summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Correct source file corruption in last checkingrog1998-09-301-2/+1
| | | | Observed by: jkh
* Revise test code for sigwait and add test code for sigsuspend.jb1998-09-302-0/+281
| | | | Submitted by: Daniel M. Eischen <eischen@vigrid.com>
* Revise test code for sigwait and add test code for sigsuspend.jb1998-09-304-18/+116
| | | | Submitted by: Daniel M. Eischen <eischen@vigrid.com>
* Fix a comment.jb1998-09-301-1/+1
|
* Cosmetic cleansing. This code requires extra work to keep the garbagejb1998-09-303-6/+9
| | | | collector thread running after a fork.
* Move the cleanup code that frees memory allocated for a dead thread fromjb1998-09-3020-390/+700
| | | | | | | | | | | | | | the thread kernel into a garbage collector thread which is started when the fisrt thread is created (other than the initial thread). This removes the window of opportunity where a context switch will cause a thread that has locked the malloc spinlock, to enter the thread kernel, find there is a dead thread and try to free memory, therefore trying to lock the malloc spinlock against itself. The garbage collector thread acts just like any other thread, so instead of having a spinlock to control accesses to the dead thread list, it uses a mutex and a condition variable so that it can happily wait to be signalled when a thread exists.
* Use snprintf instead of sprintf to avoid long source file paths fromjb1998-09-303-36/+93
| | | | | | | launching an application into space when someone tries to debug it. The dead thread list now has it's own link pointer, so use that when reporting the grateful dead.
* Implementation of an additional state called SIGWAIT (with the previousjb1998-09-3012-204/+348
| | | | | | one renamed to SIGSUSPEND) to fix sigwait(). Submitted by: Daniel M. Eischen <eischen@vigrid.com>
* NULL a pointer after it is freed to avoid trying to free it again.jb1998-09-303-0/+6
|
* - Fix the debug macros.jb1998-09-303-12/+63
| | | | | | | | | | | - Add support of a thread being listed in the dead thread list as well as the thread list. - Add a new thread state to make sigwait work properly. (Submitted by Daniel M. Eischen <eischen@vigrid.com>) - Add global variable for the garbage collector mutex and condition variable. - Delete a couple of prototypes that are no longer required. - Add a prototype for the garbage collector thread.
* Delete the XXX comments that refer to spinlock recursion. The malloc/free/jb1998-09-301-15/+1
| | | | | | | | | realloc functions check for recursion within the malloc code itself. In a thread-safe library, the single spinlock ensures that no two threads go inside the protected code at the same time. The thread implementation is responsible for ensuring that the spinlock does in fact protect malloc. There was a window of opportunity in which this was not the case. I'll fix that with a commit RSN.
* Remove SCSI support as the only driver in this kernel config, aic, is notgibbs1998-09-301-6/+6
| | | | currently supported by CAM.
* Fix typo in message.danny1998-09-303-6/+6
|
* Add several missing ioctl handlers. One needed by Sybase, the othersjfieber1998-09-304-4/+58
| | | | found while looking for the one.
* Do not allow a mounted on directory to be rmdir'ed. This removal canmckusick1998-09-301-2/+8
| | | | | | | happen when an NFS exported filesystem tries to remove a locally mounted on directory. PR: kern/7272 Submitted by: Andre Albsmeier <andre.albsmeier@mchp.siemens.de>
* Perform a hard reset on cards when the command fails. This should helpimp1998-09-301-2/+2
| | | | | those people that have cards that become wedged when a bogus command is issued that are too wedged to have a soft reset help.
* In nfs_link(), check for a cross-device mount *before* lookingmckusick1998-09-292-4/+6
| | | | | in the v_data field. Obtained from: Charles Hannum, via Frank van der Linden <frank@wins.uva.nl>
* Missing vput when cross-device link error is detected in nfs_link.mckusick1998-09-292-0/+2
|
* During truncation, have to notify the VM about the new sizemckusick1998-09-292-6/+10
| | | | | of the NFS file *before* doing the nfs_vinvalbuf operation. Otherwise some invalid data may show up in an mmap.
* Don't require an argument for -v flaggrog1998-09-292-11/+18
| | | | | | | Correct checks for null special file names Add Usage entry for -v flag Get terminology straight in man page Reviewed by: bde
* Frank sez: 'It fixes a problem with servers that return 0 valuesmckusick1998-09-292-14/+20
| | | | | | | | for some of the fsinfo RPC fields. It is strictly speaking not wrong to do this, as the spec says that "it is expected that a server will make a best effort at supporting all the attributes", but pretty unusual. You guessed it, it's NT servers that do it.' Obtained from: Frank van der Linden <frank@wins.uva.nl>
* Do not need (or want) to take a reference on an NFS file thatmckusick1998-09-292-12/+24
| | | | | | | | | is being deleted due to an forcible unmount. The problem is that vgone calls vclean() which then calls calls nfs_inactive() with VXLOCK set on the vnode. Nfs_inactive() was calling vget() to get a reference on the vnode, which in turn hung on VXLOCK. Nfs_inactive() now checks v_usecount to make sure that the vnode is not coming from vclean() before it does a vget().
* The code checks each fragment mark to see if it's valid; if the fragmentmckusick1998-09-293-6/+6
| | | | | | | | | | | | | | | | | | | | | is less than NFS_MINPACKET or greater than NFS_MAXPACKET in size, it barfs and, I think, drops the connection. However, there's no guarantee that in a multi-fragment RPC, all the fragments will be at least as large as NFS_MINPACKET. In fact, with the version of "tclnfs" we have here, which supports NFS over TCP, at least when built under SunOS 4.1.3 (i.e., with 4.1.3's user-mode ONC RPC library), I can *repeatably* cause "tclnfs" to send a request with more than one fragment, one of which is only 8 bytes long. I just do a 3877-byte write to a file, at an offset of 0. The check that "slp->ns_reclen" is greater than or equal to NFS_MINPACKET serves no useful purpose - if the NFS server code can't handle packets < NFS_MINPACKET bytes, it can't handle them over *any* protocol, so the check has to be done above the RPC-over-TCP layer - and should be removed. Obtained from: Fix from Guy Harris, forwarded by Rick Macklem.
* In the bootverbose case, print out error messages for all errors that willken1998-09-292-8/+18
| | | | | | not be retried again, even if the SF_NO_PRINT flag is set. Reviewed by: gibbs
* vm86_datacall: always use workaround since temp. malloced buffer or stackache1998-09-291-22/+18
| | | | | area can be passed (and mapped to page1!) as vesa.c does. Use contigmalloc now to get proper alignment. Bump max buffer size to PAGE_SIZE
* Add back a few useful targets lost in the reshuffle.jkh1998-09-292-8/+8
| | | | Reviewed by: jb
* Added double quotes around CHMOD description to prevent garbled output.alex1998-09-291-2/+2
| | | | | PR: 8094 Submitted by: Christoph Weber-Fahr <wefa@callcenter.systemhaus.net>
* Mark directory buffers that have no valid data with B_INVALmckusick1998-09-292-2/+12
| | | | so that they are not put in the cache.
* When adding data to a buffer, we need to clear the B_NEEDCOMMIT flagmckusick1998-09-292-2/+4
| | | | which says that the data is on server but not committed.
* Restore v1.3 - page align workaround moved to vm86_datacall nowache1998-09-291-14/+5
|
* Move workaround about page aligned data buffer directly to vm86_datacall,ache1998-09-291-4/+24
| | | | | it is impossible to use this func otherwise, i.e. all vesa calls are potentially broken. Max arg size limited to 1024 for now, bump it, if needed.
* Fixes for lkm:rvb1998-09-2910-24/+48
| | | | | 1. use VFS_LKM vs ACTUALLY_LKM_NOT_KERNEL 2. don't pass -DCODA to lkm build
* Rename a static variable, so it will not shadowed by a local variable.dt1998-09-291-7/+7
| | | | | | Now comments will be ignored, rather than put junk in the password database. Broken in: rev. 1.21
* Fix the shared library configuration stuff.markm1998-09-293-15/+15
| | | | Noticed by: Ollivier Robert
* Don't erase curproc when making a vm86() call. The previous behaviorjlemon1998-09-291-7/+2
| | | | | | was a pessimization that broke schedcpu(). (It caused the process to be remrq()'d). Reviewed by: Tor Egge
* Add history.obrien1998-09-291-1/+6
|
* Make #define NO_SWAPPING a normal kernel config option.abial1998-09-293-3/+6
| | | | Reviewed by: jkh
* Make #define NO_SWAPPING a normal kernel config option.abial1998-09-293-3/+39
| | | | | | Warn unsuspecting users against current DEVFS pitfalls. Reviewed by: jkh
* mkdosfs(1) is dead, refere to newfs_msdos(8)obrien1998-09-291-3/+3
|
* Add in a kludge for registering the "package" version of XFree86 sojkh1998-09-293-9/+24
| | | | | | | | various ports don't complain about it. It also requires that the pkg registration bits be stick into the Xbin tarball so that they'll be present in /usr/X11R6/lib/X11/pkgreg.tar.gz. The registration tarball is removed upon first use to prevent it later spamming a genuine build from ports if inadvertently extracted again.
* Small update: you need to unpack the stand-alone picobsd tarball intoabial1998-09-291-2/+3
| | | | | | src/release/picobsd, otherwise it won't work. Pointed out by: Randal Masutani <randal@comtest.com>
* Add examples for tree functions.jkh1998-09-296-2/+346
| | | | Submitted by: "Anatoly A. Orehovsky" <tolik@mpeks.tomsk.su>
* Add support for saved trees.jkh1998-09-292-4/+111
| | | | Submitted by: "Anatoly A. Orehovsky" <tolik@mpeks.tomsk.su>
* Make proper use of ${suffix} instead of hardwired suffixes.abial1998-09-294-8/+8
| | | | Pointed out by: Randal Masutani <randal@comtest.com>
* Better handling of "custom" setup.abial1998-09-292-19/+32
| | | | Submitted by: Randal Masutani <randal@comtest.com>
* Once again turn off the BRIDGING - it's not yet ready to be the default...abial1998-09-291-2/+2
| | | | Submitted by: Randal S. Masutani <randal@comtest.com>
* Add sysctl 'machdep.msgbuf_clear'. Setting it to anything causes theabial1998-09-292-8/+50
| | | | | | | kernel message buffer to be cleared. It comes handy in situations when the only logging facility you have is the msgbuf. Reviewed by: jkh
* Bring man page up to dategrog1998-09-292-97/+109
|
* Fixed printf format errors. u_long is not necessarily suitable for castingbde1998-09-291-5/+7
| | | | pointers to, and %d is not suitable for printing uint32_t's.
* FIxed printf format errors (an new one that is only detected on systemsbde1998-09-291-6/+6
| | | | | with pointers smaller than u_longs, and 2 possibly-truncating casts in the same printf).
OpenPOWER on IntegriCloud