Pull up following revision(s) (requested by bouyer in ticket #1332):
sbin/dump/main.c: revision 1.73
sbin/dump/tape.c: revision 1.54
Consistently use iswap64() and 64bit intermediate variable, as well as
64bit formats for c_tapea. Fixes tape useage report for large filesystems.

(snj)

Pull up following revision(s) (requested by bouyer in ticket #1331):
sbin/dump/dump.8: revision 1.67 via patch
sbin/dump/main.c: revision 1.72
sbin/dump/rcache.c: revision 1.25
Default the read block size (-k default value) to kern.maxphys (usually
64k these days). This gives a noticable performance boost on large filesystems.

(snj)

113[45]

(snj)

Pull up following revision(s) (requested by skrll in ticket #1135):
sys/dev/usb/if_smsc.c: revision 1.26
PR/50904: Fix variable assignment inside if-clause.

(snj)

tset was useful for fixed terminals and rarely changing configurations.
But for current software terminal emulators and varying network
logins it's just confusing.

No longer run tset when logging in.

ok @martin

(mlelstv)

Pull up following revision(s) (requested by skrll in ticket #1134):
sys/dev/usb/uhub.c: revision 1.130
Use the port status straight after the reset and not after the
usbd_clear_port_feature(dev, port, UHF_C_PORT_RESET)
The Gadget USB HUB in the IBM x3250 sborrill sent me returns the wrong
status and this would appear to leave the attached cdce(4) in the default
state.  The next device probed fails as a result.

(snj)

libc fallback is known to not produce nice alias entries. Avoid spurious
failures when testing in an empty chroot by requiring the services db.

(joerg)

Require the boot block to be used, test can't work otherwise.

(joerg)

#if 0 a KASSERT for now as it doesn't apply to Raspberry Pi 3

(skrll)

Use the nvidia,has-utmi-pad-registers property.

From Jared.

(skrll)

- GC pts_fired, and fix the comment about MAX_TIMERS
- Bump MAX_TIMERS to 36 so that we have 32 POSIX user timers which is the
  minimum required.

(christos)

Undo the __dso_protected hackery.

(christos)

undo protected symbol hack.

(christos)

Fix copy relocations against protected symbols from:
    https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65248

(christos)

Gcc, please shut up with those stupid warnings!

(martin)

str2Lst_Append tokenizes the string and uses it in the list so we can't
free the string afterwards. Keep a copy of it and cleanup at the end.

(christos)

For :ts numeric escapes \x* is hex, anything else is octal.

(sjg)

PR/50911: David Binderman: Optimize memset

(christos)

PR/50909: David Binderman: Optimize memset

(christos)

PR/50910: David Binderman: Optimize memset.

(christos)

Terminate expression with a semicolon, instead of accidentally
splicing to the next line with a comma. Noticed by Christos. Has no
actual effect, fortunately. (Compiler output is unchanged.)

(dholland)

PR/50908: David Binderman: Optimize memset's

(christos)

Fix ticket #1124

(martin)

Fix ticket #1124

(martin)

Additionally pull up r1.131 for ticket #1124 (requested by skrll)

(martin)

Additionally pull up r1.131 for ticket #1124 (requested by skrll)

(martin)

PR/50907: David Binderman: Remove useless strlen()'s and memcpy()'s.

(christos)

Ignore clang's empty loop-body warning, it is used extensively via
macros.

(joerg)

Relocation of type SHT_REL must use the current value as addend.

Fixes ctfconvert for i386 where section .debug_info uses this relocation.

Ok: Christos Zoulas

(hannken)

PR/50905: Henning Petersen: Fix useless comparison (from FreeBSD)

(christos)

PR/50909: David Binderman: Optimize memset

(christos)

PR/50904: Fix variable assignment inside if-clause.

(skrll)

Fix for ticket 1132. Not EM_IAMCU but STT_GNU_IFUNC.

(msaitoh)

Try to fix build on 32bit architectures.

(martin)

Update ticket 1118. Add ticket 1132.

(msaitoh)

Pullup the following revision to fix build break caused by ticket #1118:

sys/arch/xen/x86/cpu.c 1.102-1.103

Increase the number of entries of cpu_features from 5 to 7.

(msaitoh)

Pullup the following revision (requested by htodd in ticket #1132):

sys/sys/exec_elf.h 1.143

Add definition of EM_IAMCU. This change fixes build break caused by
ticket 1126.

(msaitoh)

Add missing percpu_putref to error path

(ozaki-r)

Avoid void * arithmetic

(christos)

Better to panic! Otherwise we end in an endless loop.

(christos)

Return instead of crashing. This is wrong, and should be fixed properly.

(christos)

Don't throw us in the debugger!

(christos)

Remove advertising clause.

(christos)

use correct offset for channel in SDP record

(plunky)

Don't try to interpret MIPS3-specific status register bit on MIPS1.
Fixes boot of NetBSD/pmax in gxemul 3max emulation mode.

(tnn)

PR/50900: David Binderman: optimize memset

(christos)

PR/50899: David Binderman: optimize memset

(christos)

PR/50898: David Binderman: remove redundant memset.

(christos)

Ticket #1125

(martin)

Pull up following revision(s) (requested by bouyer in ticket #1125):
sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.67
In xennet_xenbus_detach(), remove the event handler early (just after
xennet_stop()) so that we don't get events while slepping (e.g.
in softint_disestablish()) when some structures have already been
freed.
Problem reported and patch tested by Rohan Desai.

(martin)

use kat, a version of cat that strips comments

(christos)

Introduce kat() a built-in version of cat(1) that strips shell-like comments.

(christos)

Tickets #1103, #1117 - #1121, #1123, #1124, #1126, #1129 - #1131

(martin)

Pull up following revision(s) (requested by uwe in ticket #1131):
bin/cp/cp.c: revision 1.59
The '-i' flag should work regardless of whether the standard input is
a terminal.  The Open Group notes this historic behavior and correctly
notes that it doesn't make much sense.  Note also, that mv(1) has
always respected its '-i' regardless of whether the standard input is
a terminal.
From Timo Buhrmester.

(martin)

Pull up following revision(s) (requested by mrg in ticket #1130):
bin/cp/utils.c: revision 1.45
for display in the (post) SIGINFO handler use off_t not size_t for
file sizes.  fixes incorrect reporting errors on 32 bit platforms
with >4GB file sizes.

(martin)

Pull up following revision(s) (requested by mrg in ticket #1129):
bin/mv/mv.c: revision 1.45
add SIGINFO for mv(1), largely based upon the support in cp(1).

(martin)

Catch up to -current, via patch, requested by christos in ticket #1126:

(martin)

Note ticket #1124

(martin)

Pull up following revision(s) (requested by skrll in ticket #1124):
sys/dev/usb/ugen.c: revision 1.127
sys/dev/usb/ugen.c: revision 1.128
sys/dev/usb/ugen.c: revision 1.129
sys/dev/usb/ugen.c: revision 1.130
s/0/NULL/
One more s/0/NULL/
Remove always true conditional
Only clear the endpoint information in ugen_set_interface only if setting
the new altno suceeds.
Avoids the null de-ref in PR/50597 and PR/50810

(martin)

Pull up following revision(s) (requested by skrll in ticket #1124):
sys/dev/usb/ugen.c: revision 1.127
sys/dev/usb/ugen.c: revision 1.128
sys/dev/usb/ugen.c: revision 1.129
sys/dev/usb/ugen.c: revision 1.130
s/0/NULL/
One more s/0/NULL/
Remove always true conditional
Only clear the endpoint information in ugen_set_interface only if setting
the new altno suceeds.
Avoids the null de-ref in PR/50597 and PR/50810

(martin)

Pull up following revision(s) (requested by christos in ticket #1123):
external/bsd/blacklist/lib/bl.c: revision 1.27
lib/Makefile: revision 1.237
external/bsd/blacklist/lib/Makefile: revision 1.4
external/bsd/blacklist/lib/Makefile: revision 1.5
Move blacklistd later now that it depends on libpthread (From Rin Okuyama)
typo
Add a mutex to prevent races during initialization code from multiple threads.
Found in named.

(martin)

Pull up following revision(s) (requested by christos in ticket #1121):
lib/libc/time/localtime.c: revision 1.101
PR/50133: Martin Husemann: Can't cache $TZ.
XXX: Pullup-7

(martin)

Work around gcc on sparc64 bug

(christos)

Pull up following revision(s) (requested by gson in ticket #1120):
lib/libm/src/e_hypotf.c: revision 1.10
Fix incorrect magic numbers in scaling.  From FreeBSD commit 23397, by
way of Rin Okuyama.  Fixes PR lib/50698.

(martin)

Pull up the following revisions, requested by msaitoh in ticket #1119:

sys/arch/x86/x86/procfs_machdep.c 1.7-1.11

x86's /proc/cpuinfo fixes:
- Always output 2 digits for the cpu frequency decimal part.
- Update x86's feature bits in /proc/cpuinfo (PR#49246).
- Fix a bug that /proc/cpuinfo's CPU model was incorrect on many newer
  CPUs and CPU family was incorrect on some AMD CPUs.
- Add comment. Fix comment.

(martin)

Pull up following revision(s) (requested by msaitoh in ticket #1118):
sys/arch/x86/include/cpuvar.h: revision 1.47
sys/arch/x86/x86/cpu.c: revision 1.117
sys/arch/x86/x86/identcpu.c: revision 1.49
sys/arch/x86/include/cpu.h: revision 1.67
Retrieve cpuid7 (Structured Extended Features) into ci_feat_val.

(martin)

Pull up the following changes, requested by msaitoh in #1117:

sys/arch/x86/include/cacheinfo.h 1.20-1.21
sys/arch/x86/include/specialreg.h 1.83-1.86
usr.sbin/cpuctl/arch/i386.c 1.67-1.70

Changes for x86's cpuctl(8):
- Add some TLB information (index 0x6a-0x6d).
- Add Hardware-Controlled Performance States (HWP) bits, FPU Data
  Pointer Updated Only bit and CLFLUSHOPT bit.
- Add some AMD's bit definitions from "BIOS and Kernel Developer(BKDG)
  for AMD Family 15h Models 60h-6Fh Processors".
- Add Xeon E5-4600 v3,
- Add Xeon E3-1200 v4 and v5.
- Add 6th gen Core, Xeon E3-1500 v5 and Xeon D-1500.
- Change CPU family 0x1c from "Atom Family" to "45nm Atom Family"

(martin)

Pull up following revision(s) (requested by knakahara in ticket #1103):
sys/kern/subr_pool.c: revision 1.206
fix: "vmstat -C" CpuLayer showed only the last cpu values.

(martin)

PR/50897: David Binderman: fix memory leaks. While here, modernize error
handling, and types.

(christos)

David Binderman in PR bin/50897: fix memory leak.

(martin)

Non-x86 doesn't have AGP. This should really be handled by config logic.

(mlelstv)

bump date for entry about using NetBSD's regex, now that it works

(aymeric)

Whitespace

(skrll)

add our libc's regex implementation to libnbcompat.

This way, nbsed picks it up and it unbreaks the build of xf86-video-intel-old's
man page under Linux and probably other future changes which rightfully expect
the (net)bsd behaviour.

Full release tested on ArchLinux and NetBSD 7.0_STABLE

(aymeric)

Remove unnecessary local change.

(nakayama)

Reapply following changes to binutils 2.26 to fix PR toolchain/48630.

http://mail-index.netbsd.org/source-changes/2014/03/10/msg052578.html
http://mail-index.netbsd.org/source-changes/2014/03/10/msg052579.html

(nakayama)

Use KASSERT for conditions that cannot be met with current parameters.

(mlelstv)

Simplify the port comparison code further.

(christos)

if opening the audio device fails, try opening again with write-only.
this allows my playback-only usb device to be listed and tested.

(mrg)

kill stray &

(christos)

Fix port matching; we need to ignore ports when they are 0 not only in
the second saidx but the first one too. Fixes NAT-T issue with NetBSD
being the host behind NAT.

(christos)

gather more information from mbuf.

(christos)

Add more debugging, no functional change.

(christos)

The '-i' flag should work regardless of whether the standard input is
a terminal.  The Open Group notes this historic behavior and correctly
notes that it doesn't make much sense.  Note also, that mv(1) has
always respected its '-i' regardless of whether the standard input is
a terminal.

From Timo Buhrmester.

(uwe)

Revert to disabling nouveau by default in x86 GENERIC.

Haven't had time to debug the issues that seem to affect pretty much
everyone.  Maybe soon!

(riastradh)

Whitespace

(skrll)

Whitespace

(skrll)

Correct display of df(1) with the -i parameter

Now the "Mounted on" column should be aligned with /entries.

Patch by Michal Mazurek.

(kamil)

Fix a KASSERT.

I've managed to record via uaudia(4) on Raspberry PI now.

(skrll)

Don't modify command line buffer but create a copy of the root argument.

(mlelstv)

Make agp_i810 also depend on agp so that a 'no agp' build is possible.

(mlelstv)

allow mmap()ing blitter registers

(macallan)

Missing file in pullup #1096: Don't truncate at sub-microsecond while
preserving timestamps.

(martin)

Missing file in pullup #1096: Don't truncate at sub-microsecond while
preserving timestamps.

(martin)

Ticket #1371

(martin)

Hack to adjust config automatically to 32bit and 64bit systems.
Requested by christos in #1371.

(martin)

Ticket #1371

(martin)

Fix path in previous

(martin)

Hack to adjust config automatically to 32bit and 64bit systems.
Requested by christos in #1371.

(martin)

Ticket #1371

(martin)

Hack to adjust config automatically to 32bit and 64bit systems.
Requested by christos in #1371.

(martin)

make it work on linux, be pickier about errors, and correct variable type.

(christos)

Fix infinite recursion caused by gcc lossage.

(christos)

Appease KASSERTs

(skrll)

Improve tests of proxy arp

The tests make it clear how it behaves though, I don't know if the current
behavior is what it should be.

(ozaki-r)

Don't try to use stdin after clobbering it with a failed freopen().
Prevents an extra "Bad file descriptor" message when trying to hexdump
a single nonexistent file.

The intended behavior seems to have been to read from stdin if there
was one filename given and it wasn't valid. But this seems like a bad
idea, so prevent that case instead of hacking it up so it works.

(dholland)

Don't use one function for two different things. Two functions is the
ticket.

(dholland)

use "static"

(dholland)

Add tests for tap(4)

(ozaki-r)

Center game board on screen.  OK from christos@.

(nat)

sun4v: rft_kernel should pay attention to the number of register windows that can be restored. We cannot rely on fill traps to occur, since sun4v only has two trap levels available in privileged mode. Based on code from OpenBSD. ok martin@

(palle)

Get the RPI3 working (in aarch32 mode) by recognising Cortex A53 CPUs.
While I'm here add some A57/A72 info as well.

My RPI3 works with FB console - the uart needs some help with its clocks.

(skrll)

Note Tickets 1096, 1108 and 1128

(martin)

Note Tickets 1096, 1108 and 1128

(martin)

Pull up following revision(s) (requested by uwe in ticket #1128):
sys/kern/kern_time.c: revision 1.184
Don't leak garabage from the kernel stack on sleep(0) and equivalents.
Hat tip to perl's ext/POSIX/t/wrappers.t

(martin)

Pull up following revision(s) (requested by uwe in ticket #1128):
sys/kern/kern_time.c: revision 1.184
Don't leak garabage from the kernel stack on sleep(0) and equivalents.
Hat tip to perl's ext/POSIX/t/wrappers.t

(martin)

Pull up following revision(s) (requested by pgoyette in ticket #1108):
sys/kern/kern_module.c: revision 1.107-1.110
sys/kern/kern_ksyms.c: revision 1.82
Add some additional status messages for kern.module.verbose=TRUE

Remove extraneous ')' from previous.

KNF

In module_do_load(), consolidate checking for a pre-existing module,
and return a single error value EEXIST.  When making a recursive
call (to load required modules), treat a pre-existing module as
success.

Without this change, when a module was loaded by specific request
(as opposed to being loaded as a requirement of some other module),
we would always load the module from the file-system, and then
after making various sanity/compatability checks we would destroy
the new copy if there was a pre-existing copy.

Fixes PR kern/40764

XXX Note that if the module exists, we bypass all of the various
XXX "compatability" checks, including whether or not the existing
XXX module is of any particular class!  (In the previous code, we
XXX checked to see if the newly-loaded copy had the correct class,
XXX but not the pre-existing copy, which could have been loaded
XXX from a different path/filename.)

(martin)

Pull up following revision(s) (requested by pgoyette in ticket #1108):
sys/kern/kern_module.c: revision 1.107-1.110
sys/kern/kern_ksyms.c: revision 1.82
Add some additional status messages for kern.module.verbose=TRUE

Remove extraneous ')' from previous.

KNF

In module_do_load(), consolidate checking for a pre-existing module,
and return a single error value EEXIST.  When making a recursive
call (to load required modules), treat a pre-existing module as
success.

Without this change, when a module was loaded by specific request
(as opposed to being loaded as a requirement of some other module),
we would always load the module from the file-system, and then
after making various sanity/compatability checks we would destroy
the new copy if there was a pre-existing copy.

Fixes PR kern/40764

XXX Note that if the module exists, we bypass all of the various
XXX "compatability" checks, including whether or not the existing
XXX module is of any particular class!  (In the previous code, we
XXX checked to see if the newly-loaded copy had the correct class,
XXX but not the pre-existing copy, which could have been loaded
XXX from a different path/filename.)

(martin)

Pull up following revision(s) (requested by nakayama in ticket #1096):
bin/mv/mv.c: revision 1.44
bin/cp/utils.c: revision 1.43-1.44
lib/librumphijack/hijack.c: revision 1.112-1.115
usr.bin/touch/touch.c: revision 1.33
sbin/restore/tape.c: revision 1.68
sbin/restore/dirs.c: revision 1.51
Don't truncate at sub-microsecond while preserving timestamps.

One of motivation of this change is to make the behavior of test(1)
-nt/ot with preserved copy (like cp -p) closer to the NetBSD 6.
Of course whether full timestamps are kept or not depends also on
underlying file system.

The ifdef added in mv(1) since existing ifdefs was our local change
to compile it on solaris (though I couldn't test it):
http://mail-index.netbsd.org/tech-userlevel/2014/11/28/msg008831.html

Fix the name of failed function in warning message.

Hijack utimensat(2) so that t_vfs test passes after cp(1)/mv(1) are
changed to use the system call.  Linux also has this system call, but
not tested this on linux.

Also hijack futimens(2) so that t_sh test passes.

Define a generic ATCALL() and use it to implement utimensat()
Make ATCALL() behave for absolute paths too.

(martin)

Pull up following revision(s) (requested by nakayama in ticket #1096):
bin/mv/mv.c: revision 1.44
bin/cp/utils.c: revision 1.43-1.44
lib/librumphijack/hijack.c: revision 1.112-1.115
usr.bin/touch/touch.c: revision 1.33
sbin/restore/tape.c: revision 1.68
sbin/restore/dirs.c: revision 1.51
Don't truncate at sub-microsecond while preserving timestamps.

One of motivation of this change is to make the behavior of test(1)
-nt/ot with preserved copy (like cp -p) closer to the NetBSD 6.
Of course whether full timestamps are kept or not depends also on
underlying file system.

The ifdef added in mv(1) since existing ifdefs was our local change
to compile it on solaris (though I couldn't test it):
http://mail-index.netbsd.org/tech-userlevel/2014/11/28/msg008831.html

Fix the name of failed function in warning message.

Hijack utimensat(2) so that t_vfs test passes after cp(1)/mv(1) are
changed to use the system call.  Linux also has this system call, but
not tested this on linux.

Also hijack futimens(2) so that t_sh test passes.

Define a generic ATCALL() and use it to implement utimensat()
Make ATCALL() behave for absolute paths too.

(martin)

Don't leak garabage from the kernel stack on sleep(0) and equivalents.
Hat tip to perl's ext/POSIX/t/wrappers.t

(uwe)

fix indent

(christos)

PR/50883: David Binderman: Remove redundant code.

(christos)

PR/50882: David Binderman: Remove redundant code.

(christos)

PR/50881: David Binderman: Remove redundant code.

(christos)

PR/50880: David Binderman: Remove redundant code.
While here, fix all debugging formats.

(christos)

PR/50879: David Binderman: remove redundant code.

(christos)

David Binderman in PR bin/50884: simplify boolean expression

(martin)

Fix typo.

(msaitoh)

new openssl

(christos)

bump OpenSSL for 1.0.1s (remove SSLv2 and weak "export" ciphers)

(christos)

merge conflicts, bump libraries to 11 for consistency.

(christos)

remove comment I used to test.

(christos)

Build as PIC if requested

(christos)

Fix wrong condition in previous causing bogus 'Use "exit" to leave
shell.' messages (from kre)

(martin)

Don't crash when relocbase was 0.

(joerg)

Don't reuse LDFLAGS for arguments passed directly to ${LD}, share/mk
expects it to be passed to ${CC} or ${CXX}, so they will generally
prefix flags with -Wl.

(joerg)

Remove unused struct member.

(skrll)

Merge r274564 from FreeBSD:
Fix a couple of bugs around the handling of structs and unions of size zero.
These would cause ctfconvert(1) to return an error when attempting to
resolve valid C types.

(joerg)

Re-initialise dwc2_urb->packet_count after the memset(dwc2_urb, 0, ...)

(skrll)

Whitespace

(skrll)

Improved handling of TEST_SH so that it is possible to define it
to the name of a shell, plus options that shell needs to run it
in the correct mode to be tested: eg: TEST_SH='bash -o posix'
Also finished the implementation of tests of "set -n" now that
the NetBSD shell supports that as it should. (from kre)

(christos)

Switch sh3 to the new binutils and gdb.  Lightly tested on landisk.
sh3eb compile-tested only.

(uwe)

Remove regexp.h here temporarily

(martin)

Revert to rev 1.77 for now on Aymeric's request

(martin)

remove regsub.c from the list of files to build because it unbreaks the build
and this API is not yet endorsed. OK christos@

(aymeric)

more tests from kre

(christos)

Even more debugging improvements (from kre)

(christos)

Complete implementation of the noexec option (-n) including
disabling noexec, if the shell is interactive, each time that
a new command is about to be read.  Also correct the -I
(ignoreeof) option so that it only applies to interactive shells,
as required by posix. (from kre)

(christos)

delete clauses 3,4

(christos)

mention that the recent addition of the NetBSD implementation of regexps
to libnbcompat may lead to crashes in tools using them on non-NetBSD hosts
if these are not rebuilt from scratch.

(aymeric)

add our libc's regex implementation to libnbcompat.

This way, nbsed picks it up and it unbreaks the build of xf86-video-intel-old's
man page under Linux and probably other future changes which rightfully expect
the (net)bsd behaviour.

(aymeric)

protect __RCSID the same way other reg*.c files do, so that we can use this
file in libnbcompat

(aymeric)

Tag ctf.5 man page with ctf, not dtrace.

(riastradh)

Skip libctf if MKCTF=no irrespective of MKDTRACE, for real.

Apparently this logic is copied in lib/Makefile and in
external/cddl/osnet/lib/Makefile.  Is the latter even used?

(riastradh)

PR/50867: David Binderman: Fix parsing loop.
While here, modernize error handling, merge copy and pasted code.

(christos)

PR/50868: David Binderman: Fix operator precedence.

(christos)

PR/50869: David Binderman: Use logical and instead of arithmetic

(christos)

PR/50870: David Binderman: Use logical and instead of arithmetic

(christos)

PR/50872: David Binderman: Use logical and instead of arithmetic

(christos)

PR/50871: David Binderman: use logical and instead of arithmetic

(christos)

PR/50873: David Binderman: use logical or instead of arithmetic

(christos)

PR/50874: David Binderman: unsigned can't be negative

(christos)

Let MKCTF alone control whether we build libctf.

If MKDTRACE=yes MKCTF=no makes no sense, we can express that
somewhere else, not as a confusing set lists error at the end -- as
far as I know, there's no way to express MKDTRACE *or* MKCTF as a
condition for set list entries.

(riastradh)

#ifdef forward declaration of setpolicy like definition.

(riastradh)

Consistently use estrlcpy for ifr.ifr_name here, not strncpy.

(riastradh)

uint64_t can't be negative.

(riastradh)

Sprinkle some more zfs range lock reference count kasserts.

From a debugging session months ago, might as well share them.

(riastradh)

Record current raid configurations too in /etc/security.

(riastradh)

Cast int64_t to long long before printfing with %lld.

Not doing so broke something months ago and I forgot why.

(riastradh)

PR/50854: David Binderman: Remove redundant code.

(christos)

Fix buidling with MKPROFILE=no

(adam)

Add tests on activating a new MAC address

(ozaki-r)

Add tests deleting active/inactive links

(ozaki-r)

for display in the (post) SIGINFO handler use off_t not size_t for
file sizes.  fixes incorrect reporting errors on 32 bit platforms
with >4GB file sizes.

(mrg)

Add __printflike as appropriate.

(riastradh)

Mark skewsrcp as __diagused to unbreak non-DIAGNOSTIC builds.

(riastradh)