Commit Graph

229478 Commits

Author SHA1 Message Date
Martin Matuska
a2a3407c7e MFV r328323,328324:
Sync libarchive with vendor.

Relevant vendor changes:
  PR #893: delete dead ppmd7 alloc callbacks
  PR #904: Fix archive freeing bug in bsdcat
  PR #961: Fix ZIP format names
  PR #962: Don't modify attributes for existing directories
           when ARCHIVE_EXTRACT_NO_OVERWRITE is set
  PR #964: Fix -Werror=implicit-fallthrough= for GCC 7
  PR #970: zip: Allow backslash as path separator

MFC after:	1 week
2018-01-24 14:24:17 +00:00
Dmitry Marakasov
148ee4c3ff Support configuring arbitrary limits(1) for any daemon in rc.conf
Usage is ${name}_limits, and the argument is any flags accepted by
limits(1), such as `-n 100' (e.g. only allow 100 open files).

Approved by:	cy
Differential Revision:	https://reviews.freebsd.org/D14015
2018-01-24 14:15:06 +00:00
Wojciech Macek
072c8a3b39 Reverting r328320 2018-01-24 13:57:01 +00:00
Hans Petter Selasky
f885420d3c Properly implement the "id" callback argument in the "idr_for_each" function
in the LinuxKPI. The old implementation assumed only one IDR layer was present.
Take additional IDR layers into account when computing the "id" value.

MFC after:	1 week
Found by:	Karthik Palanichamy <karthikp@chelsio.com>
Tested by:	Karthik Palanichamy <karthikp@chelsio.com>
Sponsored by:	Mellanox Technologies
2018-01-24 13:37:07 +00:00
Andrey V. Elsukov
14a6bab1da When IPv6 packet is handled by O_REJECT opcode, convert ICMP code
specified in the arg1 into ICMPv6 destination unreachable code according
to RFC7915.

Obtained from:	Yandex LLC
MFC after:	2 weeks
Sponsored by:	Yandex LLC
2018-01-24 12:40:28 +00:00
Wojciech Macek
68c2d255bc PPC: Add KASSERT in intrcnt_add which checks for buffer overflow
Authored by:           Patryk Duda <pdk@semihalf.com>
Submitted by:          Wojciech Macek <wma@semihalf.com>
Obtained from:         Semihalf
Sponsored by:          IBM, QCM Technologies
2018-01-24 12:01:32 +00:00
Martin Matuska
f7158da37f vendor/libarchive/dist: Add new files missing in r328323 2018-01-24 11:03:18 +00:00
Martin Matuska
a2160dbed3 Update libarchive to git 2d9156bb3ad92c43c1e81546f4a2c21549ba248f
Relevant vendor changes:
  PR #893: delete dead ppmd7 alloc callbacks
  PR #904: Fix archive freeing bug in bsdcat
  PR #961: Fix ZIP format names
  PR #962: Don't modify attributes for existing directories
           when ARCHIVE_EXTRACT_NO_OVERWRITE is set
  PR #964: Fix -Werror=implicit-fallthrough= for GCC 7
  PR #970: zip: Allow backslash as path separator
2018-01-24 11:01:12 +00:00
Steven Hartland
6fb1399a4c Added missing CTLFLAG_VNET to lacp default_strict_mode
Added CTLFLAG_VNET to net.link.lagg.lacp.default_strict_mode which was missed
in r290450.

Reported by:	julian@
MFC after:	1 week
Sponsored by:	Multiplay
2018-01-24 10:13:14 +00:00
Wojciech Macek
4d249cdd4c ULE: provide defaults to ts_cpu
Fix a bug when the system has no CPU 0. When created, threads were implicitly assigned to CPU 0.
This had no practical effect since a real CPU was chosen immediately by the scheduler. However,
on systems without a CPU 0, sched_ule attempted to access the scheduler queue of the "old" CPU
when assigned the initial choice of the old one. This caused an attempt to use illegal memory
and a crash (or, more usually, a deadlock). Fix this by assigned new threads to the BSP
explicitly and add some asserts to see that this problem does not recur.

Authored by:           Nathan Whitehorn <nwhitehorn@freebsd.org>
Submitted by:          Wojciech Macek <wma@semihalf.com>
Obtained from:         Semihalf
Differential revision: https://reviews.freebsd.org/D13932
2018-01-24 07:54:05 +00:00
Eitan Adler
53362abb8e m4: add annotations to various functions
Explain to the compiler that several functions are either noreturn or
take printf like arguments

Discussed with:	swildner@DragonFlyBSD.org
2018-01-24 07:01:44 +00:00
Philip Paeps
afb91be3c9 Import tzdata 2018c
Changes: https://github.com/eggert/tz/blob/2018c/NEWS

MFC after:	3 days
2018-01-24 06:48:42 +00:00
Philip Paeps
b6a2820ba9 Import tzdata 2018c 2018-01-24 06:45:03 +00:00
Navdeep Parhar
76aca1d671 cxgb(4): Validate offset/len in the GET_EEPROM ioctl.
Reported by:	Ilja Van Sprundel <ivansprundel@ioactive.com>
2018-01-24 05:16:11 +00:00
Navdeep Parhar
09b0b8c058 Do not generate illegal mbuf chains during IP fragment reassembly. Only
the first mbuf of the reassembled datagram should have a pkthdr.

This was discovered with cxgbe(4) + IPSEC + ping with payload more than
interface MTU.  cxgbe can generate !M_WRITEABLE mbufs and this results
in m_unshare being called on the reassembled datagram, and it complains:

panic: m_unshare: m0 0xfffff80020f82600, m 0xfffff8005d054100 has M_PKTHDR

PR:		224922
Reviewed by:	ae@
MFC after:	1 week
Sponsored by:	Chelsio Communications
Differential Revision:	https://reviews.freebsd.org/D14009
2018-01-24 05:09:21 +00:00
Kristof Provost
6701c43213 pf: States have at least two references
pf_unlink_state() releases a reference to the state without checking if
this is the last reference. It can't be, because pf_state_insert()
initialises it to two. KASSERT() that this is always the case.

CID:	1347140
2018-01-24 04:29:16 +00:00
Ian Lepore
8917c34a82 Follow changes in r328307 by using new IIC_RECURSIVE flag.
The driver now ensures only one thread at a time is running in the API
functions (clock_gettime() and clock_settime()) by specifically requesting
ownership of the i2c bus without using IIC_RECURSIVE, then it does all IO
using IIC_RECURSIVE so that each individual IO operation doesn't try to
re-acquire the bus.

The other IO done by the driver happens at attach or intr_config_hooks time,
when there can't be multiple threads running with the same device instance.
So, the IIC_RECURSIVE flag can be safely ORed into the wait flags for all IO
done by the driver, because it's all either done in a single-threaded
environment, or protected within a block bounded by explict
iicbus_acquire_bus() and iicbus_release_bus() calls.
2018-01-24 03:09:56 +00:00
Ian Lepore
548778ed13 Follow changes in r328307 by using new IIC_RECURSIVE flag.
The driver now ensures only one thread at a time is running in the API
functions (clock_gettime() and clock_settime()) by specifically requesting
ownership of the i2c bus without using IIC_RECURSIVE, then it does all IO
using IIC_RECURSIVE so that each individual IO operation doesn't try to
re-acquire the bus.

The other IO done by the driver happens at attach or intr_config_hooks time,
when there can't be multiple threads running with the same device instance.
So, the IIC_RECURSIVE flag can be safely ORed into the wait flags for all IO
done by the driver, because it's all either done in a single-threaded
environment, or protected within a block bounded by explict
iicbus_acquire_bus() and iicbus_release_bus() calls.
2018-01-24 03:09:41 +00:00
Ian Lepore
ac8f374abf Fix a bug introduced with recursive bus ownership support in r321584.
The recursive ownership support added in r321584 was unconditionally in
effect all the time -- whenever a given i2c slave device instance tried to
lock the i2c bus for exclusive use when it already owned the bus, the call
returned immediately without waiting.  However, many i2c slave drivers use
bus ownership to enforce that only a single thread at a time can be using
the slave device.  The recursive locking changes broke this use case.

Now there is a new flag, IIC_RECURSIVE, which can be mixed in with the
other flags passed to iicbus_acquire_bus() to allow drivers to indicate
when recursive locking is desired.  Using the flag implies that the driver
is managing concurrent access to the device by different threads in some way.

This immediately fixes all existing i2c slave drivers except for the two
i2c RTC drivers which use the recursive locking feature; those will be
fixed in a followup commit.
2018-01-23 23:30:19 +00:00
John Baldwin
faf5dd873e Remove some KSE references from ps(1).
- Simplify the description of -H to assume 1:1 threading.
- Drop 'process' from description of 'lwp' field and the corresponding
  XO field name.
- Do add an expansion of LWP in the description of 'lwp' and 'nlwps'.
- Add 'tid' as an alias for the 'lwp' field.

Reviewed by:	imp, kib (older version)
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D14021
2018-01-23 22:48:06 +00:00
Ed Maste
4d2a062671 libcxxrt: Move mangled symbols out of extern "C++" in Version.map
r260553 added a number of mangled C++ symbols to Version.map inside of
an existing `extern "C++"` block.

ld.bfd 2.17.50 treats `extern "C++"` permissively and will match both
mangled and demangled symbols against the strings in the version map
block.  ld.lld interprets `extern "C++"` strictly, and matches only
demangled symbols.

I believe lld's behaviour is correct.  Contemporary versions of ld.bfd
also behave as lld does, so move the mangled symbols out of the
`extern "C++"` block.

PR:		225128, 185663
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
2018-01-23 22:41:13 +00:00
Kirk McKusick
8557409f20 In the C library, the setting up of the group array by various
utilities is done by calling gr_addgid() for each group to be
added (usually found by traversing /etc/group) then calling the
setgroups() system call after the group set has been created.
The gr_addgid() function (helpfully?) deduplicates the addition
of group members. So, if you call it to add a group member that
already exists, it is just dropped. Because group[0] is the
effective group-ID and is over-written when a setgid program
is run, The value in group[0] is usually duplicated so that
group value is not lost when a setgid program is run.

Historically this happened because the group value indicated
in the password file also appears in /etc/group (e.g., if you
are group staff in the password file, you will also appear in
the staff line in /etc/group). But, with the addition of the
deduplication, the attempt to add group staff was lost because
it already appeared in group[0]. So, the fix is to deduplicate
starting from group[1] which allows a duplicate of the entry in
group[0], but not in later entries.

There is some confusion about the setgroups system call because in
BSD it has (always) set the entire group including the egid group
(in group[0]). However, in Linux, it skips over group[0] and starts
setting from group[1]. See this comment from linux_setgroups:

      /*
       * cr_groups[0] holds egid. Setting the whole set from
       * the supplied set will cause egid to be changed too.
       * Keep cr_groups[0] unchanged to prevent that.
       */

To make it clear what the BSD setgroups system call does, I
added the following paragraph to the setgroups(2) manual page:

   The first entry of the group array (gidset[0]) is used as the effective
   group-ID for the process.  This entry is over-written when a setgid
   program is run.  To avoid losing access to the privileges of the
   gidset[0] entry, it should be duplicated later in the group array.
   By convention, this happens because the group value indicated in the
   password file also appears in /etc/group.  The group value in the
   password file is placed in gidset[0] and that value then gets added a
   second time when the /etc/group file is scanned to create the group set.

Reported by: Paul McMath  paulm at tetrardus.net
Reviewed by: kib
MFC after:   2 weeks
2018-01-23 22:18:45 +00:00
Ian Lepore
18d61b59bf Switch to using the bcd_clocktime conversion functions that validate the BCD
data without panicking, and have common code for handling AM/PM mode.
2018-01-23 21:36:26 +00:00
Ian Lepore
e324bd4bd6 Switch to using the bcd_clocktime conversion functions that validate the BCD
data without panicking, and have common code for handling AM/PM mode.
2018-01-23 21:31:43 +00:00
Ian Lepore
234add65f7 Switch to using the bcd_clocktime conversion functinos that validate the BCD
data without panicking, and have common code for handling AM/PM mode.
2018-01-23 21:18:15 +00:00
Ed Maste
fdab3fd7e2 copyright.h: Update license text to 'THE AUTHOR'
This matches the license text at
https://www.freebsd.org/copyright/freebsd-license.html

Sponsored by:	The FreeBSD Foundation
2018-01-23 20:38:03 +00:00
Ed Maste
7eb2159f6a Use BSD-2-Clause-FreeBSD license on linux_support.s
These files previously had a 3-clause license and 'THE REGENTS' text.
Switch to standard 2-clause text with kib's approval, and add the SPDX
tag.

Approved by:	kib
2018-01-23 20:35:43 +00:00
Brooks Davis
8b75008bea Don't escape '?'s in protocol output.
This isn't required by mandoc and is nonfunctional in groff.

PR:		224632
Reported by:	w.schwarzenfeld@utanet.at
MFC after:	3 days
Differential Revision:	https://reviews.freebsd.org/D13779
2018-01-23 20:07:51 +00:00
Alan Somers
81e04458b8 sys/netinet6: fix typos in comments. No functional change.
MFC after:	3 weeks
Sponsored by:	Spectra Logic Corp
2018-01-23 19:40:05 +00:00
Kyle Evans
13811ec82e stand/fdt: Fallback to name + ".dtbo" if we fail to load name
This behavior also matches a Linux-ism by allowing fdt_overlays to specify
names of overlays without an extension, e.g. fdt-overlays="sunxi-h3-h5-emac"

If we fail to load the file given by a name in fdt_overlays, try again with
".dtbo" appended to it. This still allows overlays to lack .dtbo extension
if user prefers it and just adds a fallback cushion.

Future work could move this from a hard-coded ".dtbo" to a loader.conf(5)
configuration option.

Reviewed by:	gonzo
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D13968
2018-01-23 18:03:13 +00:00
Warner Losh
f9a20b8fc1 There's no tainted data here, tag it as such to avoid the slew of
false positives. The files the boot loader reads are assumed to be
good.

CID: 1006663,1006665,1265013, 1265014 (possibly more)
Sponsored by: Netflix
2018-01-23 18:01:40 +00:00
Warner Losh
e77c9f0c3d getenv does not return tainted data in the boot loader. Attempt to
clue Coverity into that fact.

Sponsored by: Netflix
2018-01-23 18:01:36 +00:00
Warner Losh
4baa8d7ebd On malloc failure, be sure to close the include file that triggered
it.

CID: 1007775
Sponsored by: Netflix
2018-01-23 18:01:32 +00:00
Warner Losh
74ecc44117 Don't leak memory when displaying help.
Right now, we'll leak memory when we display a help topic because we
don't free t, s, d that we've just used when breaking out of the loop.
NB: coverity just reported t, but s and d also leak.

CID: 1007776
2018-01-23 18:01:27 +00:00
Warner Losh
1065f77afb Fix some resource leaks.
Always free dev and fstyp before strduping new values to assign to
them. Free them at the end of the loop. This keeps them from leaking
for mal-formed /etc/fstab lines.

CID: 1007777, 1007778, 1007779
Sponsored by: Netflix
2018-01-23 18:01:17 +00:00
Kyle Evans
25f0135c49 Add /boot/overlays to runtime pkg, fix distrib-dirs METALOG generation
/boot/overlays was recently added without belonging to a package. It's only
used by bootloaders at the moment, so add it to the 'runtime' package to get
added with ubldr and friends.

Fix distrib-dirs METALOG generation while we're here. History elsewhere
seems to indicate that bapt@ fixed this to pull in all attributes from
mtrees while generating the METALOG. This fix got clobbered somewhere later,
so restore it.

Reviewed by:	bapt, gjb
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D13996
2018-01-23 17:59:06 +00:00
Ed Maste
da1eeb70d3 lld: Don't mark a shared library as needed because of a lazy symbol.
Obtained from:	LLVM r323221 by Rafael Espíndola
2018-01-23 17:54:39 +00:00
Glen Barber
6c41d54855 Bump Copyright year following r328283.
MFC after:	3 days
MFC with:	r328283
Sponsored by:	The FreeBSD Foundation
2018-01-23 16:48:31 +00:00
Glen Barber
afac3ed6c3 When CHROOTBUILD_SKIP is set, evaluate the existence of /bin/sh
within the CHROOTDIR.  If it does not exist, unset CHROOTBUILD_SKIP
to prevent build failures.

Requested by:	swills
MFC after:	3 days
Sponsored by:	The FreeBSD Foundation
2018-01-23 16:41:31 +00:00
Warner Losh
dfb1d80f64 Fill in ut_id. While it's not relevant to the {OLD,NEW}_TIME entries,
we shouldn't leak stack garbage into the field.

Sponsored by: Netflix
2018-01-23 15:34:34 +00:00
Emmanuel Vadot
62de70375f sockstat: add break that was forgot in 328279
Reported by:	garga@
MFC after:	1 week
X-MFC With:	328279
Sponsored by:	Gandi.net
2018-01-23 14:33:19 +00:00
Pedro F. Giffuni
7d84ca677f extfs: Remove unused variables.
Found by:	scan-build
Reviewed by:	fsu
Differential Revision:	https://reviews.freebsd.org/D14017
2018-01-23 14:17:04 +00:00
Emmanuel Vadot
ee0afaa9d7 sockstat: Add -q option to suppress the header line
MFC after:	1 week
Sponsored by:	Gandi.net
2018-01-23 13:03:47 +00:00
Wojciech Macek
a81a290f34 PowerNV: send MSI_EOI always after MSI unmask
MSI/MSI-x interrupts are edge-triggered. If an interrupt
arrives when IRQ line is masked, it will be lost and will
never recover. Perform MSI_EOI always after unmask to give
a chance for PHB/XICS to send an interrupt again if MSI/MSI-x
pending bit is set in MSI/MSI-x BAR space.

Submitted by:          Wojciech Macek <wma@semihalf.org>
Obtained from:         Semihalf
Sponsored by:          IBM, QCM Technologies
2018-01-23 08:07:00 +00:00
Xin LI
593ff55396 Document how to load nmdm(4) from a kernel module.
Submitted by:	kevlo
MFC after:	2 weeks
2018-01-23 03:36:49 +00:00
Ryan Stone
bc3d87fd59 Increment the route table gen count after a modify
Increment the route table generation count after modifying a
route.  This signals back to TCP connections that they need to
update their L2 caches as the gateway for their route may have
changed.  This is a heavier hammer than is needed, strictly
speaking, but route changes will be unlikely enough that the
performance effects of invalidating all connection route caches
should be negligible.

MFC after:	1 week
Sponsored by:	Dell EMC Isilon
Differential Revision:	https://reviews.freebsd.org/D13990
Reviewed by:	karels
2018-01-23 03:15:44 +00:00
Ryan Stone
fc21c53f63 Reduce code duplication for inpcb route caching
Add a new macro to clear both the L3 and L2 route caches, to
hopefully prevent future instances where only the L3 cache was
cleared when both should have been.

MFC after:	1 week
Sponsored by:	Dell EMC Isilon
Differential Revision:	https://reviews.freebsd.org/D13989
Reviewed by:	karels
2018-01-23 03:15:39 +00:00
Ryan Stone
dbeab32f94 Invalidate inpcb LLE cache if cached route is invalidated
When the inpcb route cache is invalidated after a change to the
routing tables, we need to invalidate the LLE cache as well.
Previous to this change packets for the connection would continue
to use the old L2 information from the old L3 gateway, and the
packets for the connection would likely be blackholed.

MFC after:	1 week
Sponsored by:	Dell EMC Isilon
Differential Revision:	https://reviews.freebsd.org/D13988
Reviewed by:	karels
2018-01-23 03:15:35 +00:00
Justin Hibbits
1a55e1d038 Fix 64-bit booke kernel builds after the ldscript changes
Commits r326203 and r326978 broke 64-bit booke kernels by introducing a 1MB
zero-pad between the ELF header and the start of the kernel.  This didn't
cause a build failure, but caused kernels to need to be loaded into memory
1MB lower, which could easily break scripts expecting previous behavior.
This change matches the similar change made to AIM in r327358.
2018-01-23 02:52:12 +00:00
Alan Somers
76f9d2759b mlock(2): correct documentation for error conditions.
The man page is years out of date regarding errors. Our implementation _does_
allow unaligned addresses, and it _does_not_ check for negative lengths,
because the length is unsigned. It checks for overflow instead.

Update the tests accordingly.

Reviewed by:	bcr
MFC after:	3 weeks
Differential Revision:	https://reviews.freebsd.org/D13826
2018-01-22 21:45:54 +00:00