summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2016-08-31Remove unnecessary 'error' assignmentTom Cosgrove
2016-08-31remove some ISA drivers for devices not recently seen in the wild.Ted Unangst
2016-08-31In pmap_activate instead of doing disable_interrupts/enable_interruptsJonathan Gray
nested inside of splhigh/splx just disable and enable interrupts once with inline cps instructions. Remove uneeded pcb_pl1vec block as well, suggested by kettenis. ok kettenis@
2016-08-31regenJeremie Courreges-Anglas
2016-08-31Estonian keymapJeremie Courreges-Anglas
Original diff from Taavi Tani in 2005 (!), adapted to -current by Mart Tonso. ok mpi@
2016-08-31Use 'sc_route{4,6}' directly instead of casting them to 'struct route *'.Martin Pieuchot
This is another little step towards deprecating 'struct route{,_in6}'. ok florian@
2016-08-31G/C rt_defmask4.Martin Pieuchot
2016-08-31Do not flush RTF_CLONED children when adding a new route.Martin Pieuchot
New RTF_CLONING routes don't have children, but the kernel might end up removing routes from a compatible route. This bug has been introduced with the support for multiple RTF_CLONING routes. Also make sure to release possible RTF_CACHED route *before* flushing RTF_CLONED children when deleting a route. KASSERT() reported by akfaew, sthen@ and martijn@ ok benno@
2016-08-31Drop gif(4) support, etherip(4) is what you want now.Martin Pieuchot
ok sthen@, deraadt@, dlg@
2016-08-31Delete unused ioctls and associated macros. Move macros that are stillAlexandre Ratchov
used internally by low-level drivers from sys/audioio.h to dev/audio_if.h instead of deleting them.
2016-08-30pool_setiplDavid Gwynne
ok markus@
2016-08-30pool_setiplDavid Gwynne
ok claudio@ mpi@
2016-08-30pool_setiplDavid Gwynne
ok yasuoka@
2016-08-30Use MaxMode to terminate the mode iteration. Actually the old wayYASUOKA Masahiko
couldn't interate the ConsOut modes properly on qemu. input NONAKA Kimihiro
2016-08-30Enable raspos24 for efifb. This makes qemu with UEFI start working.YASUOKA Masahiko
ok deraadt
2016-08-30Revert "Implement VFS read clustering for MSDOSFS"Stefan Fritsch
This caused garbage to be written instead of blocks of 0-bytes if a file is extended by seeking past the end. This happens for example when extracting files containing lots of 0-bytes with tar. ok mpi@ Details of original commit: msdosfs_vnops.c revision 1.105 denode.h revision 1.28 date: 2016/01/13 10:00:55; author: mpi; commitid: ru9jHQwQX09BC5Bw; Implement VFS read clustering for MSDOSFS. The logic used in msdosfs_bmap() to loop calling pcbmap() comes from FreeBSD and is not really efficient but it is good enough since it is only called when generating I/O. With this diff I get a 100% improvement when reading big files from a crappy USB stick. With this and bread_cluster(9) modified to not re-fetch B_CACHED buffers, reading large contiguous files with chunk sizes of MAXPHYS is almost as fast as physio(9) on the same device. For a 'real world' example, when copying music files from a USB stick I see a speed jump from 15MB/s on -current to 24Mb/s with this diff. While here rename some 'lbn' variables into 'cn' to better reflect what we're dealing with. Tested by Mathieu, with support from deraadt@
2016-08-30Use struct stat for storing attributes in fusebufs, because using structMartin Natano
vattr in userspace is suboptimal as some related helpers are not available, e.g. VATTR_NULL() and IFTOVT(). The conversion is now done in the kernel where it belongs. As a side effect the <sys/vnode.h> include can be removed from libfuse. tweaks and ok guenther
2016-08-30Move DT_NUM from <sys/exec_elf.h> to resolv.h in ld.so where it belongs.Mark Kettenis
ok guenther@
2016-08-30Make the init code match the comment and disable watchdogJonathan Gray
reset by mouse and keyboard instead of mouse and mouse.
2016-08-30Mask sb pci irq bits for a,b,c,d instead of a,b,b,d.Jonathan Gray
Couldn't find someone with a loongson 2e to test this on hardware. "looks ok" visa@
2016-08-30Properly set "msb" and "bps" fields for both play and rec directions.Alexandre Ratchov
Fixes "audio0: different play and record parameters ... " errors, and probably other unreported errors. help and ok landry@
2016-08-30use a per-table rwlock to serialize ART updates and walks, rather thanJonathan Matthew
taking the kernel lock. ok mpi@ dlg@
2016-08-30pool_setiplDavid Gwynne
ok natano@
2016-08-30pool_setipl for nfs poolsDavid Gwynne
ok natano@
2016-08-29Change le(4) buffer address to 0x71010000, which is the same addressKenji Aoyama
with NetBSD/luna68k. LUNA has 128KB so-called '3 port RAM' starting at 0x71000000. Originally, first 64KB of this area is designed for I/O processor HD647180, and next 64KB is designed for LANCE. This makes possible to support I/O processor in the future:-)
2016-08-29Set MTU size to ~4k until the TX path is ready to deal with larger packetsMike Belopuhov
2016-08-29Don't count output errors twiceMike Belopuhov
2016-08-29Fixup packet fragment unrolling procedureMike Belopuhov
When bus_dmamap_load fails to load one of the buffers in the mbuf chain, we need to revert all changes to transmit descriptors. The code to do that was prototyped but not tested. However due to how the Tx ring is set up in xnf(4) and generic lack of proper fragment support in the Netfront design we're always limited to having 256 entries for distinct shared memory pages. The mbuf chain is traversed and attempt is made to load every data chunk into a 4k sized DMA map segment which makes it impossible to reference a buffer composed of multiple pages. Current implementation lacks support for this preventing reliable transmission of frames larger than 4k. Bug reported by Kirill Miazine <km at krot ! org>, thanks!
2016-08-28Get rid of sxipio_ioh and sxipio_iot; they're not used (anymore).Mark Kettenis
2016-08-28Add a few missing sunxi-h3 clocks and resets.Mark Kettenis
2016-08-28Fix the macros that determine the register offsets based on port and pin.Mark Kettenis
These were didn't work as I expected, and returned the wrong offsets. As a result, pinctrl and gpio code was twiddling the wrong bits.
2016-08-28Reset the host controller after a failed command request to make sure itMarcus Glocker
operates again. ok kettenis
2016-08-28Remove obsolete gre(4) ioctls GRESADDRS, GRESADDRD, GREGADDRS,Reyk Floeter
GREGADDRD, GRESPROTO, and GREGPROTO. They have been replaced by generic ioctls some time ago (eg. the "ifconfig gre0 tunnel") and there is no need to keep the old ones around. They are neither used in base nor in ports. OK sthen@
2016-08-27Deassert reset signals if there is a "resets" property. Implement mmc powerMark Kettenis
sequencing like we have for imxesdhc(4). Unfortunately that doesn't seem to be enough to be able to see the onboard sdio wireless on my Banana Pi 2+.
2016-08-27Initialize vdd to 0 in imxesdhc_bus_power() to avoiud using it uninitialized.Mark Kettenis
2016-08-27Return -1 if we try to set the frequency of a clock but couldn't find it.Mark Kettenis
2016-08-27Add support for the SD/MMC clock to the geberic clock code and use it inMark Kettenis
sximmc(4).
2016-08-27Match on "allwinner,sun8i-h3-ehci".Mark Kettenis
2016-08-27Add support for the PXN bit in level 1 translation table descriptors andMark Kettenis
enable it on CPUs that support it. When enabled, this prevents the kernel from executing userland code. ok jsg@, tom@
2016-08-27Add support for the watchdog timer found on A31 and later SoCs, which has aMark Kettenis
slightly different register layout. Also fix the watchdog such that it actually disables itself when you set the period to 0 and actually gives you a period that is close to what you asked for.
2016-08-27Handle the quirky that the FIFO register sits at a different location onMark Kettenis
the A31 and later SoCs.
2016-08-27Add support for sun8i-h3, the Allwinner H3. For this SoC, the device treeMark Kettenis
contains a single clock control unit node that handles most clocks. The driver handles this through tables that describe the gating, reset signals and the clock hierarchy. This description is (deliberately) incomplete. We will add clocks on an as-needed basis. You will need a recent device tree, that includes a "allwinner,sun8i-h3-ccu" compatible node for things to work. ok patrick@
2016-08-27qaddr_t is (now) deadPhilip Guenther
ok deraadt@
2016-08-27Declare lseek() instead of assuming it'll be provided elsewherePhilip Guenther
ok deraadt@
2016-08-26Implement bus dma support for loading raw mappings so that we can usePatrick Wildt
xhci(4) on ARM. The only way the load raw operation can get to know about the coherent flag is via the segments. Store it there when the memory is initially mapped. Also store the virtual address which we need to know when we have to flush the caches on a non-coherent mapping. ok kettenis@
2016-08-26Add support for the Allwinner H3 (sun8i-h3).Mark Kettenis
2016-08-26Since the caller already checks and handles COHERENT we don't needPatrick Wildt
to explicitly check for the flag. ok kettenis@
2016-08-26Remove cpu_cpwait() calls; they are no-ops on armv7.Mark Kettenis
ok tom@
2016-08-26Remove the code that switches around MMU domains on armv7. MMU domains areMark Kettenis
basically a relic from the past. Using them doesn't make a lot of sense the way our pmaps work. Support for MMU domains isn't present in long-descriptor translation table format, so it is clearly on its way out. Based on a diff from Artituri Alm. ok patrick@
2016-08-26Remove unused prototype for sxiccmu_enablemodule().Marcus Glocker
ok kettenis
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-18 21:51:14 +0000