summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2013-09-27IPv6 atomic fragments must not go the reassembly queue, but beAlexander Bluhm
processed immediately. Let pf step over an atomic fragment header and handle the packet like an unfragmented. OK mikeb@
2013-09-26Clear SYSFAIL upon boot, for real.Miod Vallat
2013-09-26Tinkering with the cache diagnostic registers have told me enough informationMiod Vallat
to figure out the addressing split scheme used by DG in their 6:1 CMMU designs.
2013-09-26Use the cpuid vendor string instead of the model string when enablingJonathan Gray
VIA specific amd64 code. Makes the code work with Eden X2 processors which have the same model/family as a Nano but don't claim to be one in the model string. from bytevolcano at Safe-mail.net
2013-09-26Move the logic for deleting an IPv4 address to its own function.Martin Pieuchot
No functional change. ok henning@, mikeb@
2013-09-25Unbreak the tree; no 64-bit atomic operations on 32-bit architectures for now.Mark Kettenis
found out the hard way by chris@ and mpi@
2013-09-24Make (almost all) atomic operations really atomic.Mark Kettenis
Potentially fixes a race/panic reported by claudio@ ok jsg@
2013-09-24Make sure vblank_time_lock blocks interrupts to match Linux.Mark Kettenis
Potentially fixes a race/panic reported by claudio@ ok jsg@
2013-09-24Extend le@syscon to support the ILACC chips found on-board the AV530 family.Miod Vallat
GENERIC.MP on a 2x88100(6:1)@33 4605 can now make build on local disk and local net; but it still needs to boot its kernel off the network at the moment.
2013-09-24Sync the MI LANCE code ( le(4) ) with NetBSD, except for the following:Miod Vallat
- the am7990_get() - now lance_get() - is unchanged. - the interrupt acknowledge logic is unchanged, and will disable interrupts, then acknowledge all interrupt conditions. Add ILACC (79900) support (from NetBSD). Both LANCE (am7990.c) and ILACC (am79900.c) code share as much common code (lance.c) as possible. This affects all le(4) attachments, but the changes are mostly mechanical, to split am7990-specific parts from lance-agnostic parts. Compile tested on all affected platforms. Tested on alpha, hp300, luna88k, mvme88k, sparc, sparc64 and vax.
2013-09-24a bit more tmpfs scaffolding (still not active, no kernel change)Marc Espie
reduces the actual diff needed, and allows one to build mount_tmpfs without reinstalling the includes. (still awaiting review on the uvm bits) okay krw@, tedu@
2013-09-24Stop generating traffic for a given endpoint when a transfer reportedMartin Pieuchot
an I/O error. When such thing happens, the device is more likely to be already gone or non responding, but the task responsible for aborting the pipe has not been executed. Fix a problem where the ehci(4) controller would vomit^Wprint a lot of "ehci_idone:" messages in loop leaving the machine unusable, initially reported by ajacoutot@ and later by RD Thrush. Fix tested by RD Thrush and Markus Bergkvist, thanks! ok miod@
2013-09-24Try to release the console if pckbc(4) fails to send the initialMartin Pieuchot
command byte. This allows at least my USB keyboard to be part of the console on a Lenovo E20. ok miod@
2013-09-23Remove SLOWSCSI and the millisecond delay after each phase change it introduces.Miod Vallat
This code was added back when the scsi code between the kernel and the boot blocks was shared, to work around an obscure kernel problem (and definitely not related to bootstrap, at all). While there, increase the target selection timeout from 2ms to 250ms, to match the value currently used by the kernel. Crank boot blocks version. Booting a kernel from disk is now significantly faster.
2013-09-23- change m8820x_cmmu_set_reg() and m8820x_cmmu_set_cmd() to performMiod Vallat
unconditional actions. Introduce _if_addr() and _if_mode() variants of these to act on a given cmmu type only, or on a matching address line only. This makes the code slightly larger but easier to read and a tad faster. - report each cmmu on its own line in dmesg, sorted by type (I$ first, then D$). make sure disabled line information ends up in dmesg, rather than in the early bootstrap messages. - fix the multiprocessor writeback logic in m8820x_dma_cachectl() again to be more reliable; also only rendezvous the cmmus after all cmmu operation have been issued, instead of doing a rendezvous per cpu.
2013-09-23regenJonathan Gray
2013-09-23correct some radeon entries and add new Intel Ethernet and 8 series idsJonathan Gray
2013-09-22Stop merging the per-thread and per-process flags when reportingPhilip Guenther
them via sysctl(KERN_PROC). In struct kinfo_proc the per-process flags move to p_psflags, leaving the per-thread flags in p_flags. Flag descriptions in ps(1) updated to be less obtuse. discussed with matthew@ some time ago; ok jca@, manpage bits ok jmc@
2013-09-22Delete a T32 chunk that I missed before and update the comment onPhilip Guenther
ufs_readdir() to match
2013-09-22On return, uio_offset needs to be set to the offset of the first entryPhilip Guenther
not returned instead of the offset after that one. udf_uiodir() returns -1 when there isn't enough space for the next entry, so just break from the loop and return 0 in that case. Remove the bogus printf there now that this is handled correctly. problem reported by espie@
2013-09-22Support MSI for virtio. This is useful for bhyve.Stefan Fritsch
2013-09-22The readdir vop should set uio_offset to the cookie of the entryPhilip Guenther
after the last entry that fit, but I had it setting it to the cookier of the entry after that, so it would skip one entry for each block required. Remove the no-longer-needed cnt/cntp variable while we're touching things. prodding and testing by espie@
2013-09-21Grok ELF files. Good enough to netboot kernels on AV530 again.Miod Vallat
2013-09-21regenMiod Vallat
2013-09-21Oxford OXPCIE952 parallel port puc(4) board; contributed by Markus UhlinMiod Vallat
(markus.uhlin; bredband dot net), thanks!
2013-09-21Switch to MI dart(4) driver.Miod Vallat
2013-09-21Allow for an out-of-softc indirection to reach the software copy of theMiod Vallat
important (and write-only) register state; this will allow this driver to be used as a console before it attaches, to be used shortly.
2013-09-21regenJasper Lievisse Adriaanse
2013-09-21add several devices found on the hp microserver gen8 and add some moreJasper Lievisse Adriaanse
xeon e3-1200v2 pcie ids while here as suggested by jsg@ ok jsg@
2013-09-21Create process map holes with UVM_INH_SHARE so that they don't get lost inMiod Vallat
fork-without-exec situation (such as privsep'd binaries). Fixes occasional SIGSEGV in syslogd and pflogd on sun4/4c/4e.
2013-09-21Don't invoke pmap_copy() on map holes.Miod Vallat
2006-05-09Oops, correct import this time.Miod Vallat
2013-09-20Add support for root on nfs using v3. Code adapted from NetBSD.Federico G. Schwindt
Tested on sparc by miod, octeon by aalm and armv7 by me. miod ok.
2013-09-20Unify and document usbd_transfer(9).Martin Pieuchot
Get rid of the badly named usbd_bulk_transfer() and usbd_intr_transfer() they were two identical wrappers to submit a synchronous transfer. Now all synchronous transfers are submited using usbd_setup_xfer(9) and usbd_transfer(9) explicitly using the USBD_CATCH flag when signals must be checked. Previous version ok'd by mglocker@, with subtle inputs from miod@
2013-09-20Use a working magic sequence to configure absolute mode for elantech v2.Stefan Sperling
The synaptics driver now attaches to elantech v2 touchpads. It seems these pads were always falling back to compat mode, which went unnoticed until r1.47. Fix based on a patch submitted by Cedric Tessier: http://marc.info/?l=openbsd-bugs&m=136904396518528&w=2 which in turn was probably based on parts of this patch: https://launchpadlibrarian.net/73099683/elantech.patch which updates the Linux driver code but isn't part of Torvald's Linux. Behaviour of elantech firmware versions seems to vary quite a bit so print the firmware version at attach time for reference in future bug reports. Fix tested by frantisek holop and Peter J. Philipp. ok mpi@
2013-09-20Fix a panic bug in pipex. If pipex deletes a session by the idle-timerYASUOKA Masahiko
when the userland program (npppd) is dead or frozen, the session remains in state_list after it is destroyed, it will be used after free.
2013-09-20remove debug code. not sure how this lasted so long.Ted Unangst
noticed by maxime villard
2013-09-19Tweak types to keep daddr_t address and sector address separate.Kenneth R Westerback
Prefer DL_ macros over handrolling. Fix the loop to allow for bigger (highly unlikely) bunches of bits to be broken up into rw_10 sized (<= UINT32_MAX sectors) chunks. Add check to make sure i/o request starts at a sector address.
2013-09-19Do no try to clear the endpoint, as it if has stalled, if theMartin Pieuchot
USB transfer reported an I/O error. This is more likely to have no effect apart generating more errors because when this happens, the device is generally already physically detached but the *detach() task has not run yet. So simply return and wait for the device to be removed like ucom(4) does. Partially fix an issue reported by RD Thrush rt at thrush.com. ok miod@
2013-09-19get ethernet working on the edgerouter liteJonathan Matthew
- add atphy(4) to configs - allow separate rx and tx clock settings - add phy mapping for the erl - add tx/rx clock settings for erl ok miod@ bcallah@
2013-09-18sync the execbuffer relocation code with linux 3.8.13Jonathan Gray
with the fastpath and cpu relocs disabled for now. eb_* functions based on code in FreeBSD. ok kettenis@
2013-09-18return after done in atascsi_disk_unmap.David Gwynne
2013-09-17size_t could not be < 0.Sylvestre Gallon
Thanks to Pedro Martelletto.
2013-09-17Add support for AMD Family 15h Models 10h-1Fh Processors.Mark Kettenis
2013-09-17regenMark Kettenis
2013-09-17Rename existing AMD Family 15h Models 00h-0Fh Processor entries and addMark Kettenis
a bunch of AMD Family 15h Models 10h-1Fh Processor entries.
2013-09-17Change vlan(4) detach procedure to not use a hook but a list of vlansMartin Pieuchot
on the parent interface. This is similar to what bridge(4), trunk(4) or carp(4) are doing and allows us to get rid of the detachhook. ok reyk@, mikeb@
2013-09-17Fix a misaligned backslashPhilip Guenther
2013-09-17KNF - extraneous blank line snuck in during last commitMike Larkin
2013-09-16work out phy addresses before attaching cnmac and pass them to mii_attachJonathan Matthew
rather than hiding them in the mii read/write functions. ok dlg@ jasper@ bcallah@
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-04 11:20:59 +0000