dragonfly.git
15 years agoMerge from vendor branch GCC:
Simon Schubert [Sun, 5 Jun 2005 18:26:15 +0000 (18:26 +0000)]
Merge from vendor branch GCC:
Import of virgin gcc 4.0.0 distribution.
Unneccessary files and dirs have been deleted and recorded in
README.DELETED.

15 years agoImport of virgin gcc 4.0.0 distribution.
Simon Schubert [Sun, 5 Jun 2005 18:26:15 +0000 (18:26 +0000)]
Import of virgin gcc 4.0.0 distribution.
Unneccessary files and dirs have been deleted and recorded in
README.DELETED.

15 years agoMake call to arc_ioctl the default case.
Joerg Sonnenberger [Sun, 5 Jun 2005 16:05:33 +0000 (16:05 +0000)]
Make call to arc_ioctl the default case.

15 years agoMerge the pointer to the link-layer address into ifnet and remove
Joerg Sonnenberger [Sun, 5 Jun 2005 12:35:24 +0000 (12:35 +0000)]
Merge the pointer to the link-layer address into ifnet and remove
ifnet_addrs. Change the rest of the code accordingly.

15 years agounbreak world: only include sys/thread2.h in kernel, not in userland
Simon Schubert [Sun, 5 Jun 2005 12:17:46 +0000 (12:17 +0000)]
unbreak world: only include sys/thread2.h in kernel, not in userland

15 years agoAdd f_owner (user who mounted the filesystem), f_type (filesystem type ID),
Joerg Sonnenberger [Sun, 5 Jun 2005 12:16:53 +0000 (12:16 +0000)]
Add f_owner (user who mounted the filesystem), f_type (filesystem type ID),
f_(a)sync(reads,writes) (counters for sync and async read / writes).
Also add some spare fields for latter additions.

15 years agoAdd include file protection wrapper.
Jeroen Ruigrok/asmodai [Sun, 5 Jun 2005 09:40:46 +0000 (09:40 +0000)]
Add include file protection wrapper.

Wrap function definition in #if 0/#endif until they are added.

15 years agoInstead of using ifnet_addrs and following ifa_ifp, use ifindex2ifnet
Joerg Sonnenberger [Sun, 5 Jun 2005 00:25:10 +0000 (00:25 +0000)]
Instead of using ifnet_addrs and following ifa_ifp, use ifindex2ifnet
directly.

15 years ago- int -> pid_t for pid
Liam J. Foy [Sat, 4 Jun 2005 20:35:06 +0000 (20:35 +0000)]
- int -> pid_t for pid

15 years agoInstead of checking for ifnet_addrs[ifp->index - 1] == NULL to detect
Joerg Sonnenberger [Sat, 4 Jun 2005 14:41:57 +0000 (14:41 +0000)]
Instead of checking for ifnet_addrs[ifp->index - 1] == NULL to detect
running from if_detach, introduce in_ifdetach and call that after the
address list has been processed. This reduces the differences between
IPv4 and IPv6 stack.

15 years agoconvert spl*() to critical sections.
Simon Schubert [Sat, 4 Jun 2005 14:25:45 +0000 (14:25 +0000)]
convert spl*() to critical sections.

15 years ago * Include string.h to get strlen() prototype.
Eirik Nygaard [Sat, 4 Jun 2005 13:42:28 +0000 (13:42 +0000)]
 * Include string.h to get strlen() prototype.
 * Shut up gcc warning.

15 years agoThe acpi module was failing to load due to exposed crit_*() functions. Add
Matthew Dillon [Sat, 4 Jun 2005 03:22:08 +0000 (03:22 +0000)]
The acpi module was failing to load due to exposed crit_*() functions.  Add
a missing #include to get the crit_*() inlines.

Found by: Matt Dillon when his power button stopped working, click click click!

15 years agoReplace cpu_mb1() and cpu_mb2() with cpu_mfence(), cpu_lfence(), cpu_sfence(),
Matthew Dillon [Fri, 3 Jun 2005 23:57:34 +0000 (23:57 +0000)]
Replace cpu_mb1() and cpu_mb2() with cpu_mfence(), cpu_lfence(), cpu_sfence(),
and cpu_ccfence().  These provide memory and compiler fences to guarentee
read/write ordering of memory and to prevent the compiler itself from
generating reordered code in particular cases.

Most of the cases where care must be taken are tail-chasing FIFO cases,
especially in the IPI messaging code.

At the moment on SMP boxes we use a locked bus cycle on (%%esp) for the
lfence and mfence.  The only other choices are to use a cpuid instruction
or one of the {S,L,M}FENCE instructions.  cpuid is usually horrible, and
the *FENCE instructions do not exist on older cpus.

Linux seems to use the locked bus cycle or *FENCE instruction method.

Although I have no conclusive evidence, a number of crash dumps provided
by David Rhodus has led me to believe that speculative reads by modern cpus,
in particular in HTT situations with Intel cpus, can survive many more
instructions then previously believed.  The only safe solution is to use
an instruction sequence which guarentees proper operation.

I would prefer to avoid the use of a locked bus cycle but at least in the
IPIQ case (the most common case we have to worry about), only one locked
bus cycle is required and then the entire IPIQ can be processed without
further locked cycles.

15 years agoAnother instance of IF_LLSOCKADDR.
Joerg Sonnenberger [Fri, 3 Jun 2005 23:31:36 +0000 (23:31 +0000)]
Another instance of IF_LLSOCKADDR.

15 years agoAdd a new macro IF_LLSOCKADDR which maps a ifnet pointer to the
Joerg Sonnenberger [Fri, 3 Jun 2005 23:23:03 +0000 (23:23 +0000)]
Add a new macro IF_LLSOCKADDR which maps a ifnet pointer to the
associated link layer sockaddr. Use it to implement IF_LLADDR.
Replace most references of ifnet_addrs with one of those macros.

This makes the sometimes explicitly encoded assumption of each
interface having a link-layer address implicit and is the
preparation for removing ifnet_addrs.

15 years agoExpand itjc_bus_setup, it declares variables.
Joerg Sonnenberger [Fri, 3 Jun 2005 23:14:26 +0000 (23:14 +0000)]
Expand itjc_bus_setup, it declares variables.

15 years agoRename label to not collidate with local variable. Conditionalize error,
Joerg Sonnenberger [Fri, 3 Jun 2005 23:10:23 +0000 (23:10 +0000)]
Rename label to not collidate with local variable. Conditionalize error,
it's not yet used. Use #ifdef instead of depending on 0 for undefined
macros.

15 years agoInclude sys/thread2.h to unbreak build.
Joerg Sonnenberger [Fri, 3 Jun 2005 23:06:31 +0000 (23:06 +0000)]
Include sys/thread2.h to unbreak build.

15 years agoAdd missing parameter.
Joerg Sonnenberger [Fri, 3 Jun 2005 23:05:22 +0000 (23:05 +0000)]
Add missing parameter.

15 years agovm_contig_pg_free() must busy the page before freeing it in the case
Matthew Dillon [Fri, 3 Jun 2005 22:58:09 +0000 (22:58 +0000)]
vm_contig_pg_free() must busy the page before freeing it in the case
of a contigmalloc failure.

15 years agoreorder declarations
Joerg Sonnenberger [Fri, 3 Jun 2005 22:57:56 +0000 (22:57 +0000)]
reorder declarations

15 years agoAugment the panic when attempting to switch from a FAST interrupt to include
Matthew Dillon [Fri, 3 Jun 2005 22:57:27 +0000 (22:57 +0000)]
Augment the panic when attempting to switch from a FAST interrupt to include
the thread address, since it may not be readily apparent from the kernel
dump.

15 years agoUse IF_DRAIN.
Joerg Sonnenberger [Fri, 3 Jun 2005 22:56:26 +0000 (22:56 +0000)]
Use IF_DRAIN.

15 years agoAdd a sysctl, debug.use_malloc_pattern, that explicitly initializes data
Matthew Dillon [Fri, 3 Jun 2005 22:55:58 +0000 (22:55 +0000)]
Add a sysctl, debug.use_malloc_pattern, that explicitly initializes data
buffers allocated WITHOUT M_ZERO to -1 rather then leaving prior garbage
intact.  Works with INVARIANTS builds only.

15 years agoRemove spl*() in disk/{ata,buslogic,ccd} and replace them with
Sascha Wildner [Fri, 3 Jun 2005 21:56:23 +0000 (21:56 +0000)]
Remove spl*() in disk/{ata,buslogic,ccd} and replace them with
critical sections.

15 years agoDocument thread-safety.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 21:28:49 +0000 (21:28 +0000)]
Document thread-safety.

15 years agoDon't commit half of a change. Prefix parameter names and local arguments
Joerg Sonnenberger [Fri, 3 Jun 2005 21:19:09 +0000 (21:19 +0000)]
Don't commit half of a change. Prefix parameter names and local arguments
with _ in inline functions.

15 years agocpu_mb2() needs to load %eax with $0 before running cpuid.
Matthew Dillon [Fri, 3 Jun 2005 20:20:44 +0000 (20:20 +0000)]
cpu_mb2() needs to load %eax with $0 before running cpuid.

15 years agoDon't name arguments in prototypes. Nuke __STDC__ conditional.
Joerg Sonnenberger [Fri, 3 Jun 2005 20:16:28 +0000 (20:16 +0000)]
Don't name arguments in prototypes. Nuke __STDC__ conditional.

15 years agoRemove *spl() from netinet6 replacing them with critical sections.
Eirik Nygaard [Fri, 3 Jun 2005 19:56:08 +0000 (19:56 +0000)]
Remove *spl() from netinet6 replacing them with critical sections.

15 years agoRemove spl*() in net/{altq,bpf.c,bridge,dummynet,ef,gif,gre,hostcache.c}
Sascha Wildner [Fri, 3 Jun 2005 18:20:37 +0000 (18:20 +0000)]
Remove spl*() in net/{altq,bpf.c,bridge,dummynet,ef,gif,gre,hostcache.c}
and replace them with critical sections.

[forced commit to correct message]

15 years agoRemove spl*() in net/{altq,bpf.c,bridge,dummynet,ef,gif,gre,hostcache.c}
Sascha Wildner [Fri, 3 Jun 2005 18:04:14 +0000 (18:04 +0000)]
Remove spl*() in net/{altq,bpf.c,bridge,dummynet,ef,gif,gre,hostcache.c}
netgraph and replace them with critical sections.

15 years agoRemove spl*() calls from i386, replacing them with critical sections.
Matthew Dillon [Fri, 3 Jun 2005 17:14:51 +0000 (17:14 +0000)]
Remove spl*() calls from i386, replacing them with critical sections.
Leave spl support intact for the moment (it will be removed soon).  Adjust
the interrupt mux to use a critical section for 'old' interrupt handlers
not using the new serialization API (which is nearly all of them at the
moment).

15 years agoRemove *spl() from dev/disk/{advansys,aha,ahb,aic7xxx,amd} replacing them with
Eirik Nygaard [Fri, 3 Jun 2005 16:57:13 +0000 (16:57 +0000)]
Remove *spl() from dev/disk/{advansys,aha,ahb,aic7xxx,amd} replacing them with
critical sections.

15 years agoRemove spl*() calls from net/i4b, replacing them with critical sections.
Matthew Dillon [Fri, 3 Jun 2005 16:50:13 +0000 (16:50 +0000)]
Remove spl*() calls from net/i4b, replacing them with critical sections.

15 years agoMake -I only apply to rm's run in the foreground. Silently discard it if
Matthew Dillon [Fri, 3 Jun 2005 16:00:23 +0000 (16:00 +0000)]
Make -I only apply to rm's run in the foreground.  Silently discard it if
the rm is run in the background.

Reported-by: "George Georgalis" <george@galis.org>
15 years agoBump FreeBSD identifier to 1.16 to signal to which version we synchronised.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 11:54:37 +0000 (11:54 +0000)]
Bump FreeBSD identifier to 1.16 to signal to which version we synchronised.

15 years agoFix last two return() calls to comply to style.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 11:51:33 +0000 (11:51 +0000)]
Fix last two return() calls to comply to style.
Explicitly kept from previous commit to not confuse content and style.

15 years agoSync to FreeBSD 1.14/1.15:
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 11:50:17 +0000 (11:50 +0000)]
Sync to FreeBSD 1.14/1.15:

Use nanoseconds and then lexicographic ordering when the seconds of the
[acm]time are the same.

Written by: dwmalone

Changed to comply to our style.

15 years agoAdd first stab at a statvfs.h.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 11:29:32 +0000 (11:29 +0000)]
Add first stab at a statvfs.h.

It most likely needs to cut back on the sys/types.h inclusion to limit
namespace pollution.

15 years agoAdd id_t, a general identifier type, per IEEE Std 1003.1, 2004 Edition.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 11:07:34 +0000 (11:07 +0000)]
Add id_t, a general identifier type, per IEEE Std 1003.1, 2004 Edition.
It should be able to hold a gid_t/pid_t/uid_t.

15 years agoAdd blkcnt_t and blksize_t per IEEE Std 1003.1, 2004 Edition.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 09:45:53 +0000 (09:45 +0000)]
Add blkcnt_t and blksize_t per IEEE Std 1003.1, 2004 Edition.

I do not understand why both FreeBSD and NetBSD use an unsigned integer,
when the specification clearly states an signed integer should be used.

15 years agoAdd fsblkcnt_t and fsfilcnt_t.
Jeroen Ruigrok/asmodai [Fri, 3 Jun 2005 09:17:36 +0000 (09:17 +0000)]
Add fsblkcnt_t and fsfilcnt_t.
Part of IEEE Std 1003.1, 2004 Edition.

15 years agoCovert netproto/ipsec into using critical sections instead of SPL ops.
Hiten Pandya [Fri, 3 Jun 2005 00:22:27 +0000 (00:22 +0000)]
Covert netproto/ipsec into using critical sections instead of SPL ops.

15 years agoRemove spl*() calls from netinet, replacing them with critical sections.
Matthew Dillon [Thu, 2 Jun 2005 23:52:42 +0000 (23:52 +0000)]
Remove spl*() calls from netinet, replacing them with critical sections.
A slight rearrangement of COMMON_START() in tcp_usrreq.c was necessary to
ensure that the inp is loaded after  entering the critical section.

15 years agoCorrect typo in comment for vshiftl().
Hiten Pandya [Thu, 2 Jun 2005 23:50:30 +0000 (23:50 +0000)]
Correct typo in comment for vshiftl().

15 years agoRemove spl*() calls from the netproto/atm driver, replacing them with
Matthew Dillon [Thu, 2 Jun 2005 22:37:52 +0000 (22:37 +0000)]
Remove spl*() calls from the netproto/atm driver, replacing them with
critical sections.

15 years agoRemove spl*() calls in netgraph and replace them with critical sections.
Sascha Wildner [Thu, 2 Jun 2005 22:11:46 +0000 (22:11 +0000)]
Remove spl*() calls in netgraph and replace them with critical sections.

15 years agoAfter some thought, replace the splhigh()/spl0() combination in swapout()
Matthew Dillon [Thu, 2 Jun 2005 22:03:49 +0000 (22:03 +0000)]
After some thought, replace the splhigh()/spl0() combination in swapout()
with a critical section.

15 years agoFix a warning by conditionalizing a forward declaration.
Matthew Dillon [Thu, 2 Jun 2005 21:55:22 +0000 (21:55 +0000)]
Fix a warning by conditionalizing a forward declaration.

15 years agoRemove variable names from procedure declarations so userland doesn't
Matthew Dillon [Thu, 2 Jun 2005 21:51:31 +0000 (21:51 +0000)]
Remove variable names from procedure declarations so userland doesn't
trip up on the header file.

15 years agoConditionalize thread2.h so /usr/src/usr.bin/fstat can include vm_object.h.
Matthew Dillon [Thu, 2 Jun 2005 21:51:00 +0000 (21:51 +0000)]
Conditionalize thread2.h so /usr/src/usr.bin/fstat can include vm_object.h.

15 years agoRemove spl*() calls from the crypto device driver, replacing them with
Matthew Dillon [Thu, 2 Jun 2005 21:40:55 +0000 (21:40 +0000)]
Remove spl*() calls from the crypto device driver, replacing them with
critical sections.

15 years agoRemove spl*() calls from the atm code, replacing them with critical sections.
Matthew Dillon [Thu, 2 Jun 2005 21:36:09 +0000 (21:36 +0000)]
Remove spl*() calls from the atm code, replacing them with critical sections.

15 years agoConditionalize an #include so libcam can build the file.
Matthew Dillon [Thu, 2 Jun 2005 21:12:45 +0000 (21:12 +0000)]
Conditionalize an #include so libcam can build the file.

15 years agoRemove spl*() calls and replace them with critical sections.
Sascha Wildner [Thu, 2 Jun 2005 20:57:21 +0000 (20:57 +0000)]
Remove spl*() calls and replace them with critical sections.

15 years agoRemove spl*() calls from the bus/ infrastructure, replacing them with
Matthew Dillon [Thu, 2 Jun 2005 20:41:05 +0000 (20:41 +0000)]
Remove spl*() calls from the bus/ infrastructure, replacing them with
critical sections.  Remove splusb() from everywhere, replacing it with
critical sections.

15 years ago<sys/vmmeter.h> need not be included twice.
Sascha Wildner [Thu, 2 Jun 2005 19:57:07 +0000 (19:57 +0000)]
<sys/vmmeter.h> need not be included twice.

15 years agoANSIfy.
Jeroen Ruigrok/asmodai [Thu, 2 Jun 2005 18:45:06 +0000 (18:45 +0000)]
ANSIfy.

15 years agoDetail thread-safety conformance.
Jeroen Ruigrok/asmodai [Thu, 2 Jun 2005 18:44:07 +0000 (18:44 +0000)]
Detail thread-safety conformance.

15 years agoClean up type-o's.
Matthew Dillon [Thu, 2 Jun 2005 16:20:51 +0000 (16:20 +0000)]
Clean up type-o's.

15 years agoUse volatile sig_atomic_t variables and async-signal-safe functions for
Sascha Wildner [Thu, 2 Jun 2005 16:16:37 +0000 (16:16 +0000)]
Use volatile sig_atomic_t variables and async-signal-safe functions for
signal handling.

Submitted by: Sepherosa Ziehau <sepherosa@softhome.net>

15 years agoOnly do the OS version requirements check if DESTDIR is non-existant,
Matthew Dillon [Thu, 2 Jun 2005 06:08:07 +0000 (06:08 +0000)]
Only do the OS version requirements check if DESTDIR is non-existant,
empty, or "/".  Otherwise we cannot build nrelease on boxes with older
kernels.

Reported-by: "Simon 'corecode' Schubert" <corecode@fs.ei.tum.de>
15 years agoFix a bugaboo in the last commit. The wildcard patterns were not accounting
Matthew Dillon [Wed, 1 Jun 2005 22:59:14 +0000 (22:59 +0000)]
Fix a bugaboo in the last commit.  The wildcard patterns were not accounting
for kernel versions older then 1.3.1.

15 years agoFix a recursive clock_lock() on SMP systems which was deadlocking on boot.
Matthew Dillon [Wed, 1 Jun 2005 22:55:19 +0000 (22:55 +0000)]
Fix a recursive clock_lock() on SMP systems which was deadlocking on boot.
(related to the new system timer API).

15 years agoAdd a simple API tha allows the interrupt timer to efficiently convert
Matthew Dillon [Wed, 1 Jun 2005 22:25:14 +0000 (22:25 +0000)]
Add a simple API tha allows the interrupt timer to efficiently convert
system hardware ticks (e.g. ACPI at 3x the frequency) to interrupt timer
ticks (e.g. 8254 timer 0).

Remove the sysctl code that allows the system hardware timer to be selected
after boot.

15 years agoBack out last commit, this wasn't supposed to crep in.
Joerg Sonnenberger [Wed, 1 Jun 2005 20:47:14 +0000 (20:47 +0000)]
Back out last commit, this wasn't supposed to crep in.

15 years agoUse the ACPI timer as the system clock if possible. This should free up
Matthew Dillon [Wed, 1 Jun 2005 20:40:33 +0000 (20:40 +0000)]
Use the ACPI timer as the system clock if possible.  This should free up
the PC speaker on most systems and fix VESA BIOS and other issues.

15 years agoDisable the ability to change the system clock with a sysctl. More
Matthew Dillon [Wed, 1 Jun 2005 20:38:37 +0000 (20:38 +0000)]
Disable the ability to change the system clock with a sysctl.  More
work is needed before this will work (if we ever want to make it work,
that is).  The system clock selected at boot time is the one that will
be used.

15 years agoif_printf / device_printf. remove stored device_t reference as it is
Joerg Sonnenberger [Wed, 1 Jun 2005 20:21:47 +0000 (20:21 +0000)]
if_printf / device_printf. remove stored device_t reference as it is
no longer needed.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoMore cleanups, add the API implementation to select the system clock.
Matthew Dillon [Wed, 1 Jun 2005 20:19:47 +0000 (20:19 +0000)]
More cleanups, add the API implementation to select the system clock.

15 years agoRemoving rest of debugging code which slipped into the commit.
Joerg Sonnenberger [Wed, 1 Jun 2005 20:04:53 +0000 (20:04 +0000)]
Removing rest of debugging code which slipped into the commit.
This doesn't change the semantic.

Noticed-by: Andreas Kohn
15 years ago* Add __unused where necessary.
Sascha Wildner [Wed, 1 Jun 2005 18:37:20 +0000 (18:37 +0000)]
* Add __unused where necessary.

* Change local variable name to avoid name shadowing.

* Bump WARNS to 6.

Submitted by: Sepherosa Ziehau <sepherosa@softhome.net>

15 years ago- Kill keyword 'register'
Liam J. Foy [Wed, 1 Jun 2005 18:25:45 +0000 (18:25 +0000)]
- Kill keyword 'register'

15 years agoStop branding DragonFly binaries with the FreeBSD ABI.
Joerg Sonnenberger [Wed, 1 Jun 2005 18:07:10 +0000 (18:07 +0000)]
Stop branding DragonFly binaries with the FreeBSD ABI.
Increment __DragonFly_version to get this into programs,
we can do sanity checks on it later.

15 years agoImplement a new cputimer infrastructure to allow us to support different
Matthew Dillon [Wed, 1 Jun 2005 17:43:46 +0000 (17:43 +0000)]
Implement a new cputimer infrastructure to allow us to support different
time bases.  This is in preparation for adding support for the ACPI timer
as the system's free-running counter instead of the i8254 timer 1 or timer 2.

15 years agoAdd back support for SYSV binaries. Those are nasty since
Joerg Sonnenberger [Wed, 1 Jun 2005 16:51:46 +0000 (16:51 +0000)]
Add back support for SYSV binaries.  Those are nasty since
ELFOSABI_NONE == ELFOSABI_SYSV, but doing another loop to
handle them doesn't hurt much, since it's not the critical path
anyway.

15 years agoTeach file about DragonFly's ELF ABI tagging.
Joerg Sonnenberger [Wed, 1 Jun 2005 15:09:11 +0000 (15:09 +0000)]
Teach file about DragonFly's ELF ABI tagging.

Go ahead for local patch by maintainer.

15 years agoIncrement subversion to allow matching for ELF ABI tagging enabled kernels.
Joerg Sonnenberger [Wed, 1 Jun 2005 14:36:48 +0000 (14:36 +0000)]
Increment subversion to allow matching for ELF ABI tagging enabled kernels.

15 years agoMerge from vendor branch FILE:
Joerg Sonnenberger [Wed, 1 Jun 2005 14:34:46 +0000 (14:34 +0000)]
Merge from vendor branch FILE:
Update to file 4.13. Put the contrib files into contrib/file-4 instead
of contrib/file-4.13 since we want to do more incremental updates and
this keeps the repository size down.

15 years agoUpdate to file 4.13. Put the contrib files into contrib/file-4 instead
Joerg Sonnenberger [Wed, 1 Jun 2005 14:34:46 +0000 (14:34 +0000)]
Update to file 4.13. Put the contrib files into contrib/file-4 instead
of contrib/file-4.13 since we want to do more incremental updates and
this keeps the repository size down.

15 years agoUpdate to file 4.13. Put the contrib files into contrib/file-4 instead
Joerg Sonnenberger [Wed, 1 Jun 2005 14:34:46 +0000 (14:34 +0000)]
Update to file 4.13. Put the contrib files into contrib/file-4 instead
of contrib/file-4.13 since we want to do more incremental updates and
this keeps the repository size down.

15 years ago* Remove timevaladd() and move timevalfix() and timevalsub() to
Sascha Wildner [Wed, 1 Jun 2005 03:05:40 +0000 (03:05 +0000)]
* Remove timevaladd() and move timevalfix() and timevalsub() to
  usr.bin/kdump where they belong.

* Include missing <stdlib.h>, add prototypes, staticize, constify,
  remove void casts, add __unused.

* Minor style(9) cleanup.

* Bump WARNS to 6.

Submitted by:     Haakon Schad Bergsaker <hakonsb@math.uio.no>
Modifications by: me

15 years agoAdd DragonFly cvs id.
Sascha Wildner [Wed, 1 Jun 2005 03:02:30 +0000 (03:02 +0000)]
Add DragonFly cvs id.

15 years agoRemove no longer relevant file.
Sascha Wildner [Tue, 31 May 2005 22:49:19 +0000 (22:49 +0000)]
Remove no longer relevant file.

15 years ago- Minor restructure. Don't bother calling makemsg() unless we can create
Liam J. Foy [Tue, 31 May 2005 19:11:04 +0000 (19:11 +0000)]
- Minor restructure. Don't bother calling makemsg() unless we can create
the client.

Ok: joerg

15 years agoDon't match executables with ELFOSABI_NONE against the brand list.
Joerg Sonnenberger [Tue, 31 May 2005 17:45:20 +0000 (17:45 +0000)]
Don't match executables with ELFOSABI_NONE against the brand list.
For binaries without matching brand, try to match PT_NOTE header
of type 1 (which is equal to .note.ABI-tag as section).

Implement matching for FreeBSD, the up-coming DragonFly and Linux
ABI tag.

This implementation has some known limitations:
- only one PT_NOTE header of type 1 is checked (the last)
- the PT_NOTE header has to be part of the first page
- ELF branding is prefered, no further matching is done (prevents multiple
  ABI emulations for the same ELF OS ABI)

Inspired-by: NetBSD
15 years agoRemove M_EXT_OLD, rename m_ext.ext_nref.new to m_ext.ext_ref and
Joerg Sonnenberger [Tue, 31 May 2005 14:11:43 +0000 (14:11 +0000)]
Remove M_EXT_OLD, rename m_ext.ext_nref.new to m_ext.ext_ref and
m_ext.ext_nfree.new to m_ext.ext_free.

15 years ago- Fix memory leak if realloc(3) failed. Use reallocf(3).
Liam J. Foy [Tue, 31 May 2005 13:50:09 +0000 (13:50 +0000)]
- Fix memory leak if realloc(3) failed. Use reallocf(3).

Ok: joerg

15 years agoIf we want to abuse an API by providing a callback doing nothing, we
Joerg Sonnenberger [Tue, 31 May 2005 12:31:21 +0000 (12:31 +0000)]
If we want to abuse an API by providing a callback doing nothing, we
can at least follow the new API to do nothing.

15 years agoConvert to new m_ext API.
Joerg Sonnenberger [Tue, 31 May 2005 12:29:05 +0000 (12:29 +0000)]
Convert to new m_ext API.

15 years agocleanup
Joerg Sonnenberger [Tue, 31 May 2005 10:44:13 +0000 (10:44 +0000)]
cleanup

15 years agoNo need to bzero softc.
Joerg Sonnenberger [Tue, 31 May 2005 08:30:14 +0000 (08:30 +0000)]
No need to bzero softc.

15 years agoLet fxp_release take a device_t directly and change some device_printf
Joerg Sonnenberger [Tue, 31 May 2005 08:27:47 +0000 (08:27 +0000)]
Let fxp_release take a device_t directly and change some device_printf
to if_printf. This allows the elimination of fxp_softc.dev.
Also add a missing '\n' in fxp_attach.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoRemove __inline hints, let the compiler figure out the details.
Joerg Sonnenberger [Tue, 31 May 2005 08:23:29 +0000 (08:23 +0000)]
Remove __inline hints, let the compiler figure out the details.
This prevents GCC 3.4.4 from inline fxp_scb_wait, but since we access
external memory anyway, the calling overhead doesn't matter.

15 years agoUse PCI accessor functions instead of messing directly with the
Joerg Sonnenberger [Tue, 31 May 2005 08:19:04 +0000 (08:19 +0000)]
Use PCI accessor functions instead of messing directly with the
config space.
Use if_printf in fxp_watchdog.
Don't hardware interface name for if_initname, use the device name.
Remove bogus source path and include of machine/clock.h.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoRemove bogus check if interface was already attached. The function is
Joerg Sonnenberger [Tue, 31 May 2005 08:12:48 +0000 (08:12 +0000)]
Remove bogus check if interface was already attached. The function is
called only from *_attach and if those are called twice, enough is broken
already.

15 years agoif_printf / device_printf.
Joerg Sonnenberger [Tue, 31 May 2005 08:10:44 +0000 (08:10 +0000)]
if_printf / device_printf.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoif_printf / device_printf cleanup.
Joerg Sonnenberger [Tue, 31 May 2005 08:06:47 +0000 (08:06 +0000)]
if_printf / device_printf cleanup.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
While here, use a local eaddr to build the fake MAC address and stop
printing it in fwe_attach, ether_ifattach does it already.