Consistent spelling.

(wiz)

Implement workaround for:
arm11 Errata 364296:Possible Cache Data Corruption with Hit-Under-Miss

Remove hack in userret which is redundant with workaround.

workaround code from <imre.deak@teleca.com>

(matt)

Sync with head.

(simonb)

Add a few more conditions to the setting of the symbol size for external
dynamic symbols.

(matt)

Add pci_mapreg_submap():  This function is pci_mapreg_map() with two
additional arguments, offset and maxsize.  This new functionality
eases handling certain tasks within the direct rendering manager, though
I hope others will also find it useful.

pci_mapreg_map() is now merely a wrapper around pci_mapreg_submap();
the latter contains all of the code from the former.

ok christos@

(bjs)

Add an optional radius configuration section to the racoon.conf file. This
is similar to the the LDAP configuration section and overrides settings in
the system radius configuration file.

(mgrooms)

Rewrite the introduction.
NOGCCERROR also prevents -Wa,--fatal-warnings.

(lukem)

If WARNS>0 and !defined(NOGCCERROR), add
  -Wa,--fatal-warnings
to CFLAGS to set assembler warnings to be fatal.
(Done via CFLAGS instead of AFLAGS so that all assembling, not just
that of .S/.s files, uses the option.)

(lukem)

Document NOGCCERROR.
Improve PAXCTL_FLAGS & WARNS.

(lukem)

Add TNF to __COPYRIGHT

(lukem)

Express explicitly that VGA_POST needs X86EMU.

(joerg)

Correct typo to fix the build.

(tteras)

deal specially with sym_zero, returned by a weak symbol that was not found,
by retrieving it's name instead of printing random garbage.

(christos)

- fix p for threads
- more sane ordering for thread command, username, and pid

(christos)

Add support for compiling SA as an option. Implied by COMPAT_40.

i386 kernels both with COMPAT_40 and with no compat options (and thus
no SA) compile.

No functional changes intended.

(wrstuden)

Enable MAKEVERBOSE support.

(lukem)

Switch __COPYRIGHT() to using __SECTIONSTRING() now that userland
copyrights don't have \n in them.
(Tested on i386.  There might be a couple of applications specific
to a given platform that should be fixed.)

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.
Tweak to use a consistent format.

(lukem)

Do not set default gss id if xauth is used.

(tteras)

Remove the \n and tabs from the __COPYRIGHT() strings.
Tweak to use a consistent format.

(lukem)

Sync with head.

(simonb)

Remove the \n and tabs from the __COPYRIGHT() strings.
Tweak to use a consistent format.

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.
Tweak to use a consistent format.

(lukem)

Replace gcc initialisers with c99 initialisers.

(gmcgarry)

UBC license does not require an advertizing clause any more - this files
where probably missed in the initial sweep due to strange formatting
(last clause not explicitly numbered).

FreeBSD converted them (rev. 1.21 pf rarpd.8 and rev 1.41 of rarpd.c
in their repository), nearly four years ago.

(martin)

AFLAGS is actually options to ${CC} (not ${AS}) when compiling .s/.S files.
(For gcc, this means that flags to as(1) need to be passed as -Wa,flag)

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.
Tweak to use a consistent format.

(lukem)

beautify dmesg with MPVERBOSE.
before:

pci0 at hypervisor0 bus 0: configuration mode 1hypervisor0: added to list as bus 0

pchb0 at pci0 dev 0 function 0

now:

pci0 at hypervisor0 bus 0: configuration mode 1
hypervisor0: added to list as bus 0
pchb0 at pci0 dev 0 function 0

(cegger)

Add HAVE_PCC tests for -Wno-pointer-sign.

(gmcgarry)

Have a COW with fscow_run().  Fixes problem of allocating files
on ffs with indirect blocks.

found by simonb

(pooka)

Repeat after me: do not pollute sys with #ifdef _RUMPKERNEL

(pooka)

Implement, and document, an ugly hack which allows cross-building of
NetBSD on Darwin - Mac OS X Leopard - to complete successfully

hack  cross-building hack on Darwin
mdate  20 July 2008
who    agc
file  src/distrib/common/Makefile.mdset      : 1.33
descr
      Darwin has problems with getopt() when used in mdsetimage -v,
      due to the difference between BSD and libiberty() getopt
      implementations, more fully described in
      http://mail-index.netbsd.org/current-users/2008/06/27/msg003242.html
      and the subsequent thread.  For just now, we just have an ugly
      workaround not to call mdsetimage with the -v argument on Darwin
kcah

(agc)

don't include i82365var.h, fixes build error, from Robert Swindells

(drochner)

Correct typo to fix the build.

(tron)

Comment out references to the "-P"/"--perl-regexp" option,
which is not supported on NetBSD.  Fixes PR 39122.

(apb)

Remove trailing whitespace.

(wiz)

Add PMAP_KMPAGE defintion (forgotten commit).  Also add PVF_DIRTY which will
be used by an update pmap.c shortly (to note what pages may have dirty
cache lines).

(matt)

Remove #ifdef __FreeBSD__ around DRM_UPDATE_DRAW ioctl, as we have drawable
code (and if it does not work, we must make it work!).

Add missing entry for DRM_MODESET_IOCTL.

(bjs)

Separate generic vendor id handling to a new function and use it.

(tteras)

Do not set default gss id if xauth is used, otherwise gss-id attribute
might be sent even if it was not requested.

(tteras)

Call mountpoint VFS_FSYNC() from specfs fsync.

Makes rump kern/38057-happy.

(pooka)

Fix previous.

(martin)

Initialize vg_path before using it. This fixes lvcreate for me.

(haad)

Use BUILDSYMLINKS for the nvi.1 manpage instead of copying the vi.1 file
to nvi.1
Resolves the problem that this file was not removed during a make cleandir.

(veego)

Make the user-namespaced rumpuser_cv_has_waiters() of type int
instead of bool to avoid unnecessary problems in trying to provide
the bool type.

(pooka)

put back -p, add sort order for pid, username, command. XXX: should add reverse
sorting order too.

(christos)

Add CLEANFILES for krb5-config

(veego)

Make struct pcib_softc explicit in our softc.

(martin)

Explicitly add struct pcib_softc to the softc since the pcib functions
we call expect this.

(martin)

Rearange softc so it starts with a struct pcib. We reuse the x86 pcib code
which assumes this.

(martin)

Make the softc externally visible, so other bridges reusing this code
don't have to "get it right" manually.

(martin)

Specify volatile to some pointers to access DMA registers.

(tsutsui)

No need to specify '&' to get array's address.
No binary changes.

(tsutsui)

Add CLEANFILES for ${BUILT_SRCS} (vis.h)

(veego)

Fix pointer arithmetic botch in rev 1.37.

(tsutsui)

assert -> KASSERT

(pooka)

You must untypo what you have typoed: rumpnode_if.h -> rumpvnode_if.h

(pooka)

The build options for the library parts are mostly located in this
file.  So, in order to better support UPDATE builds, add a dependency
on this file for object files in subdirectories using this file.

(he)

Add a CLEANFILES line to remove the generated file defs.h

(veego)

Add locking around veriexec operations to prevent all sorts of badness
happening.  This fixes kern/38646.

(blymn)

add forgotten initialization/destruction for tsk_lock.

(bjs)

Add a forgotten mtsleep().

(bjs)

When doing pointer arithmetic to compute limit cast bootp to pointer
type of correct signedness.  Caught by lint.

(uwe)

Remove the \n and tabs from the __COPYRIGHT() strings.
(Tweak some to use a consistent format.)

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.

(lukem)

Make sure callouts are initialized before any access (including stopping
them)

(martin)

Remove the \n and tabs from the __COPYRIGHT() strings.

(lukem)

Remove the \n and tabs from the __COPYRIGHT() strings.

(lukem)

Remove the \n from the __COPYRIGHT example; what(1) stops extracting
at them, so we lose most of the message.

(lukem)

add -1 to usage

(christos)

ioctl compatibility for v4l2 and drm

(jmcneill)

move an XXXNETBSD along with the assignment that moved

(mrg)

merge cornflakes from git-change-04893aa99abfbed8eb6d7067a974fa1f31193c87.

(mrg)

fix a merge error

(mrg)

merge cornflakes from git-change-04893aa99abfbed8eb6d7067a974fa1f31193c87.

(mrg)

import mesa-drm git at change 04893aa99abfbed8eb6d7067a974fa1f31193c87.

Status:

Vendor Tag: mesa-drm
Release Tags: git-change-04893aa99abfbed8eb6d7067a974fa1f31193c87

(mrg)

- From Paul Goyette: correct the description of mutex_spin_exit.
- Remove sentence about lockmgr(9), as it is finally replaced.
- Note atomic_ops(3) and membar_ops(3) man pages.

(rmind)

Fix previous commit. device-mapper driver should build fine now.

(haad)

Add support for BSD raw devices to filesystem code. After LV creation link
to block and raw device should be created now.

(haad)

Pull in agp_i810.h so the agp borrow hack works on i915drm.

(jmcneill)

PCI_PRODUCT works on pa_id, not pa_device

(jmcneill)

dev->pci_device should be assigned PCI_PRODUCT(pa->pa_id), not the id
itself.

(bjs)

use ALIGN() instead of rolling our own.

(bjs)

set dev->kdev = kdev at open time, not in drm_ioctl.

(bjs)

Limit sectorsize to max. 4096 bytes. This prevents a panic when declaring
the node pool since standard pools are limited upto 4096 aparently.
Pagesize limit even?

(reinoud)

Use the allocation descriptor types selected on mount when growing from an
internal to an extent recording.

(reinoud)

Use ecma-167 constants instead of homegrown to prevent double mapping.

(reinoud)

Choose which time of allocation descriptors to use on new nodes based on
the number of partition mappings.

(reinoud)

Implement drm_drawable gunk and drm_locked_task, untested. i915drm links now.

(jmcneill)

buf_destroy() an embedded buffer before returning memory to pool
issue detected by LOCKDEBUG panicing about "allocation contains active lock"

(kardel)

(haad)

Initialize timecounter_lock at IPL_HIGH. Prevents a "locking against myself"
when com.c (options PPS_SYNC) delivers a PPS time stamp during clock
interupt processing.

(kardel)

Change IPL of DRM_SPININIT macro to IPL_NONE.  Currently, the only
mutex initialized by this macro is the dma lock, which certainly
should not be IPL_VM.  The priorities of the other locks should be
revisited.

Also, in DRM_WAIT_ON, enforce the policy of "drm returns -errno
to shared code" (this used to be implemented via a macro).

(bjs)

bare-bones entry for mesa-drm

(mrg)

support new drm.

XXX: not fully enabled yet.  config(1) has a makeoptions bug
with complex conditionals.  see files.drm / files.pcidrm for
some instructions on how to enable it.

(mrg)

XXX: remove the (volatile uint32_t *) cast from atomic_read() macro for now.

now all bsd drm drivers build except for i915 which fails to link.

(mrg)

XXX: always return DRM_MIGHT_BE_AGP for now -- need to implement the
XXX: releant bus controller grovelling for netbsd pci.

(mrg)

XXX rename some macro internal variable names to not clash with users.

(mrg)

rename some inner local variables away from outer scope name conflicts

(mrg)

rename a local variable from free to myfree

(mrg)

pull over many many changes from netbsd-current sys/dev/{,pci/}drm/.
make the mesa-drm style Makefile's actually run on netbsd (but i doubt
they work yet.)

this still has a lot to go.  it doesn't work yet, but it's getting
close.

most of this is from other authors than myself:  yorrick, drochner,
jared, bjs, all contributed a lot more than i have so far, but there's
a bunch of clean up from myself in here as well.

(mrg)

include bsd.subdir.mk

(mrg)

import mesa-drm git from version 5d27fd94afaaf434c3a92af0075420b550055bfb,
which is approximately june 25th or so.

Status:

Vendor Tag: mesa-drm
Release Tags: git-change-5d27fd94afaaf434c3a92af0075420b550055bfb

(mrg)

fix debugging printf

(christos)

s/msgs/msgs(1)/ in a comment.  From PR 39167.

(apb)

man page fixes from PR 39164.

(apb)

Comment out references to the "-P"/"--perl-regexp" option,
which is not supported on NetBSD.  Fixes PR 39122.

(apb)

document -j from Dennis den Brok

(christos)

Sync with head.

(simonb)

Sync with head.

(simonb)

Hopefully fix gcc type puned warning.

(reinoud)

enable DIAGNOSTIC

(pooka)

support cv_has_waiters()

(pooka)

emulate vlog()

(pooka)

add pthread_cond_has_waiters_np()

(pooka)

Don't biohazardwait B_ASYNC buffers.

(pooka)

fix comment

(reinoud)

"mount -u -o log" works now - remove the code that explicity disabled
this.

(simonb)

In ffs_fsync() pass FSYNC_VFS to ffs_full_fsync() for a VBLK vnode so
that the correct "struct mount" is referenced.

Fixes WAPBL for the "mount update" case, so remove the "anti-kern/38057"
hack that was previous there to guard against this.

Based on suggestion from yamt@.  yamt suggest this could be cleaner
that the current VFS_FSYNC method too.  Another day...

(simonb)

thanks David, but the initialisation might not work, dependent on
the %union used -- remove the "const" instead (and add a "static"
which desn't seem to hurt)

(drochner)

On dirhash lookup, don't retrieve info you already have.

(reinoud)

Remove double check

(reinoud)

First retrieve length and hash values *before* advancing to the next entry.
On hashline collision it would check a random one next.

(reinoud)

Instead of bombing out when trying to mount a disc RW when UDF_READWRITE is
not defined, downgrade it to a read-only mount.

(reinoud)

groff puts this code through g++ which is rather happier if we give
yyvalzero an initialiser.

(dsl)

PR/39155: Wolfgang Solfrank: Invalid parameter to turnstile_exit in rw_downgrade.

(rmind)

Change dm_format_dev to select right device.
Remove debug printf from library.

(haad)

Fix previsious commit, there was typo.

(haad)

Use htonl like ping/traceroute6 do, from Markus Mayer

(christos)

Implement directory hashing to speed up directory traversals. Speed
improvements of at least 4 times in untarring and roughly 100 to 500 times
on file creation in big directories. Lookup of files was O(n*n) and is now
O(1) even for file creation. Free spaces in the directory are kept in a
seperate list for fast file creation.

The postmark benchmark gives:

UDF old:
pm>set transactions 2000
pm>set number 3000
pm>run
Creating files...Done
Performing transactions..........Done
Deleting files...Done
Time:
        1593 seconds total
        681 seconds of transactions (2 per second)

Files:
        3956 created (2 per second)
                Creation alone: 3000 files (4 per second)
                Mixed with transactions: 956 files (1 per second)
        990 read (1 per second)
        1010 appended (1 per second)
        3956 deleted (2 per second)
                Deletion alone: 2912 files (9 per second)
                Mixed with transactions: 1044 files (1 per second)

Data:
        5.26 megabytes read (3.38 kilobytes per second)
        21.93 megabytes written (14.10 kilobytes per second)
pm>

UDF new:
pm>set transactions 2000
pm>set number 3000
pm>run
Creating files...Done
Performing transactions..........Done
Deleting files...Done
Time:
        19 seconds total
        3 seconds of transactions (666 per second)

Files:
        3956 created (208 per second)
                Creation alone: 3000 files (230 per second)
                Mixed with transactions: 956 files (318 per second)
        990 read (330 per second)
        1010 appended (336 per second)
        3956 deleted (208 per second)
                Deletion alone: 2912 files (970 per second)
                Mixed with transactions: 1044 files (348 per second)

Data:
        5.26 megabytes read (283.66 kilobytes per second)
        21.93 megabytes written (1.15 megabytes per second)

(reinoud)

fix a longstanding bug, from Otto Moerbeck:
if we're reducing a rule that has an empty
right hand side and the yacc stackpointer is pointing at the very
end of the allocated stack, we end up accessing the stack out of
bounds by the implicit $$ = $1 action

do it a bit differently than in Openbsd: avoid pulling in <string.h>
because this is not compatible with wacky namespace games we do in libc

(drochner)

include <stdlib.h> explicitely, for exit(), getprogname() etc.
(it happens to be pulled in by the yacc skeleton, but we shouldn't
rely on this)

(drochner)

Mark groff_mom.0, groff_mom.7, mom.tmac, and om.tmac as obsolete.
Fixes build after 'mom removal'.

(oster)

Revert part of previous changes:
> make makeiplcookie() return PSL bits via ipl2psl_table[]
> rather than IPL_FOO indexes
because ipl2psl_table[] could be updated during autoconf(9)
and we can't use it as cookie before cpu_configure(9).

(tsutsui)

Dont round up nanosecond count; if its nearing wrap one could get a count
of >1 second resulting in invalid time stamps on disc.

(reinoud)

Disable warning about nonexisting proc dir.

(haad)

make this compile

(cegger)

devive_private -> device_private

(cegger)

use device_xname

(cegger)

remove redundant declaration of tvtohz(). It's declared in <sys/timevar.h>.
Makes i386 ALL kernel build again.

(cegger)