summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2003-10-05Stop the chip from stripping VLAN headers (the driverHenric Jungheim
ignores them) and tell the rest of the stack that "em" support VLAN frames. The "em" driver should now work with VLAN(4) pseudo-devs (and it should no longer be silently stripping VLAN headers when bridging). prodding by Attila Nagy ok jason@
2003-10-05Oops, commited from the wrong treeMiod Vallat
2003-10-05More cmmu routines cleanup:Miod Vallat
- remove unused routines (which were mostly here for debugging anyway) - do not include <machine/m88*.h> from <machine/cmmu.h> but rather only from the modules which need them - try to simplify the MVME187 codepath in the 8820x code, whenever possible
2003-10-05Do not use empty No_SSBR and No_Precheck macros, but rather pass emptyMiod Vallat
arguments to PREP*() macros. Slightly reorganize code in badaddr() - also do not make some internal labels unnecessarily visible from outside code.
2003-10-05Kill vm_offset_t and vm_size_t, in favor of the [pv]addr_t and [pv]size_tMiod Vallat
typedefs.
2003-10-05De-registerMiod Vallat
2003-10-05typosMiod Vallat
2003-10-05Correct db_{read,write}_bytes prototypes and use the official ones.Miod Vallat
2003-10-05Despite what the comments say, the instruction cache has never been turned on,Miod Vallat
since day one. It's never too late to fix it...
2003-10-05If attaching a wsdisplay console causes a font switch, do not update the PROMMiod Vallat
view of the cursor position, so that it comes back to a correct location when the system is shut down.
2003-10-05a few more devices to come later (commented out now)Michael Shalayeff
2003-10-05reload the itmr as soon as we get the intr to avoid lagging on slower machinesMichael Shalayeff
2003-10-05Fix tree breakage. We are not ready for RealTek PHYs, so back out thatKenneth R Westerback
addition, leaving only desired Broadcom PHY additions.
2003-10-05Fix tree breakage. We are not ready for RealTek PHYs, so back out thatKenneth R Westerback
addition, leaving only desired Broadcom PHY additions.
2003-10-05Misc. PHYs: more Broadcom, some RealTek.Kenneth R Westerback
The Broadcom devices are needed for the bge sync being attempted. From FreeBSD. As were the pcidevs committed earlier.
2003-10-05Misc. PHYs: more Broadcom, some RealTek.Kenneth R Westerback
The Broadcom devices are needed for the bge sync being attempted. From FreeBSD. As were the pcidevs committed earlier.
2003-10-05do not panic if microtime does not work to measure rng performance; from ↵Michael Shalayeff
andrisp@hipo.lv
2003-10-05from freebsd: read one word less (for type) and take care of odd sized recordsMichael Shalayeff
2003-10-04syncTheo de Raadt
2003-10-04new serverworks partsTheo de Raadt
2003-10-04Some more Broadcomm devices. Needed for sync to latest FreeBSD/NetBSD bge.Kenneth R Westerback
2003-10-04Some more Broadcomm devices. Needed for sync to latest FreeBSD/NetBSD bge.Kenneth R Westerback
2003-10-04vrele swap and/or root vnode should we fail during mountroot; found by ↵Michael Shalayeff
mpech@ and tedu@ ok
2003-10-04- document that boot will by default try to load /bsd, /obsd then /bsd.old.Jason McIntyre
- indent examples, and put descriptions before rather than after them. ok mickey@
2003-10-04fix xfs_devpoll so it works.Bob Beck
ok millert@
2003-10-04syncAlexander Yurchenko
2003-10-04VIA VT8237 southbridgeAlexander Yurchenko
ok deraadt@
2003-10-04(boot): device: filename -> device:filenameJason McIntyre
nice CAVEAT, BTW.
2003-10-04bpf support for atm cards; from jason@ackley.netTheo de Raadt
none of us can test this, but that does not mean it has to sit in the pr database
2003-10-04syncTheo de Raadt
2003-10-04Support for Matrix Orbital USB LCD Modules; cjeker@diehard.n-r-g.comTheo de Raadt
2003-10-03Use the commen DMT constants, and try to provide at least as many informationMiod Vallat
in "machine frame" than regdump() does. XXX this whole ddb/ directory needs a huge KNF spanking, badly
2003-10-03Merge tty_attach() in ttymalloc() and tty_detach() in ttyfree(). The need forMiod Vallat
separate tty registering is gone now that sparc has switched to wscons, and this makes the code less error-prone. Also, remove tests for ttymalloc() failure, since it uses M_WAITOK. ok millert@ deraadt@, tested by various people as well besides me...
2003-10-03Bring several fixes from FreeBSD to our current pipe implementation:Miod Vallat
- when allocating or growing a pipe buffer is not possible, do not panic but report the error and handle it correctly. (1.73) - "The pipe_write() code was locking the pipe without busying it first in certain cases, and a close() by another process could potentially rip the pipe out from under the (blocked) locking operation." (from Al Viro, 1.81) - "Remove test in pipe_write() which causes write(2) to return EAGAIN on a non-blocking pipe in cases where select(2) returns the file descriptor as ready for write. This in turns causes libc_r, for one, to busy wait in such cases. Note: it is a quick performance fix, a more complex fix might be required in case this turns out to have unexpected side effects." (1.141)
2003-10-02Upon invoking the freshly loaded binary, populate registers via embedded asmMiod Vallat
statements, and be sure of what we get, rather than use "hinted" register local variables; bugcrt can now be compiled with compiler optimization enabled.
2003-10-02Fix some DAE processing logic - the kernel would incorrectly end up retryingMiod Vallat
already processed, or unwanted, DAE in certain circumstances by mistake.
2003-10-02Ack external interrupts before reenabling them on 187.Miod Vallat
2003-10-02Round VM_MAX_KERNEL_ADDRESS to a page boundary.Miod Vallat
2003-10-02do not try to send incomplete fragments on ENOBUFS caseJun-ichiro itojun Hagino
(behavior change from 4.4bsd). dhartmei ok
2003-10-02correct endian handling of ip->ip_off.Jun-ichiro itojun Hagino
do not try to send incomplete fragments on ENOBUFS case (behavior change from 4.4bsd). dhartmei ok
2003-10-01it turned that only adding the cardbus part was not enough to haveFederico G. Schwindt
D-Link DFE690 working. add the missing bits here. from Joshua Johnson <joshua.johnson at charter dot net>.
2003-10-01Unused bits.Miod Vallat
2003-10-01There is no point in chown/chmod outside the install target.Miod Vallat
2003-10-01use random number generator to generate IPv6 fragment ID/flowlabel.Jun-ichiro itojun Hagino
cleanup IPv6 flowlabel handling. deraadt ok
2003-10-01correct m_cat misuse. i remember it was ok'ed by someone but don't remember ↵Jun-ichiro itojun Hagino
who...
2003-10-01Fix stack leak into pfloghdr (the pad field is not initialized),Can Erkin Acar
which ends up in the pflog pcap file. From dhartmei@ ok dhartmei@, frantzen@, henning@
2003-10-01Compile kernel and bootblocks with -mmemcpy, to get rid of libgcc dependancy.Miod Vallat
Also, honor default compilation options in stand/ whenever possible. This requires an up-to-date compiler, but it's not exactly like many people are going to be affected by this change...
2003-10-01out-of-bound memory access. deraadt okJun-ichiro itojun Hagino
2003-10-01Fix off-by-one.Chad Loder
OK jason@, henric@
2003-10-01make ifindex2ifnet growing code more correct. (found by iij seil team)Jun-ichiro itojun Hagino