summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
...
* Now that _BSD_CLK_TCK_ and _BSD_CLOCKS_PER_SEC_ are the same on allmike2002-09-036-321/+0
| | | | | architectures, move the definition directly into <time.h> and finish the removal of <machine/ansi.h>.
* Remove an unneeded PROC_LOCK, which caused lock recursion panics.jake2002-09-021-1/+1
| | | | | | Print a warning about old applications with no signal trampoline. Reported by: marius@alchemy.franken.de
* Use the new kern_*() functions to avoid using the stack gap iniedowse2002-09-022-90/+49
| | | | linux_fcntl*() and linux_getcwd().
* Align _BSD_CLK_TCK_ and _BSD_CLOCKS_PER_SEC_ with most othermike2002-09-021-2/+2
| | | | | | | platforms. This introduces some binary incompatibilities for dynamically linked programs which make use of clock(3) and times(3). Approved by: jake
* Split up __getcwd so that kernel callers of the internal versioniedowse2002-09-022-18/+32
| | | | can specify whether the buffer is in user or system space.
* Align _BSD_CLK_TCK_ and _BSD_CLOCKS_PER_SEC_ with most othermike2002-09-021-2/+2
| | | | | platforms. This introduces some binary incompatibilities for dynamically linked programs which make use of clock(3) and times(3).
* Split fcntl() into a wrapper and a kernel-callable kern_fcntl()iedowse2002-09-022-65/+84
| | | | | implementation. The wrapper is responsible for copying additional structure arguments (struct flock) to and from userland.
* Hook up libkern/strlcpy.c and libkern/strlcat.c after repocopy.brooks2002-09-022-0/+4
| | | | | Obtained from: OpenBSD Discussed on: -arch
* Make SCSI_DELAY setable at boot time and runtime via thebrooks2002-09-026-25/+89
| | | | | | kern.cam.scsi_delay tunable/sysctl. Reviewed by: mdodd, njl
* Grammer cleanupdillon2002-09-021-2/+2
|
* fix bogus CTR3 message.davidxu2002-09-021-1/+1
| | | | Reviewed by: julian@freebsd.org (mentor)
* o Remove an initialized but unused variable from pmap_remove_all().alc2002-09-021-2/+0
|
* Moved elf brand identification into a function. Fully identify thejake2002-09-021-105/+75
| | | | | | | | brand early in the process of loading an elf file, so that we can identify the sysentvec, and so that we do not continue if we do not have a brand (and thus a sysentvec). Use the values in the sysentvec for the page size and vm ranges unconditionally, since they are all filled in now.
* o Synchronize updates to struct vm_page::cow with the page queues lock.alc2002-09-022-7/+8
|
* Fixed more indentation bugs.jake2002-09-021-3/+3
|
* Remove NOFORTH=yes, it seems to work now thanks to scottl.jake2002-09-021-1/+0
|
* Continue de-counting i4b. Devices i4bctl, i4bcapi, iavc, i4bq921,brooks2002-09-0286-512/+70
| | | | | | | | i4bq931, i4b, isic, iwic, ifpi, ifpi2, ifpnp, ihfc, and itjc are no longer count devices. Also remove a few other instances of N<DEVICE> being used to control compilation of whole files. Reviewed by: hm
* Fix things so that:mjacob2002-09-014-34/+278
| | | | | | | | | a) we don't believe what the board tells us all the time (if the BIOS hasn't run, port page 2 and port page 0 tend to be garbage) b) add the missing code to set parameters for the SPI cards. MFC after: 0 days
* Add a missing #include <sys/lockmgr.h>.iedowse2002-09-011-0/+1
|
* Minor fixupsscottl2002-09-011-1/+4
|
* Sync up: remove device counts.marcel2002-09-011-2/+2
|
* Use the new kern_* functions to avoid the need to store argumentsiedowse2002-09-018-382/+292
| | | | | | | | in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
* Added fields for VM_MIN_ADDRESS, PS_STRINGS and stack protections tojake2002-09-0121-87/+302
| | | | | | sysentvec. Initialized all fields of all sysentvecs, which will allow them to be used instead of constants in more places. Provided stack fixup routines for emulations that previously used the default.
* Restore Rev. 1.40 (remove "Keyboard yes/no" printf).kan2002-09-012-2/+2
| | | | | | | GCC 3.2 overflows boot2 by 12 bytes, this patch brings it back within the boundaries, with 12 bytes available for future bloat. Approved by: obrien
* Add a new function linux_emul_convpath(), which is a version ofiedowse2002-09-012-32/+64
| | | | | | | | linux_emul_find() that does not use stack gap storage but instead always returns the resulting path in a malloc'd kernel buffer. Implement linux_emul_find() in terms of this function. Also add LCONVPATH* macros that wrap linux_emul_convpath in the same way that the CHECKALT* macros wrap linux_emul_find().
* Split out a number of mostly VFS and signal related syscalls intoiedowse2002-09-017-216/+591
| | | | | | | | | | | | a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
* Remove drm_linux.h, move the two useful defines into drm_drv.h. Use fd lockinganholt2002-09-013-167/+73
| | | | | | | | on -current. Actually copy in data from userspace to kernel in the linux-compat ioctl path. Make sure ioctl sizes are as expected in the handler functions. Reviewed by: rwatson
* Since we have vp and td cached in local variables, use those insteadrwatson2002-09-011-1/+1
| | | | | | | of derefencing the VOP arguments again when calling the UFS code. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
* Unbreak LINT a little by not attempting to configure the nonexist optionbde2002-09-011-3/+0
| | | | AHC_DEBUG_SEQUENCER.
* Use the correct bit value for a debug optionsscottl2002-09-011-1/+1
|
* Remove options that don't actually exist (in this form).scottl2002-09-011-5/+0
|
* Bump __FreeBSD_version for _BSD_FOO_T_ to _FOO_T_DECLARED switch.mike2002-09-011-1/+1
| | | | | This should have been committed about a week ago to coincide with the actual changes.
* Upgrade to latest QLogic f/w from their intranet site for 2100, 2200 andmjacob2002-09-013-15262/+15533
| | | | | | 2300 cards. MFC after: 1 day
* Stylistic fixes for the inclusion of the opt_puc.h header, so it comes from thejmallett2002-09-013-7/+6
| | | | | | | right places, and so it's sorted consistently with how all the other N-hundreds of files ding this do it. Submitted by: Hiten Pandya <hiten@uk.FreeBSD.org>
* 'aicasm optional ahc ahd' means 'build aicasm only if BOTH ahc and ahdpeter2002-08-311-1/+6
| | | | are specified'. Ie: it is a logical and, not a logical or.
* Reduce the maximum KVA reserved for swap meta structures from 70 to 32 MB.dillon2002-08-312-5/+7
| | | | | | Reduce the swap meta calculation by a factor of 2, it's still massive overkill. X-MFC after: immediately
* More cleaning up and unhacking:phk2002-08-315-152/+249
| | | | | | | | | | | | | | | | | | | Don't expect all RIDs to be PCI rids. The previous code made at least 1 mistake, even for PCI. Give the card definitions a chance to specify a init function. Use this instead of the gross superio hack. Move the win877 init function to puc_pci.c where it belongs. RIDs can actually be zero, don't set badmuxed if so. Set a less incorrect end for the construct SYS_RES_IOPORT entries, I guess both sio and lpt happen to use 8 IO ports, but that shouldn't really be hardcoded this way. Fixup puc_pccard.c to match. We're getting closer.
* Make this compile.jake2002-08-311-1/+2
|
* Don't enable PUC_DEBUG by accidentphk2002-08-312-2/+0
|
* Split the puc driver in pci specific and generic parts.phk2002-08-315-99/+302
| | | | | | | Add a pccard frontend for it as well. The PCcard stuff does not work yet because there is still some PCImagic left in puc.c
* Make modems work:imp2002-08-311-20/+23
| | | | | | | | | | | | | | | | | Don't set the CCR_OPTION register when registering/deregistering interrupts unless the card is a MFC card. This makes a lot of cards work that didn't work before. These bits are only defined for MFC cards. Remove COOKIE_FOR_WARNER kludges that phk introduced. Add more debug info. Annoy the purists by committing two minor style nits at the same time. Many thanks to phk. He tracked this problem down to what was screwing things up, so fixing it became trivially easy once the problem was understood.
* Add cleandepend target.gibbs2002-08-311-0/+4
|
* Different and more correct workaround for broken cards under NEWCARD:phk2002-08-311-4/+8
| | | | | | | don't spam the configuration index register. I'm not sure where those bits truly belong, but spamming the config index sure doesn't do us any good.
* Pass the cleandir target on to the aicasm Makefile. This shouldgibbs2002-08-311-0/+4
| | | | fix the buildkernel target.
* Ignore all config entries which have a non-zero start I/O address.phk2002-08-311-0/+4
| | | | | | | This is a bandaid which makes some more PCCards work under NEWCARD which would otherwise not work. Cards with only fixed address windows still loose.
* Complete the hookup of ahd/ahc module building.gibbs2002-08-314-15/+42
|
* Enable ahd/ahc register pretty printing by default. This expeditesgibbs2002-08-312-2/+10
| | | | handling of bug reports.
* Add support for ahd/ahc register pretty printing in diagnostics.gibbs2002-08-312-12/+35
| | | | | | | | | This feature can be disabled via the AHD/AHC_REG_PRETTY_PRINT kernel option. The ahc driver now uses the same debug options mechanism as ahd: AHC_DEBUG - Compile in debugging code AHC_DEBUG_OPTS - String of debug options as listed in aic7xxx.h
* Sync perforce Ids.gibbs2002-08-312-2/+2
|
* Identify new controllers and the aic7901A.gibbs2002-08-311-33/+55
| | | | | | | | Move interrupt enable to the OSM. Remove some debugging messages. Remove chip bug workarounds for non-production asics.
OpenPOWER on IntegriCloud