- use warn/err properly
- wrap long lines

(christos)

Minor wording correction.

(dsieger)

Use 'a && b && ...' contructs for running multiple commands on a single line
in Makefiles, to properly detect/report failures.

(njoly)

Remove dead code (it is commented out).

Delete the detach routine, it's obviously not finished.

(dyoung)

Added myself

(pettai)

Add the _COMPONENT definition and missing prototypes. Some related cosmetics.

(jruoho)

Preparing a device for wakeup involves:

  (a) turning on all power resources required by the device; and
  (b) executing _DSW (or _PSW) control method.

This implements (b). Ok jmcneill@.

(jruoho)

Make module_{lookup,enqueue}() static now that it's possible again
(effectively reverts my kern_module rev. 1.53 from some months ago)

(pooka)

Changes to 1.99.21/20100305

+ clean up some lint
+ the obligatory ''build on mac os x fixes'' - <inttypes.h> header file

(agc)

Remove no longer used -DNFS.

(pooka)

Set env vars before returning from fsu_mount

(stacktic)

note module changes

(pooka)

Update to use newstyle rump_module interfaces.

(pooka)

During bootstrap, process all modinfos in a DSO in one go.  Get
rid of dependency tricks, since they are no longer necessary.

(pooka)

regen for rump_module interface change

(pooka)

Use improved kernel module interfaces: instead of adding + loading
modules in bootstrap, just add them.  Load them later the same way
as the kernel does: module_init_class().

Change the signature of rump_module_init() to take a vector instead
of just one module.  All modules in a DSO should be init'd at the
same time because they might depend on each other, and code outside
the rump kernel cannot know which way.  (binary kernel modules are
still loaded with rump_sys_modctl() the usual way).

(pooka)

Fix console attach.

(skrll)

Move builtin modules to a list in init and load them from there
instead of using linksets directly.  This has two implications:

1) It is now possible to "unload" a builtin module provided it is
  not busy.  This is useful e.g. to disable a kernel feature as
  an immediate workaround to a security problem.  To re-initialize
  the module, modload -f <name> is required.
2) It is possible to use builtin modules which were linked at
  runtime with an external linker (dlopen + rump).

(pooka)

Link the kernel with -q option to be able to relocate it with the boot loader.

(phx)

A1200 (BPPC) BAT init.

(phx)

Removed useless code.

(phx)

Do not match on the elroy family. From OpenBSD.

(skrll)

Add missing license.

ok drochner@

(jruoho)

Reset sendsig_sigcontext_vec at fini to avoid having it point at
garbage and to allow re-init of the module (if DIAGNOSTIC).

(pooka)

clean up some lint that crept in.

(agc)

Update netpgp to version 1.99.20/20100304 - portability improvements, and bug fixes:

Changes to 1.99.20/20100304

+ move args to some functions around to be consistent
+ use uint*_t where appropriate
+ fix bug in verify memory
+ add documentation to manual pages to show how to do combined
  signing/encryption and decryption/verification
+ make verification of ascii-armoured memory work the same as binary
+ eliminate use of strdup(3), strcasecmp(3), and strptime(3). NetBSD/pkgsrc
  PR 42922 applies - need to define _XOPEN_SOURCE and _BSD_SOURCE for
  newer linux platforms with glibc 2.10.1. solved a bit differently, by
  implementing strdup(3) and strcasecmp(3) independently, and using regexps
  to avoid calling strptime(3).

(agc)

Remove <dev/acpi/acpica.h> from all files. It is included from
<dev/acpi/acpivar.h>. Ditto for <dev/sysmon/sysmonvar.h>, <sys/bus.h>,
<dev/pci/pcivar.h>, and <dev/isa/isavar.h>.

Also nuke a lot of unused and invalid headers. Some of these are audibly
provided by standard headers (namely <sys/param.h> and <sys/device.h>), some
have nothing to do with ACPI devices (e.g. <sys/syslog.h>), and some are
nonexistent local includes (e.g. "mpu_ym.h"). Moreoever, try to group the
includes into their respective blocks.

Tested with GENERIC and ALL (i386). No functional change.

(jruoho)

As pointed out by Rhialto, "degrees Kelvin" is redundant.  So fix a
couple of comments related to converstion between Kelvin and degC.

(pgoyette)

Use 8k buffer, since 4k is too short for the default list (might
save one round of modctl()).

(pooka)

regen (for lfs syscalls #ifdef removal).

(pooka)

Protect all header files against multiple inclusions.

(jruoho)

ftp 20100305: fix http date parsing

(lukem)

Back to using 'RFC xxxx' instead of 'RFCxxxx'

(lukem)

From Stefan Bauer: Fix multiple typoes and manpage formatting errors.

(tteras)

From Stefan Bauer: Fix multiple typoes and manpage formatting errors.

(tteras)

DTrace: add a dependency on the dtrace module.

(darran)

DTrace: fix the curthread and mutex data types.

(darran)

DTrace: Explicitly check for OS X to decide if echo needs -e.

(darran)

Simplify this a bit (and reduce differences with if_rtw_pci.c): we
don't have to save the base address or write it to the BAR.

(dyoung)

This is *always* compiled with #define rbus 1, so get rid of the
conditional compilation.

Make a few changes of this kind:

- ih = cardbus_intr_establish(cc, cf, ...);
+ ih = Cardbus_intr_establish(ct, ...);

- cardbus_intr_disestablish(cc, cf, ih);
+ Cardbus_intr_disestablish(ct, ih);

Tested by plugging a Syba CardBus to FireWire adapter into an HP
Pavilion N3270, adding an address (169.254.0.7/16) to fwip0, attaching
and pinging a MacBook Pro.

(dyoung)

Remove ACPI_BUG_DEBUG by using ACPI_DEBUG_PRINT(x) instead.

(jruoho)

Add missing semicolon.

(jruoho)

Make cosmetic changes in order to reduce differences with
sys/dev/cardbus/if_rtw_cardbus.c: remove an unnecessary #include.
Change a few cut & paste instances of ADM8211 to RTL8180.  Make the
suspend & resume functions static, add declarations for them at the top
of the file, and move the functions themselves to the bottom.

(dyoung)

Extract both instances of the code that prints the PCI Secondary Status
Register into a subroutine.  Use terminology similar to that which
PCI System Architecture (4th Edition) uses.  For comparison with PCI
reference books, specify flags and fields with __BIT(n) and __BITS(m, n)
instead of hexadecimal constants.

(dyoung)

Fix race condition on reallocation of huge category.

We need to remove the old region before mremap() since if it relesae the
old region, other thread may map it for the same huge category allocation
and insert it to the tree before we acquire a lock after mremap().

Fixes PR/42876.

(enami)

This is *always* compiled with #define rbus 1, so get rid of the
conditional compilation.

(dyoung)

Get rid of the #if 1 surrounding #include <dev/cardbus/rbus.h>.

This is *always* compiled with #define rbus 1, so get rid of the
conditional compilation.

ath, atw, and rtw work fine after this change.

(dyoung)

This is *always* compiled with #define rbus 1, so get rid of the
conditional compilation.

(dyoung)

Remove unnecessary #ifdef INET-enclosed #includes.

This is *always* compiled with #define rbus 1, so get rid of the
conditional compilation.

(dyoung)

Remove ACPI_BUT_DEBUG.

(jruoho)

Cleanup:

  * Semantics.
  * No need for <dev/acpi/acpica.h>.
  * Reduce the amount of error reporting.
  * Remove ACPI_BUT_DEBUG and ACPIBUT_F_VERBOSE.

No functional change.

(jruoho)

VALD_ACPI_DEBUG is dead.

(jruoho)

Remove VALD_ACPI_DEBUG and AVALD_F_VERBOSE. No functional change.

(jruoho)

Enable (uncomment) ACPI_DEBUG, ACPI_DEBUG_ALLOC, and ACPI_MUTEX_DEBUG.

(jruoho)

Parse HTTP 'Date' entries in the `C' locale rather than the user's.
Fix from [bin/42917] (with minor changes), from KAMADA Ken'ichi.

(lukem)

Fix two memory leaks.

(jruoho)

Fix ACPI_DEBUG build failures reported by Greg A. Woods.

(jruoho)

It turns out that ppb at cardbus was not even in ALL.  It has probably
not been compiled in a very long time.  With some minor changes, it was
possible to make it compile:

Use kmem_zalloc()/kmem_free() instead of alloca()/memset().

Use %zu and %zx for printf'ing bus_size_t.

Delete the declaration of an unused local variable.

(dyoung)

Delete unfinished and unreachable code: ppb_cardbus_setup(),
ppb_cardbus_enable(), ppb_cardbus_disable().

(dyoung)

binutils-2.20.1 out.

(wiz)

Simplify interrupt (dis)establishment by two source transformations:

-      cardbus_intr_disestablish(cc, cf, ih);
+      Cardbus_intr_disestablish(ct, ih);

-      ih = cardbus_intr_establish(cc, cf, ...);
+      ih = Cardbus_intr_establish(ct, ...);

Tested by kiyohara@.

(dyoung)

Note about proper support for NE2000 8 bit mode.

(tsutsui)

Convert padding for unknown capabilities as well.

(roy)

Convert padding and give terminfo some default assumptions about termcap.

(roy)

From Pierre POMES: fixed admin port initialization

(vanhu)

This really should be ACPI_BUS_COMPONENT

Maybe I should stay out the acpi stuff?  :)

(pgoyette)

Fix this correctly.  Thanks jruoho@

(pgoyette)

Instead of "int hz", use "extern int hz".

(jruoho)

Provide the _COMPONENT definition and use ACPI_FREE(x) instead of
AcpiOsFree(). If the memory tracking of ACPICA is enabled, the latter is
guaranteed to leak memory, possibly also corrupting kernel memory.

(jruoho)

Add roundup(9).

(jruoho)

avoid a build error in the previous for !DEBUG kernels.

(mrg)

Document the counting and rounding macros from <sys/param.h>.

(jruoho)

- in _bus_dmamap_unload(), pmap_page_protect() and pmap_clear_reference()
  switch the dcache_flush_page() into a dcache_flush_page_all()
- in both pmap_kremove()/pmap_remove(), remove the blast_dcache() call
  and replace it with dcache_flush_page_all()
- in pmap_get_page() [used to allocate PTP's], always call pmap_zero_page(pa)
- flush the dcache of the dst page in pmap_{copy,zero}_page() by redirecting
  throught a C function that calls the (renamed) asm.  the old asm code had a
  comment about needing to do this...
- add a couple of membar #Sync's that the USIII manual recommends

based on discussions with chuq@, skrll@ and martin@.

these help my SB2000 / SB2500 with both disk / nfs builds and other tasks,
sometimes lasting for several hours before failing or asserting.

(mrg)

Replace ACPI_FREE() with AcpiOsFree() so we no longer need to define
_COMPONENT (we don't have a bit defined for SMBUS anyway).  This was
uncovered by turning on ACPI_DEBUG for the i386 ALL kernel config.

Reported by Greg Woods on current-users@

(pgoyette)

Replace unsafe use of TAILQ_FOREACH: as the comment says, the
structures are pulled off the list in the loop and it's anyone's
guess where they go after that.

(pooka)

Give usage() a chance.
XXX: UKFS_DEVICE_ARGVPROBE() interface

(pooka)

"dtrace,zfs" means "MKDTRACE=yes" *and* "MKZFS=yes" which is not what
we want. Invent a flag "solaris" which is the or of those two flags.

(tron)

The directory "modules/solaris" depends on ZFS as well.

(tron)

The "solaris" kernel module also gets build if ZFS is enabled.

(tron)

Don't attempt to read or write component label stuff from/to 'dead disks'.
Update used spares with the correct parity map bits too.

Addresses PR#42904 by Louis Guillaume.  Fix confirmed by submitter.
Thanks!

(oster)

update description of deffs

(pooka)

reflect previous change in comment

(pooka)

Remove fs_foo.h support from deffs now that nothing uses it.

(pooka)

Put several fixes to ne(4) driver for better 8 bit mode support,
especially on RTL8019AS which is also used for non-ISA local bus of
embedded controllers and some m68k machines like atari and x68k.

* move RTL8019 probe and attach code from each bus attachment
  to MI ne2000_detect() and ne2000_attach()
* change a method for backend and attachment to specify 8 bit mode
  to use a new sc->sc_quirk member, instead of sc->sc_dmawidth
* handle more NE2000 8 bit mode specific settings, including
  bus_space(9) access width and available size of buffer memory
* add a function to detect NE2000 8 bit mode
  (disabled by default, but enalbed by options NE2000_DETECT_8BIT
    to avoid possible regression on various ISA clones)
* fix ipkdb attachment accordingly (untested)

Tested on two NE2000 ISA variants (RTL8019AS and another clone named UL0001)
in both 8 bit and 16 bit mode on i386. "Looks good" from nonaka@.

See my post on tech-kern for details:
http://mail-index.NetBSD.org/tech-kern/2010/02/26/msg007423.html

(tsutsui)

Abort an error.

(roy)

regen: ENOSYS nfssvc honestly

(pooka)

One more overcomplex ENOSYS bites the dust.

(pooka)

Remove nfssvc non-emulations I missed yesterday.

(pooka)

Add prototype and forward declarations to make the empty nfs stub compile.

(martin)

When implementing "read directory", when there are too many empty entries
in a row, and we need to try to read the next block, and have passed a
non-NULL cookie pointer to VOP_READDIR, ensure that we free the cookie
buffer before re-doing VOP_READDIR, so that we don't leak memory.
This fix is similar to nfs_serv.c revisions 1.115 + 1.124.

This should fix the long-standing problem observed by e.g. using Linux-
emulated programs to take backup of servers, which is one of the problems
which were reported in PR#42661.

Thanks to pooka@ for the hints for traversing the VOP* layer.

(he)

Remove ACPI_EXTRA_DEBUG.

(jruoho)

Convert the single instance of ACPI_EXTRA_DEBUG to ACPIVERBOSE.

(jruoho)

Refactor acpi_make_devnode() to make it a little more readable.

No functional change intended.

(jruoho)

s/cpu_fork/cpu_lwp_fork/ in comment.

(skrll)

make suspend/resume actually work:
- we have to DRM_LOCK()/DRM_UNLOCK(), thus s/rad_dev/dev/ in a variable name
- only call into radeon_cp_{stop,resume}() if the device is active

with this my nforce4 dual core amd system is able to suspend/resume with both
X and drm active.

(mrg)

treat the onboard radeon found in Mac Minis like the ones found in iBooks
From Hiroaki Urata via PR 42931

(macallan)

Mention support for Gumstix Verdex Pro and SMSC LAN9118 Family Ethernet
driver, which were added by kiyohara@.

(rmind)

Add a check for duplicate keys. The check is run once and quadratic in
the hash collision chain length, which is expected to be fairly low.

(joerg)

You have found a scroll of genocide --More--
What class of monsters do you wish to genocide? --More--
> fs_foo.h
Wiped out all fs_foo.h

(pooka)

Remove #ifdef KERNFS by using weak symbols.  Notably, this approach
won't work when kernfs is a module.  But then again, kernfs as a
module (i.e. current situation) doesn't contain IPSEC support, so
it's not really any worse either.

(pooka)

Simplify interrupt (dis)establishment by two source transformations:

-      cardbus_intr_disestablish(cc, cf, ih);
+      Cardbus_intr_disestablish(ct, ih);

-      ih = cardbus_intr_establish(cc, cf, ...);
+      ih = Cardbus_intr_establish(ct, ...);

Compiles.  Untested.  Let me know if you use this hardware with NetBSD
any longer.

(dyoung)

Remove fs_lfs now that the syscall is always defined.

(pooka)

remove redundant checks of PK_MARKER.

(yamt)

comment

(yamt)

fix multi-thread problems wrt read/write buffer.  PR/42651.

(yamt)

Fix some historical powerpc curiocity which had kernfs depend on
startsym/endsym, and which crept into most powerpc ports.

(pooka)

Use roundup2() instead of hardcoding the CACHE_LINE_SIZE rounding
operation.

(jym)

-DFFS is no longer necessary

(pooka)

don't create unused fs_nfs.h

(pooka)

Reflect removal of fs_nfs.h and -DNFS

(pooka)

Get rid of dependency on fs_nfs.h, i.e. source modules with
conditional content depending on if the NFS client is wanted or
not.  The server can now be made an independent module not depending
on the nfs client.

Tested with rump_nfs (standalone client), rump_nfsd (standalone
nfsd) and a qemu installation with both the client and the server.

(pooka)

DTrace: another fix for OS X builds, stdint.h location.

(darran)

Make machine/spr.h as obsolete.

(matt)

Add missing <powerpc/FOO/spr.h> to a few files missed on the first pass.

(matt)

Add missing <powerpc/FOO/spr.h> to a few files missed on the first pass.
Don't export <machine/spr.h> to userland, only <powerpc/spr.h>

(matt)

For the nfs throttling kludge, test against v_tag == VT_NFS instead
of v_op (the latter imposes linkage).

(pooka)

Remove unnecessary #ifdef NFS.

(pooka)

regen:
- remove nfssvc / fs_nfs.h / #ifdef NFS
- make getfh unconditional

(pooka)

remove nfssvc emulation (or non-emulation)

(pooka)

DTrace: Get DTrace to build on OS X platforms.

(darran)

Kill fs_nfs.h and #ifdef NFS by making getfh() support unconditional
and nfssvc unsupport unconditional.  nfs server was just a fancy
NOP anyway for all except Ultrix.  I know I'm boring but I couldn't
imagine why someone would want to run an Ultrix nfs server (and I
also doubt that it would work anymore with all the changes to fs
exporting etc).

(pooka)

Document new hp700 'frame' command.

(skrll)

DTrace: Add CTF generation to the make framework.  Only enabled when
MKDTRACE=yes.

(darran)

DTrace: add libelf, libdwarf, libctf, and CTF tools to the toolchain,
built when MKDTRACE=yes.

(darran)

Add a "mach frame" command to hppa ddb.

(skrll)

Simplify interrupt (dis)establishment by two source transformations:

-      cardbus_intr_disestablish(cc, cf, ih);
+      Cardbus_intr_disestablish(ct, ih);

-      ih = cardbus_intr_establish(cc, cf, ...);
+      ih = Cardbus_intr_establish(ct, ...);

Tested with ath & atw units on a Soekris net4521.

(dyoung)

DTrace: Add the sdt.h header to the set lists.

(darran)

fs_lfs.h is no longer necessary

(pooka)

regen: lfs megamaid syscalls -> MODULAR

(pooka)

load lfs syscalls in modload

(pooka)

Make lfs syscalls loadable.  This nukes fs_lfs.h & #ifdef LFS.
(I don't mind if someone wants to go further and OBSOL them).

(pooka)

Make is possible to add extra output at the top of syscallargs.h.
Use this feature to stick sys/mount.h in there.

(pooka)

/*
* XXX: Get extra reference to LFS vfsops.  This prevents unload,
* but also prevents kernel panic due to text being unloaded
* from below lfs_writerd.  When lfs_writerd can exit, remove
* this!!!
*/

(pooka)

Format string cosmetics (mainly from %d to %u).

(jruoho)

Remove the ufs_vnops.c hacks from the lfs module, since ffs (and
hence ufs) has been required by lfs for ages now.

(pooka)

Remove fs_mfs.h from users because it is now unnecessary and don't
generate fs_mfs.h anymore.

(pooka)

Make mfs_initminiroot() mandatory.  Allows to remove #ifdef MFS.

(pooka)

Make mfs_initminiroot() mandatory.  Allows to remove #ifdef MFS.

(pooka)

Re-zero the attach args before setting them up for a new call to
config_found_ia().  In this way we avoid passing an uninitialized
pcibus_attach_args.pba_pc that will cause problems, later.

(dyoung)

Don't generate unused fs_thefs.h headers.

(pooka)

Log successful and unsuccessful attempts to change passwords, via -l
or pam, to ease IT audit guideline compliance.  Patch from Richard
Hansen of BBN in private mail.

Proposed on tech-kern with positive comments, except a suggestion I
didn't implement:

A possible future enhancement is refraining from logging if the old
password is empty, as some people abort password changing that way.
However, it's not clear if this complies with most guidelines that
require password change logging, and at first glance that appears to
be a fairly difficult change.

(gdt)

regen for conditional lfs syscall removal

(pooka)

-fs_lfs.h

(pooka)

remove unused lfs compat stubs

(pooka)

-fs_lfs.h

(pooka)