summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2013-01-15If timing between keys is less than (by default) 1 millisecond, assumeNicholas Marriott
the text is being pasted. assume-paste-time option changes the value (0 disables). Based on a diff from Marcin Kulik.
2013-01-15Bring back symbol visibility support. This had been done, then backed out,Miod Vallat
a long time ago, because I had not realized symbol visibility required a lot of shared library major bumps to avoid mixed binaries. However, at the moment, none of the platforms using gcc 3 supports shared libraries (this is being worked on), and the remaining gcc 2 platforms won't support gcc 3 until switching from a.out to ELF, which makes the risk of mixing visibility-unaware binaries with visibility-aware binaries moot. Reminded by brad@.
2013-01-15Pass an EFBIG error to user land when the maximum splicing lengthAlexander Bluhm
has been reached. This creates a read event on the spliced source socket that can be noticed with select(2). So the kernel passes control to the relay process immediately. This could be used to log the end of an http request within a persistent connection. deraadt@ reyk@ mikeb@ like the idea
2013-01-15Add IMSG_WRITE_FILE and associated bits to allow the unprivilegedKenneth R Westerback
process to ask that a file be written by the privileged process. Not yet used.
2013-01-15tidyMarc Espie
okay miod@
2013-01-15build crt{begin,end}S.o with -fpic.Miod Vallat
2013-01-15When configured for ELF object format output, recognize #abdiff, #got_rel andMiod Vallat
#plt symbol modifiers, and produce the appropriate relocation types. This allows as(1) to correctly assemble PIC code.
2013-01-15syncTodd T. Fries
2013-01-15tidy up the vio docs; ok sfJason McIntyre
2013-01-15Switch mfs to bufq. This simplifies the code, gives us queue limits forJoel Sing
free and uses FIFO queueing (rather than the current LIFO queueing). Tested by sthen@ ok beck@ krw@
2013-01-15Changing the socket buffer flags sb_flags was not interrupt safeAlexander Bluhm
as |= and &= are non-atomic operations. To avoid additional locks, put the flags that have to be accessed from interrupt into a separate sb_flagsintr 32 bit integer field. sb_flagsintr is protected by splsoftnet. Input from miod@ deraadt@; OK deraadt@
2013-01-15Process futex requeuing even when the thread times out or is signaled.Philip Guenther
Bug report from frantisek holop (minusf at obiit.org) ok pirofiti@
2013-01-15The socket splicing tests allow the kernel to return EFBIG in caseAlexander Bluhm
the splicing maximum has been reached.
2013-01-15Flushing prefixes of autoconfigured IPv6 addresses with ndp -PAlexander Bluhm
caused an uvm fault. The function in6_purgeaddr() might already free the prefix before prelist_remove() is called. So call in6_purgeaddr() for all addresses first and then loop over the remaining prefixes and call prelist_remove(). OK claudio@
2013-01-15Implement a function that gets and correctly initialises a ccb for I/O,Joel Sing
along with a function that enqueues a ccb on a workunit. ok krw@
2013-01-15implement bus_space_foo_8 for amd64. it will panic on attempts to do 8David Gwynne
byte operations on I/O space with "invalid I/O operation", but works great for mem mappings. compiled AND tested on myx(4) on amd64. ok deraadt@ kettenis@ miod@ pirofti@
2013-01-15Always initialise the discipline name, not just when we are doing a create.Joel Sing
2013-01-15Make the FAT setattr handling more consistent about unsupportedMartynas Venckus
attributes; making it more "user-friendly". Mostly from FreeBSD. OK guenther@, krw@.
2013-01-15Don't pull <math.h> just because it can return HUGE_VAL in theMartynas Venckus
corner cases. OK millert@.
2013-01-03Reengineer the socket splicing regression tests:Alexander Bluhm
- Move the tests from splice to new sosplice directory for consistent naming. - Split the API tests and the TCP splicing tests into separate directories. - Create some tests for the upcoming UDP splicing. - Tests can be run in obj directories now. - The API tests can run both on a local and on a remote machine now. - Fix the forking TCP tests which splice and read or write simultaneously. - Bunch of little fixes for races in the tests. - Deduplicate code, move checks into common functions.
2013-01-15Fix typos in comments and change the syntactic sugar of a function toClaudio Jeker
reduce the diff between Fx and Ox. OK gcc
2013-01-15put a tab after #define to make the diff between Fx and Ox smaller.Claudio Jeker
No binary change.
2013-01-15Keep a function pointer to the per-discipline I/O interrupt handler in theJoel Sing
discipline data structure. To be used with an upcoming diff.
2013-01-15dont use amd64 is currently broken cos it has noDavid Gwynne
bus_space_write_raw_region_8. disabling it for now.
2013-01-15Factor out workunit initialisation code.Joel Sing
ok krw@
2013-01-15Fix segmentation of buffers that straddle a boundary.Mark Kettenis
ok dlg@
2013-01-15Allow SIGKILL to terminate coredumping processes. Semantics decidedTheo de Raadt
with kettenis guenther and beck. ok guenther
2013-01-15Slice & dice coredump write requests into MAXPHYS blocks, andTheo de Raadt
yield between operations. Re-grab the vnode every operation, so that multiple coredumps can be saved at the same time. ok guenther beck etc
2013-01-15Loongson: Remove some of the unused defines.Paul Irofti
Increases readability.
2013-01-15use bus_space_write_raw_region_8 on 64bit archs when writing to the ringsDavid Gwynne
2013-01-15Loongson: Unbreak the tree after my last commit.Paul Irofti
Patch appended my final work on glxclk to an earlier version that I forgot was there.
2013-01-14map the registers PREFETCHABLE so things that can do write combining canDavid Gwynne
try and do write combining like the myx doco likes.
2013-01-14avoid extra bus_space barriers in the interrupt handler.David Gwynne
2013-01-14use one interrupt handler for the thtc headend, rather than seperateTheo de Raadt
interrupt handlers for the ports. this ensures that the 2nd port does not get starved. ok dlg
2013-01-14after OpenBSD requested a etherip rev 3 to solve a byte order weakness, andTheo de Raadt
worked with them to define a specific format, ietf quietly changed their mind and did something different. First step of transitioning over two release -- accept both types, but continue sending old format. pointed out by david.dahlberg@fkie.fraunhofer.de ok claudio
2013-01-14Remove an unnecessary function castPhilip Guenther
ok jsing@
2013-01-14When showing threads, say so. Fix up some comments and names at the same time.Philip Guenther
original diff by zhuk@ ok jsing@ mpi@ zhuk@
2013-01-14Loongson: Replace the system clock provided by the CPU with a GLX MFGPT.Paul Irofti
CPU throttling was not possible due to the fact that the system clock was the CPU clock. So slowing down the CPU would also slow down the passing of time. This commit adds a driver for the MFGPT1 clock from the AMD companion chip found on lemote and hooks it up as the system clock. It also changes the frequency value of hz from the default, which was 100, to 128. That's because the scaling on MFGPT clocks is represented by powers of two. Okay miod@.
2013-01-14update to nginx-1.2.6Robert Nagy
2013-01-14kroute.c rev 1.69 of ospfd for ospf6dFlorian Obser
------------------------------------------------------------------------ date: 2009/06/02 20:16:59; author: claudio; state: Exp; lines: +13 -3 Track reject and blackhole routes and allow them to be redistributed even though they point to the loopback. Mainly used for redistribute default since on default free routers we need to have a fake route now. After discussion with Tonnerre Lombard, idea OK henning@ ------------------------------------------------------------------------ ok claudio@
2013-01-14Fix the sotfware.Joel Sing
2013-01-14Call the correct callback on node disassociation.Joel Sing
2013-01-14Free the directory block buffer at the completion of a readdir,Joel Sing
significantly reducing NTFS kernel memory usage. Without this the buffer is not usually freed until the vnode is reclaimed. Tested by Sebastian Neuper. ok beck@ krw@
2013-01-14No longer need the "miniproc" glue, as we only have kinfo_proc now.Philip Guenther
ok deraadt@ pirofti@ kettenis@
2013-01-14when posting descriptors to the chips rings, avoid going write barrierDavid Gwynne
write barrier write barrier when using myx_write to post descriptors. instead let its go write write write barrier by using the appropriate bus_space write directly followed by a single bus_space barrier. the story above is mostly true, except that myx wants use to write all the descriptors except the first, barrier, and then write the first one out to signale that the chip can proceed. it is also worth noting that the barriers cover more address space than what we actually wrote to. this makes the code much simpler, and avoids generating extra fence operations (which is what barrier functions end up as on most of our archs) when we wrap around the end of the ring. the bus_space doco encourages this. bus_space use was discussed with krw@ kettenis@ deraadt@
2013-01-14Do not bogusly attempt to mprotect ld.so's .rodata area RW at bootstrapMiod Vallat
time: the logic is wrong, and there are no relocations to apply to .rodata anyway. ok jsing@ kettenis@
2013-01-14sm_usmerr should be initialized with OIDVAL, not MIB; ok reyk@Miod Vallat
2013-01-14the myri doco suggests its nice to post stuff by filling in everythingDavid Gwynne
in the rings except the first descriptor. once you've written as much as you can out, then you go back and post the first descriptor to signal that the chip should go ahead and work.
2013-01-14First convert packet to lease, which validates option data andKenneth R Westerback
discards bad options. THEN check to see if any required options are missing and reject both OFFER and ACK packets that lack required options. Since it is the latter's lease we actual bind. Move required option check into packet_to_lease() instead of duplicating it.
2013-01-14Ansify and apply style(9) to function definitions.Joel Sing
ok krw@
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-15 23:13:24 +0000