]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
21 years agoMark libpam as c99- and WARNS5-clean.
Dag-Erling Smørgrav [Thu, 1 May 2003 14:55:06 +0000 (14:55 +0000)]
Mark libpam as c99- and WARNS5-clean.

21 years agoMake GCC 3.3 STFU about copyright[].
David E. O'Brien [Thu, 1 May 2003 14:45:27 +0000 (14:45 +0000)]
Make GCC 3.3 STFU about copyright[].

21 years agoMark fetch(1) as c99- and WARNS6-clean.
Dag-Erling Smørgrav [Thu, 1 May 2003 14:40:55 +0000 (14:40 +0000)]
Mark fetch(1) as c99- and WARNS6-clean.

21 years agoRemove DIOCGPC98 ioctl.
Yoshihiro Takahashi [Thu, 1 May 2003 14:40:16 +0000 (14:40 +0000)]
Remove DIOCGPC98 ioctl.

21 years agoMark libfetch as c99-clean. Always build at WARNS level 2 rather than
Dag-Erling Smørgrav [Thu, 1 May 2003 14:39:43 +0000 (14:39 +0000)]
Mark libfetch as c99-clean.  Always build at WARNS level 2 rather than
juggling with levels 2 and 3, as this has lead to world breakage for
NOCRYPT users in the past.

21 years agoAdd a mechanism to allow Makefiles to specify the particular C dialect
Dag-Erling Smørgrav [Thu, 1 May 2003 14:36:27 +0000 (14:36 +0000)]
Add a mechanism to allow Makefiles to specify the particular C dialect
in which the source code is written.  This is controlled by the CSTD
variable, which can have one of the following values:

  - "k&r" => -traditional
  - "c89" or "c90" => -std=iso9899:1990
  - "c94" or "c95" => -std=iso9899:199409
  - "c99" => -std=iso9899:1999

The corresponding option is added to CFLAGS regardless of WARNS level.
This also removes -ansi from WARNS level 6, but adds -Wno-long-long to
work around a weird gcc bug (-ansi, which is supposedly equivalent to
-std=iso9899:1990, seems to turn long long warnings off instead of on)

If CSTD is undefined, CFLAGS are unchanged except for the -ansi /
-Wno-long-long change mentioned above for WARNS level 6.

21 years agoOops, DIOCGPC98 should be DIOCSPC98.
Yoshihiro Takahashi [Thu, 1 May 2003 14:30:59 +0000 (14:30 +0000)]
Oops, DIOCGPC98 should be DIOCSPC98.

21 years agoThe -t flag in the retired apmconf.8 is the same as "-h false" in the
Anton Berezin [Thu, 1 May 2003 14:19:00 +0000 (14:19 +0000)]
The -t flag in the retired apmconf.8 is the same as "-h false" in the
apm.8.  Since the -t flag in the apm.8 is something different entirely,
the reference to -t was quite confusing.  Fix that.

21 years agoStart running amd64 builds on 9ball.
Dag-Erling Smørgrav [Thu, 1 May 2003 14:14:54 +0000 (14:14 +0000)]
Start running amd64 builds on 9ball.

21 years agoMerged from src/sbin/fdisk/Makefile revision 1.10.
Yoshihiro Takahashi [Thu, 1 May 2003 13:53:41 +0000 (13:53 +0000)]
Merged from src/sbin/fdisk/Makefile revision 1.10.

21 years ago- Move decoding pc98_partition function into geom_pc98_enc.c.
Yoshihiro Takahashi [Thu, 1 May 2003 13:44:24 +0000 (13:44 +0000)]
- Move decoding pc98_partition function into geom_pc98_enc.c.
- Add encoding pc98_partition function.

21 years agoMerged from src/sbin/fdisk/fdisk.c revision 1.69.
Yoshihiro Takahashi [Thu, 1 May 2003 13:27:11 +0000 (13:27 +0000)]
Merged from src/sbin/fdisk/fdisk.c revision 1.69.

21 years agoUpgrade to Groff 1.19.
Ruslan Ermilov [Thu, 1 May 2003 13:22:21 +0000 (13:22 +0000)]
Upgrade to Groff 1.19.

21 years agoMerge local changes (8 bit clean ctype support) onto mainline.
Ruslan Ermilov [Thu, 1 May 2003 13:18:55 +0000 (13:18 +0000)]
Merge local changes (8 bit clean ctype support) onto mainline.

21 years agoMerge Groff 1.19 changes.
Ruslan Ermilov [Thu, 1 May 2003 13:18:29 +0000 (13:18 +0000)]
Merge Groff 1.19 changes.

21 years agoUse stock (FSF) version of this file.
Ruslan Ermilov [Thu, 1 May 2003 13:17:56 +0000 (13:17 +0000)]
Use stock (FSF) version of this file.

21 years agoRemoved yacc(1) output files.
Ruslan Ermilov [Thu, 1 May 2003 13:17:18 +0000 (13:17 +0000)]
Removed yacc(1) output files.

21 years agoFix to build pc98 boot loader after support amd64.
Yoshihiro Takahashi [Thu, 1 May 2003 13:17:06 +0000 (13:17 +0000)]
Fix to build pc98 boot loader after support amd64.

21 years agoThis commit was generated by cvs2svn to compensate for changes in r114405,
Ruslan Ermilov [Thu, 1 May 2003 13:15:22 +0000 (13:15 +0000)]
This commit was generated by cvs2svn to compensate for changes in r114405,
which included commits to RCS files with non-trunk default branches.

21 years agoRemoved files not present in v1.19 import.
Ruslan Ermilov [Thu, 1 May 2003 13:15:22 +0000 (13:15 +0000)]
Removed files not present in v1.19 import.

21 years agoThis commit was generated by cvs2svn to compensate for changes in r114402,
Ruslan Ermilov [Thu, 1 May 2003 13:09:50 +0000 (13:09 +0000)]
This commit was generated by cvs2svn to compensate for changes in r114402,
which included commits to RCS files with non-trunk default branches.

21 years agoVirgin import of FSF groff v1.19
Ruslan Ermilov [Thu, 1 May 2003 13:09:50 +0000 (13:09 +0000)]
Virgin import of FSF groff v1.19

21 years agoNote when we issue a request to the disk driver, so that we can
Greg Lehey [Thu, 1 May 2003 13:06:00 +0000 (13:06 +0000)]
Note when we issue a request to the disk driver, so that we can
measure the time it spends.

21 years agoFix compiling problem, p_tracee is in my local repository for
David Xu [Thu, 1 May 2003 12:16:06 +0000 (12:16 +0000)]
Fix compiling problem, p_tracee is in my local repository for
threaded process debugging, not ready for this time.

21 years agoUse only a 64bit hash filter table for ADM-Centaur cards like the
Martin Blapp [Thu, 1 May 2003 09:31:01 +0000 (09:31 +0000)]
Use only a 64bit hash filter table for ADM-Centaur cards like the
Accton EN2242 and the ADMtek AN985 cards.

PR: 32699
Submitted by: Jean-Luc Richier <Jean-Luc.Richier@imag.fr>
Reviewed by: phk
MFC after: 2 weeks

21 years agoDrop Giant lock before suspended, pick up it after resumed.
David Xu [Thu, 1 May 2003 07:29:25 +0000 (07:29 +0000)]
Drop Giant lock before suspended, pick up it after resumed.
thread_suspend_check() is used in exit1() which still needs
Giant lock.

21 years agoRemove VINUMDEBUG. vinum(8) always compiles with VINUMDEBUG to be able
Greg Lehey [Thu, 1 May 2003 07:19:29 +0000 (07:19 +0000)]
Remove VINUMDEBUG.  vinum(8) always compiles with VINUMDEBUG to be able
to cope with kernel modules with VINUMDEBUG.

21 years agoDo not attempt to free NULL dinodes (i_din1 or i_din2) in ffs_ifree().
Tim J. Robbins [Thu, 1 May 2003 06:41:59 +0000 (06:41 +0000)]
Do not attempt to free NULL dinodes (i_din1 or i_din2) in ffs_ifree().
These fields can be left as NULL if ffs_vget() allocates an inode but
fails before the dinode memory has been allocated. There are two cases
when this can occur: when we lose a race and another process has added
the inode to the hash, and when reading the inode off disk fails.

The bug was observed by Kris on one of the package-building machines.
See http://marc.theaimsgroup.com/?l=freebsd-current&m=105172731013411&w=2
In Kris's case, it was the bread() that failed because of a disk error.

The alternative to this patch is to ensure that ffs_vget() does not call
vput() when the inode that hasn't been properly initialised.

21 years agoFree i_din2 instead of i_din1 in ffs_ifree() on UFS2 filesystems.
Tim J. Robbins [Thu, 1 May 2003 06:38:27 +0000 (06:38 +0000)]
Free i_din2 instead of i_din1 in ffs_ifree() on UFS2 filesystems.
This is purely a cosmetic change because these members are in a
union together.

21 years agoAdd description of '-o' and '-s'.
Hidetoshi Shimokawa [Thu, 1 May 2003 06:33:13 +0000 (06:33 +0000)]
Add description of '-o' and '-s'.

21 years agoI'm pleased to announce that Promise is now supporting the FreeBSD
Søren Schmidt [Thu, 1 May 2003 06:20:50 +0000 (06:20 +0000)]
I'm pleased to announce that Promise is now supporting the FreeBSD
project by providing documentation (under NDA) and hardware for
testing. This commit is the first result of the cooperation, and
adds support for several of their new controllers that we didn't
support before (and probably newer would have without this arrangement).

Add support for the Promise SATA150 TX2/TX4 and the Promise TX4000
controllers. This also adds support for various motherboard fitted
Promise SATA/ATA chips.
Note that this code uses memory mapped registers to minimize overhead.
I belive FreeBSD has made another first in the Open Source world
by being able to release support for this :)

21 years agoRelax the test for when to use LBA instead of CHS size.
Søren Schmidt [Thu, 1 May 2003 06:17:30 +0000 (06:17 +0000)]
Relax the test for when to use LBA instead of CHS size.

21 years agoFix compile errors on ia64 in dagetcapacity. Set block_len and maxsector
Kenneth D. Merry [Thu, 1 May 2003 05:16:13 +0000 (05:16 +0000)]
Fix compile errors on ia64 in dagetcapacity.  Set block_len and maxsector
to 0 initially.  It seems that the ia64 backend isn't as "smart" as the
i386 backend, which realized that those variables were only set or used
when error == 0, and thus were not used uninitialized.

21 years agoExtend the scope of the vm_object locking in vm_object_split().
Alan Cox [Thu, 1 May 2003 05:06:33 +0000 (05:06 +0000)]
Extend the scope of the vm_object locking in vm_object_split().

21 years agoCommit a missed change to keep in sync with the MI elf loader.
Peter Wemm [Thu, 1 May 2003 04:39:22 +0000 (04:39 +0000)]
Commit a missed change to keep in sync with the MI elf loader.

21 years agoArgh. This was broken by the last-minute elf32/elf64/"elf kernel" changes.
Peter Wemm [Thu, 1 May 2003 04:31:33 +0000 (04:31 +0000)]
Argh.  This was broken by the last-minute elf32/elf64/"elf kernel" changes.

21 years agoOnly use the pc bios stuff on i386's. I think this might even be unused
Peter Wemm [Thu, 1 May 2003 04:23:15 +0000 (04:23 +0000)]
Only use the pc bios stuff on i386's.  I think this might even be unused
there too.

21 years agoAdd AMD64 support to dev/fb. It isn't optimized.
Peter Wemm [Thu, 1 May 2003 04:21:05 +0000 (04:21 +0000)]
Add AMD64 support to dev/fb.  It isn't optimized.

21 years agoUpdate the hardcoded bus tags for early console support for amd64.
Peter Wemm [Thu, 1 May 2003 04:19:19 +0000 (04:19 +0000)]
Update the hardcoded bus tags for early console support for amd64.

21 years agoI changed the numbering of the MODINFOMD_SMAP during the commit, so
Peter Wemm [Thu, 1 May 2003 04:18:02 +0000 (04:18 +0000)]
I changed the numbering of the MODINFOMD_SMAP during the commit, so
recognize the old number for my development boxes so I can use old
loader/pxeboot for a while if I need to.

21 years agoNuke; repocopied to elf32_freebsd.c where it lives on.
Peter Wemm [Thu, 1 May 2003 03:57:19 +0000 (03:57 +0000)]
Nuke; repocopied to elf32_freebsd.c where it lives on.

21 years agoEnable the i386 loader to load and run an amd64 kernel. If this puts
Peter Wemm [Thu, 1 May 2003 03:56:30 +0000 (03:56 +0000)]
Enable the i386 loader to load and run an amd64 kernel.  If this puts
things over floppy size limits, I can exclude it for release builds or
something like that.  Most of the changes are to get the load_elf.c file
into a seperate elf32_ or elf64_ namespace so that you can have two
ELF loaders present at once.  Note that for 64 bit kernels, it actually
starts up the kernel already in 64 bit mode with paging enabled.  This
is really easy because we have a known minimum feature set.

Of note is that for amd64, we have to pass in the bios int 15 0xe821
memory map because once in long mode, you absolutely cannot make VM86
calls.  amd64 does not use 'struct bootinfo' at all.  It is a pure loader
metadata startup, just like sparc64 and powerpc.  Much of the
infrastructure to support this was adapted from sparc64.

21 years agoLock an update to a vm_object's ref_count.
Alan Cox [Thu, 1 May 2003 03:51:05 +0000 (03:51 +0000)]
Lock an update to a vm_object's ref_count.

21 years agoHack to enable getting two different elf32 and elf64 loaders in the
Peter Wemm [Thu, 1 May 2003 03:46:12 +0000 (03:46 +0000)]
Hack to enable getting two different elf32 and elf64 loaders in the
same i386 loader binary.

21 years ago<b30> is 'IA64' - ie: you're running on an ia64 in 32 bit mode.
Peter Wemm [Thu, 1 May 2003 03:44:40 +0000 (03:44 +0000)]
<b30> is 'IA64' - ie: you're running on an ia64 in 32 bit mode.

21 years agoFix printf warning caused by recent CAM change.
Hidetoshi Shimokawa [Thu, 1 May 2003 03:38:04 +0000 (03:38 +0000)]
Fix printf warning caused by recent CAM change.

21 years agoBack out last commits. The elf64/elf32 kernel name thing was more pain
Peter Wemm [Thu, 1 May 2003 03:33:28 +0000 (03:33 +0000)]
Back out last commits.  The elf64/elf32 kernel name thing was more pain
than it was worth.

21 years agoSlight reorg and added AMD64 support. A couple of the MODINFOMD_* values
Peter Wemm [Thu, 1 May 2003 03:31:18 +0000 (03:31 +0000)]
Slight reorg and added AMD64 support.  A couple of the MODINFOMD_* values
that were added to sparc64 and later powerpc, really should have been in
the MI area.  But changing that now with insufficient preperation will
just cause too much pain.

Move MD_FETCH() to the MI sys/linker.h file to avoid another two copies
of it.

21 years ago - Update the vm_object locking in vm_object_reference().
Alan Cox [Thu, 1 May 2003 03:29:20 +0000 (03:29 +0000)]
 - Update the vm_object locking in vm_object_reference().
 - Convert some dead code in vm_object_reference() into a comment.

21 years agoLock accesses to the vm_object's ref_count and resident_page_count.
Alan Cox [Thu, 1 May 2003 03:10:38 +0000 (03:10 +0000)]
Lock accesses to the vm_object's ref_count and resident_page_count.

21 years agoSync up with the files in the hammer branch in the p4 tree to get basic
Peter Wemm [Thu, 1 May 2003 02:59:24 +0000 (02:59 +0000)]
Sync up with the files in the hammer branch in the p4 tree to get basic
AMD64 support.  There is still more to add.

21 years agoAdd the -j and -y options to the synopsis.
Tim J. Robbins [Thu, 1 May 2003 02:37:46 +0000 (02:37 +0000)]
Add the -j and -y options to the synopsis.

21 years agoFlush streams before calling system() so that the output appears in the
Tim J. Robbins [Thu, 1 May 2003 02:36:27 +0000 (02:36 +0000)]
Flush streams before calling system() so that the output appears in the
right place in the output stream when redirected to a file (when full
buffering is enabled). Noticed by schweikh on freebsd-standards.

21 years agoDon't emulate a MBR by handling the MBR::type attribute. It is
Marcel Moolenaar [Thu, 1 May 2003 01:51:28 +0000 (01:51 +0000)]
Don't emulate a MBR by handling the MBR::type attribute. It is
not needed at all. The BSD class will attach to a GPT class without
it.

21 years agoAdd the #includes previously in individual .c files.
Greg Lehey [Thu, 1 May 2003 01:40:16 +0000 (01:40 +0000)]
Add the #includes previously in individual .c files.

Change name of history file to History to avoid name conflicts.

21 years agoMove most #includes to vext.h.
Greg Lehey [Thu, 1 May 2003 01:39:49 +0000 (01:39 +0000)]
Move most #includes to vext.h.

Change name of history file to History to avoid name conflicts.

Don't try to make devices unless devfs has been removed.

Don't accept resetconfig command from a file.

Abused by:  Jens Schweikhardt <schweikh@schweikhardt.net>

Remove dead code (#if 0)

vinum_mirror: Don't try to create mirrors with 0 drives.

Found by: mismatch between enum kw in two different files.

21 years agoMove most #includes to vext.h.
Greg Lehey [Thu, 1 May 2003 01:39:42 +0000 (01:39 +0000)]
Move most #includes to vext.h.

Change name of history file to History to avoid name conflicts.

Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

Don't try to chown directories if they haven't been created.

21 years agoMove most #includes to vext.h.
Greg Lehey [Thu, 1 May 2003 01:38:55 +0000 (01:38 +0000)]
Move most #includes to vext.h.

vinum_info: Use increasing verbosity to show time spent in disk
    driver.

Correct a comment.

21 years agoRewrite minor number decoding. Now we have only three types of
Greg Lehey [Thu, 1 May 2003 01:35:03 +0000 (01:35 +0000)]
Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

21 years agoFor consistency's sake, on command failure, throw 1, not -1.
Greg Lehey [Thu, 1 May 2003 01:34:38 +0000 (01:34 +0000)]
For consistency's sake, on command failure, throw 1, not -1.

Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

21 years agoRewrite minor number decoding. Now we have only three types of
Greg Lehey [Thu, 1 May 2003 01:34:05 +0000 (01:34 +0000)]
Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

Correct formats for some error messages.  Don't cast the value to
match the format.

21 years agoRewrite minor number decoding. Now we have only three types of
Greg Lehey [Thu, 1 May 2003 01:33:34 +0000 (01:33 +0000)]
Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

Tidy up comments.

Check for null rqgs.  This continue to be reported, though I can't
work out why.

Correct formats for some error messages.  Don't cast the value to
match the format.

Use microtime, not getmicrotime, for timing debug entries.

21 years agoTerminate a comment.
Greg Lehey [Thu, 1 May 2003 01:33:01 +0000 (01:33 +0000)]
Terminate a comment.

21 years agoDon't make definition of kw_debug dependent on VINUMDEBUG. In
Greg Lehey [Thu, 1 May 2003 01:32:31 +0000 (01:32 +0000)]
Don't make definition of kw_debug dependent on VINUMDEBUG.  In
userland, define the keyword even if the kernel module doesn't have
debug code.

21 years agoUse strrchr, not rindex. This is for compatibility with other
Greg Lehey [Thu, 1 May 2003 01:32:08 +0000 (01:32 +0000)]
Use strrchr, not rindex.  This is for compatibility with other
systems, and requires a #define.

21 years ago#ifdef VINUMDEBUG, not #if VINUMDEBUG.
Greg Lehey [Thu, 1 May 2003 01:31:48 +0000 (01:31 +0000)]
#ifdef VINUMDEBUG, not #if VINUMDEBUG.

21 years agoDon't make definition of kw_debug dependent on VINUMDEBUG. It's only
Greg Lehey [Thu, 1 May 2003 01:31:20 +0000 (01:31 +0000)]
Don't make definition of kw_debug dependent on VINUMDEBUG.  It's only
an enum value, and dropping it can lead to some spectacular surprises
in userland.

21 years agoRewrite minor number decoding. Now we have only three types of
Greg Lehey [Thu, 1 May 2003 01:30:59 +0000 (01:30 +0000)]
Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

As a result of the minor number changes, split out the superdevice
handling into a separate function, vinum_super_ioctl.  This was most
of the code of vinumioctl.

attachobject: Improve error checking.

21 years agoUse microtime, not getmicrotime, for timing debug entries.
Greg Lehey [Thu, 1 May 2003 01:30:31 +0000 (01:30 +0000)]
Use microtime, not getmicrotime, for timing debug entries.

init_drive: Rephrase error message text.
    Remove dead code (inside #if 0).

Change name of find_drive_by_dev to the more descriptive
find_drive_by_name.

Tidy up comments.

21 years agoRemove "to do" comments.
Greg Lehey [Thu, 1 May 2003 01:29:52 +0000 (01:29 +0000)]
Remove "to do" comments.

get_emppty_drive: Fix a day one bug with strcpy parameters.

Change name of find_drive_by_dev to the more descriptive
find_drive_by_name.

Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

21 years agoRewrite minor number decoding. Now we have only three types of
Greg Lehey [Thu, 1 May 2003 01:28:42 +0000 (01:28 +0000)]
Rewrite minor number decoding.  Now we have only three types of
object: subdisks, plexes and volumes.  The encoding for plexes and
subdisks no longer reflects the object to which they belong.  The
super devices are high-order volume numbers.  This gives vastly more
potential volumes (4 million instead of 256).

Remove an unnecessary goto.

vinumopen: Return EINVAL, not ENXIO, on an attempt to open a
referenced plex.

21 years agoCommit MD parts of a loosely functional AMD64 port. This is based on
Peter Wemm [Thu, 1 May 2003 01:05:25 +0000 (01:05 +0000)]
Commit MD parts of a loosely functional AMD64 port.  This is based on
a heavily stripped down FreeBSD/i386 (brutally stripped down actually) to
attempt to get a stable base to start from.  There is a lot missing still.
Worth noting:
- The kernel runs at 1GB in order to cheat with the pmap code.  pmap uses
  a variation of the PAE code in order to avoid having to worry about 4
  levels of page tables yet.
- It boots in 64 bit "long mode" with a tiny trampoline embedded in the
  i386 loader.  This simplifies locore.s greatly.
- There are still quite a few fragments of i386-specific code that have
  not been translated yet, and some that I cheated and wrote dumb C
  versions of (bcopy etc).
- It has both int 0x80 for syscalls (but using registers for argument
  passing, as is native on the amd64 ABI), and the 'syscall' instruction
  for syscalls.  int 0x80 preserves all registers, 'syscall' does not.
- I have tried to minimize looking at the NetBSD code, except in a couple
  of places (eg: to find which register they use to replace the trashed
  %rcx register in the syscall instruction).  As a result, there is not a
  lot of similarity.  I did look at NetBSD a few times while debugging to
  get some ideas about what I might have done wrong in my first attempt.

21 years agoKPT_MIN_ADDRESS and KPT_MAX_ADDRESS are not used anywhere. And if they
Peter Wemm [Thu, 1 May 2003 00:10:38 +0000 (00:10 +0000)]
KPT_MIN_ADDRESS and KPT_MAX_ADDRESS are not used anywhere.  And if they
were, they are not safe to use outside of the kernel since these values
can change at kernel compile time - ie: we do not want them compiled into
userland binaries.

21 years agoKill MID_MACHINE, its a.out specific, the only platform that supports
Marcel Moolenaar [Wed, 30 Apr 2003 23:16:33 +0000 (23:16 +0000)]
Kill MID_MACHINE, its a.out specific, the only platform that supports
it is i386. All of the other platforms should remove it too.
-- peter@

21 years agoRepocopy from x86_64/... to amd64/...
Peter Wemm [Wed, 30 Apr 2003 22:51:59 +0000 (22:51 +0000)]
Repocopy from x86_64/... to amd64/...
Rename visible x86_64 references to amd64.
Kill MID_MACHINE, its a.out specific, the only platform that supports it
is i386.  All of the other platforms should remove it too.

21 years agoAdd AMD64 hooks
Peter Wemm [Wed, 30 Apr 2003 22:22:29 +0000 (22:22 +0000)]
Add AMD64 hooks

21 years agoAdd support for IC Book Ironclad Pro/Lite 8-port cards.
Maxim Sobolev [Wed, 30 Apr 2003 22:15:47 +0000 (22:15 +0000)]
Add support for IC Book Ironclad Pro/Lite 8-port cards.

Sponsired by: IC Book Labs
MFC After: 2 weeks

21 years agoWe use i386 boot code on AMD64.
Peter Wemm [Wed, 30 Apr 2003 22:13:36 +0000 (22:13 +0000)]
We use i386 boot code on AMD64.

21 years agoFix transcription error. Use == NULL, not != NULL. Fortunately this
Peter Wemm [Wed, 30 Apr 2003 22:09:26 +0000 (22:09 +0000)]
Fix transcription error.  Use == NULL, not != NULL.  Fortunately this
was harmless.

21 years agoUse the 64 bit sized struct kinfo_proc for AMD64.
Peter Wemm [Wed, 30 Apr 2003 22:06:39 +0000 (22:06 +0000)]
Use the 64 bit sized struct kinfo_proc for AMD64.

21 years agoLook for an elf32 kernel (powerpc) and elf64 kernel (sparc64) as well
Peter Wemm [Wed, 30 Apr 2003 22:05:48 +0000 (22:05 +0000)]
Look for an elf32 kernel (powerpc) and elf64 kernel (sparc64) as well
as a plain "elf kernel".

21 years agoACPI will always be present on AMD64 - it will never be an autodetect
Peter Wemm [Wed, 30 Apr 2003 22:02:39 +0000 (22:02 +0000)]
ACPI will always be present on AMD64 - it will never be an autodetect
module.

21 years agoAlso look for an "elf64 kernel" (for sparc64) and "elf32 kernel" (for
Peter Wemm [Wed, 30 Apr 2003 22:00:16 +0000 (22:00 +0000)]
Also look for an "elf64 kernel" (for sparc64) and "elf32 kernel" (for
powerpc) when building metadata.

21 years agoThe PAM module pam_krb5 does not have "session" capabilities.
Mark Murray [Wed, 30 Apr 2003 21:57:54 +0000 (21:57 +0000)]
The PAM module pam_krb5 does not have "session" capabilities.
Don't give examples of such use, this is bogus.

21 years agoAMD64 uses the new-style cpu_switch()/cpu_throw() calling conventions.
Peter Wemm [Wed, 30 Apr 2003 21:45:03 +0000 (21:45 +0000)]
AMD64 uses the new-style cpu_switch()/cpu_throw() calling conventions.

21 years agoCreate a 'legacy' node for AMD64 as well as i386. While we'll never
Peter Wemm [Wed, 30 Apr 2003 21:41:41 +0000 (21:41 +0000)]
Create a 'legacy' node for AMD64 as well as i386.  While we'll never
have to use it since all AMD64 machines are supposed to have acpi etc,
I'm using it during development so I can avoid the acpi code for now.
Yes, this is cheating.

21 years agoEliminate a compiler warning with gcc3.3 on AMD64, where speed is
Peter Wemm [Wed, 30 Apr 2003 21:39:28 +0000 (21:39 +0000)]
Eliminate a compiler warning with gcc3.3 on AMD64, where speed is
a 32 bit int which can never be > ULONG_MAX / 8.  Its an 'always true'
warning.

21 years agoMore factual updates for kerberos 5.
Mark Murray [Wed, 30 Apr 2003 21:15:41 +0000 (21:15 +0000)]
More factual updates for kerberos 5.

21 years agoRemove 80386 bandaids from code repocopied from i386. rtld_start.S still
Peter Wemm [Wed, 30 Apr 2003 21:09:06 +0000 (21:09 +0000)]
Remove 80386 bandaids from code repocopied from i386.  rtld_start.S still
todo.

21 years agoAMD64 support (another IEEEFP platform)
Peter Wemm [Wed, 30 Apr 2003 21:06:30 +0000 (21:06 +0000)]
AMD64 support (another IEEEFP platform)

21 years agoAMD64 support; repocopied from i386
Peter Wemm [Wed, 30 Apr 2003 21:05:33 +0000 (21:05 +0000)]
AMD64 support; repocopied from i386

21 years agoTeach libdisk that AMD64 works just like i386
Peter Wemm [Wed, 30 Apr 2003 21:03:16 +0000 (21:03 +0000)]
Teach libdisk that AMD64 works just like i386

21 years agoMake the defaults for Kerberos 5 a little more up-to-date.
Mark Murray [Wed, 30 Apr 2003 20:58:49 +0000 (20:58 +0000)]
Make the defaults for Kerberos 5 a little more up-to-date.

21 years agoReplace beforeinstall: with plain FILES.
Ruslan Ermilov [Wed, 30 Apr 2003 20:48:03 +0000 (20:48 +0000)]
Replace beforeinstall: with plain FILES.

21 years agoChange the name of the kadmind binary to match reality, now that
Mark Murray [Wed, 30 Apr 2003 20:42:41 +0000 (20:42 +0000)]
Change the name of the kadmind binary to match reality, now that
KerberosIV is no longer an issue.

21 years agoStyle fixes to struct sigacts member comments. Sort function prototypes.
John Baldwin [Wed, 30 Apr 2003 19:57:21 +0000 (19:57 +0000)]
Style fixes to struct sigacts member comments.  Sort function prototypes.

21 years agoForgot to remove Giant around call to kern_sigaction() in
John Baldwin [Wed, 30 Apr 2003 19:45:13 +0000 (19:45 +0000)]
Forgot to remove Giant around call to kern_sigaction() in
freebsd4_sigaction() in revision 1.232.

21 years agoAxe a stale comment.
John Baldwin [Wed, 30 Apr 2003 19:41:04 +0000 (19:41 +0000)]
Axe a stale comment.

21 years agoAdd __amd64__ to the list of things that use IEEEFP
Peter Wemm [Wed, 30 Apr 2003 19:32:25 +0000 (19:32 +0000)]
Add __amd64__ to the list of things that use IEEEFP

21 years agoTell malloc.c that AMD64 uses the same pagesize as i386.
Peter Wemm [Wed, 30 Apr 2003 19:30:34 +0000 (19:30 +0000)]
Tell malloc.c that AMD64 uses the same pagesize as i386.