4, Bridge >= 4 and XBridge flavours.
2009-08-18On dual interrupt IOC3 designs, disestablish the ethernet interrupt if noMiod Vallat
ethernet driver attaches; prevents interrupt storms on Octane caused by the way ARCS initializes the chip, when not booting from the network.
2009-08-18Fix after readdisklabel() change.Matthias Kilian
Looks fine to miod@, who also reminded me to not forget to call flashunlock(sc).
2009-08-17fix readdpmelabel() declaration since the definition was changedDariusz Swiderski
from returning char* to int, last char* return removed as well (cleanup after deraadt@ commit, it compiles now) while there initialize hfspartend to some sane value (sync with macppc version) ok kettenis@
2009-08-17All callers of qli_get_fw_state() bzero the mbox memory, no need to do itMiod Vallat
again with a wrong size. ok marco@
2009-08-17sizeof(ptr) -> sizeof(*ptr) as intended; ok djm@ millert@Miod Vallat
2009-08-17Use ANSI function declarations. No binary change.Joel Sing
ok deraadt@
2009-08-17dd 'show all bufs' to show all the buffers in the systemJasper Lievisse Adriaanse
ok beck@ thib@
2009-08-17initialize retry to zero; otherwise if firmware sends odd val16Martynas Venckus
from the very beginning; we might never retry. ok mglocker@
2009-08-17the null check makes more sense when it's done before derefMartynas Venckus
ok marco@, jsing@
2009-08-17remove prototype of non-existent midi_get_hwif()Alexandre Ratchov
suggested by jsg@
2009-08-17By popular demand and peer pressure, check-in work in progress work to supportMiod Vallat
the Yeelong Lemote mips-based netbook. Kernel bits only for now, needs polishing; most of this work done during h2k9 last month, although the porting effort started earlier this year.
2009-08-16remove use of BITS and BIT macrosJonathan Gray
2009-08-16tweak a switch statement to appease lintJonathan Gray
2009-08-16remove use of BITS and BIT macros.Jonathan Gray
2009-08-16start getting rid of some of these horrific bit macros, removeJonathan Gray
usage of BITS(). There is a binary change due to the way these macros are further used in the MASK_AND_RSHIFT specifically for ATW_SR_RFTYPE_MASK and ATW_SR_BBPTYPE_MASK.
2009-08-16remove prototypes of a bunch of functions that had their implementationsJonathan Gray
removed in pfsync v5.
2009-08-16remove moscom_get_status() as it not called or required.Jonathan Gray
2009-08-16remove prototype for uark_open for which there is no implementation.Jonathan Gray
2009-08-16Update to cope with new readdisklabel() return value.Miod Vallat
2009-08-15use static for inline so we get the desired behaviour with compilersJonathan Gray
using c99 inline semantics.
2009-08-14First cut at UDF 2.[56] support, allowing read-only access to HDDVDKenneth R Westerback
and Blu-ray disks. Previously working DVDs should still work. Done at f2k9 with phessler@. Vnode bug squashing by beck@. Thanks to Bryan Brake for sending HDDVD/Blu-ray hardware and disks to f2k9 in Stockholm. ok beck@ dlg@ phessler@
2009-08-14add ddb functions to be able to show all the nfsnodes in the systemThordur I. Bjornsson
and rewrite the nfsreq code to use pool_walk(). OK beck@, blambert@
2009-08-14- fix function name in vprint()Jasper Lievisse Adriaanse
ok otto@
2009-08-14enable lisa(4); tested by Marco Knol; ok deraadtConstantine A. Murenin
2009-08-14Use the nfs_hashlock to protect the nfs_nodetree hanging of the mount.Thordur I. Bjornsson
What can happen is that a recycling of a vnode could pull one from out under us (since NFS has issues with ref counts...). Dance around getnewvnode() since we can end up recycling vnodes that where formerly owned by NFS, causing recursive locking. We where lucky with the old hashtables has the race was rare but now with more aggresive recycling we loose, just as theo found out on vax. help from oga, beck and blambert (beck mostly screamed though). ok oga@, beck@, blambert@
2009-08-14- let this compile after soreceive() changesJasper Lievisse Adriaanse
ok otto@
2009-08-14make the uvm device lock a mutex.Owain Ainsworth
This is the same diff that was backed out after c2k9 in the date-based revert. ok ariane@
2009-08-14Avoid double-release of vnodes in several functionsBret Lambert
ok thib@
2009-08-14- use u_char consistentlyJasper Lievisse Adriaanse
- minor formatting ok oga@ otto@
2009-08-14- fix format string for daddr64_t to %lldJasper Lievisse Adriaanse
ok otto@
2009-08-14- deregister; no binary changeJasper Lievisse Adriaanse
ok thibski@
2009-08-14Initialize len variable; found by Martynas VenckusBret Lambert
ok thib@
2009-08-14Another change to make my life easier when parsing pdus.Claudio Jeker
2009-08-13add a shutdown hook to stop unsolicited responses and the CORB andJacob Meuser
RIRB DMA engines
2009-08-13Remove unrelated bit from last commit which breaks at least 2 arches.Bret Lambert
Bad blambert@, no biscuit.
2009-08-13- ansify function declarations, no binary changeJasper Lievisse Adriaanse
"fine" thib@
2009-08-13Forgot to add break to the switch cases. Silly.Paul Irofti
Okay deraadt@.
2009-08-13make scsi_done set ITSDONE on the xs. this means hba drivers dont have toDavid Gwynne
do it (but doesnt preclude them using it internally). discussed with krw@ miod@ and deraadt@
2009-08-13More changes to make the parsing of those evil messages a bit easier.Claudio Jeker
2009-08-13rwlock assertion functions, currently unusedBret Lambert
ok art@
2009-08-13Add wake/suspend support for pckbd(4).Paul Irofti
Patch initially from mlarkin@. KNF and refactoring by me. Suggestions and okay deraadt@.