Commit Graph

246370 Commits

Author SHA1 Message Date
Cy Schubert
964c4a37b0 Chase r344883. ntp is 4.2.8p13.
MFC after:	3 days
2019-11-21 20:15:22 +00:00
Eric Joyner
52e8f6a331 bitstring: add functions to find contiguous set/unset bit sequences
Add bit_ffs_area_at and bit_ffc_area_at functions for searching a bit
string for a sequence of contiguous set or unset bits of at least the
specified size.

The bit_ffc_area function will be used by the Intel ice driver for
implementing resource assignment logic using a bitstring to represent
whether or not a given index has been assigned or is currently free.

The bit_ffs_area, bit_ffc_area_at and bit_ffs_area_at functions are
implemented for completeness.

I'd like to add further test cases for the new functions, but I'm not
really sure how to add them easily. The new functions depend on specific
sequences of bits being set, while the bitstring tests appear to run for
varying bit sizes.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>

Submitted by:	Jacob Keller <jacob.e.keller@intel.com>
Reviewed by:	asomers@, erj@
MFC after:	1 week
Sponsored by:	Intel Corporation
Differential Revision:	https://reviews.freebsd.org/D22400
2019-11-21 19:57:56 +00:00
Warner Losh
c3bd3d1cc7 Install bsd.sysdir.mk.
Submitted by: Jung-uk Kim
2019-11-21 19:54:10 +00:00
Eric Joyner
2d7c33897f bitstring: exit early if _start is past size of the bitstring
bit_ffs_at and bit_ffc_at both take _start parameters which indicate to
start searching from _start onwards.

If the given _start index is past the size of the bit string, these
functions will calculate an address of the current bitstring which is
after the expected size. The function will also dereference the memory,
resulting in a read buffer overflow.

The output of the function remains correct, because the tests ensure to
stop the loop if the current bitstring chunk passes the stop bitstring
chunk, and because of a check to ensure the reported _value is never
past _nbits.

However, if <sys/bitstring.h> is ever used in code which is checked by
-fsanitize=undefined, or similar static analysis, it can produce
warnings about reading past the buffer size.

Because of the above mentioned checks, these buffer overflows do not
occur as long as _start is less than _nbits. Additionally, by definition
bit_ffs_at and bif_ffc_at should set _result to -1 in any case where the
_start is after the _nbits.

Check for this case at the start of the function and exit early if so,
preventing the buffer read overflow, and reducing the amount of
computation that occurs.

Note that it may seem odd to ever have code that could call bit_ffc_at
or bit_ffs_at with a _start value greater than _nbits. However, consider
a for-loop that used bit_ffs and bit_ffs_at to loop over a bit string
and perform some operation on each bit that was set. If the last bit of
the bit string was set, the simplest loop implementation would call
bit_ffs_at with a start of _nbits, and expect that to return -1. While
it does infact perform correctly, this is what ultimately triggers the
unexpected buffer read overflow.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>

Submitted by:	Jacob Keller <jacob.e.keller@intel.com>
Reviewed by:	asomers@, erj@
MFC after:	1 week
Sponsored by:	Intel Corporation
Differential Revision:	https://reviews.freebsd.org/D22398
2019-11-21 19:36:11 +00:00
John Baldwin
bddf73433e NIC KTLS for Chelsio T6 adapters.
This adds support for ifnet (NIC) KTLS using Chelsio T6 adapters.
Unlike the TOE-based KTLS in r353328, NIC TLS works with non-TOE
connections.

NIC KTLS on T6 is not able to use the normal TSO (LSO) path to segment
the encrypted TLS frames output by the crypto engine.  Instead, the
TOE is placed into a special setup to permit "dummy" connections to be
associated with regular sockets using KTLS.  This permits using the
TOE to segment the encrypted TLS records.  However, this approach does
have some limitations:

1) Regular TOE sockets cannot be used when the TOE is in this special
   mode.  One can use either TOE and TOE-based KTLS or NIC KTLS, but
   not both at the same time.

2) In NIC KTLS mode, the TOE is only able to accept a per-connection
   timestamp offset that varies in the upper 4 bits.  Put another way,
   only connections whose timestamp offset has the 28 lower bits
   cleared can use NIC KTLS and generate correct timestamps.  The
   driver will refuse to enable NIC KTLS on connections with a
   timestamp offset with any of the lower 28 bits set.  To use NIC
   KTLS, users can either disable TCP timestamps by setting the
   net.inet.tcp.rfc1323 sysctl to 0, or apply a local patch to the
   tcp_new_ts_offset() function to clear the lower 28 bits of the
   generated offset.

3) Because the TCP segmentation relies on fields mirrored in a TCB in
   the TOE, not all fields in a TCP packet can be sent in the TCP
   segments generated from a TLS record.  Specifically, for packets
   containing TCP options other than timestamps, the driver will
   inject an "empty" TCP packet holding the requested options (e.g. a
   SACK scoreboard) along with the segments from the TLS record.
   These empty TCP packets are counted by the
   dev.cc.N.txq.M.kern_tls_options sysctls.

Unlike TOE TLS which is able to buffer encrypted TLS records in
on-card memory to handle retransmits, NIC KTLS must re-encrypt TLS
records for retransmit requests as well as non-retransmit requests
that do not include the start of a TLS record but do include the
trailer.  The T6 NIC KTLS code tries to optimize some of the cases for
requests to transmit partial TLS records.  In particular it attempts
to minimize sending "waste" bytes that have to be given as input to
the crypto engine but are not needed on the wire to satisfy mbufs sent
from the TCP stack down to the driver.

TCP packets for TLS requests are broken down into the following
classes (with associated counters):

- Mbufs that send an entire TLS record in full do not have any waste
  bytes (dev.cc.N.txq.M.kern_tls_full).

- Mbufs that send a short TLS record that ends before the end of the
  trailer (dev.cc.N.txq.M.kern_tls_short).  For sockets using AES-CBC,
  the encryption must always start at the beginning, so if the mbuf
  starts at an offset into the TLS record, the offset bytes will be
  "waste" bytes.  For sockets using AES-GCM, the encryption can start
  at the 16 byte block before the starting offset capping the waste at
  15 bytes.

- Mbufs that send a partial TLS record that has a non-zero starting
  offset but ends at the end of the trailer
  (dev.cc.N.txq.M.kern_tls_partial).  In order to compute the
  authentication hash stored in the trailer, the entire TLS record
  must be sent as input to the crypto engine, so the bytes before the
  offset are always "waste" bytes.

In addition, other per-txq sysctls are provided:

- dev.cc.N.txq.M.kern_tls_cbc: Count of sockets sent via this txq
  using AES-CBC.

- dev.cc.N.txq.M.kern_tls_gcm: Count of sockets sent via this txq
  using AES-GCM.

- dev.cc.N.txq.M.kern_tls_fin: Count of empty FIN-only packets sent to
  compensate for the TOE engine not being able to set FIN on the last
  segment of a TLS record if the TLS record mbuf had FIN set.

- dev.cc.N.txq.M.kern_tls_records: Count of TLS records sent via this
  txq including full, short, and partial records.

- dev.cc.N.txq.M.kern_tls_octets: Count of non-waste bytes (TLS header
  and payload) sent for TLS record requests.

- dev.cc.N.txq.M.kern_tls_waste: Count of waste bytes sent for TLS
  record requests.

To enable NIC KTLS with T6, set the following tunables prior to
loading the cxgbe(4) driver:

hw.cxgbe.config_file=kern_tls
hw.cxgbe.kern_tls=1

Reviewed by:	np
Sponsored by:	Chelsio Communications
Differential Revision:	https://reviews.freebsd.org/D21962
2019-11-21 19:30:31 +00:00
Ian Lepore
e3c42ad809 Rewrite iicdev_writeto() to use a single buffer and a single iic_msg, rather
than effectively doing scatter/gather IO with a pair of iic_msgs that direct
the controller to do a single transfer with no bus STOP/START between the
two buffers.  It turns out we have multiple i2c hardware drivers that don't
honor the NOSTOP and NOSTART flags; sometimes they just try to do the
transfers anyway, creating confusing failures or leading to corrupted data.
2019-11-21 19:13:05 +00:00
Warner Losh
5bebf8b402 Remove two obsolete comments that reference splhigh/splx. 2019-11-21 18:49:54 +00:00
John Baldwin
c0632867ba Use kmod.opts.mk to fix standalone build. 2019-11-21 18:21:30 +00:00
John Baldwin
9ab286b7c1 Add a kmod.opts.mk.
This Makefile sets KERN_OPTS.  This permits kernel module Makefiles to
use KERN_OPTS to control the value of variables such as SRCS that are
used by bsd.kmod.mk for KERN_OPTS values that honor WITH/WITHOUT
options for standalone builds.
2019-11-21 18:14:26 +00:00
Hans Petter Selasky
c4e11f2231 Add USB ID for Diamond Multimedia BVU195 Display Link device.
Submitted by:	darius@dons.net.au
PR:		242128
MFC after:	1 week
Sponsored by:	Mellanox Technologies
2019-11-21 16:42:25 +00:00
Warner Losh
3cc5d6a470 Introduce bsd.sysdir.mk to consolidate looking for the kernel.
Reviewed by: jhb
Differential Revision: https://reviews.freebsd.org/D22144
2019-11-21 15:59:33 +00:00
Mateusz Piotrowski
6e02be63f4 bsdinstall.8: Fix description of DISTRIBUTIONS environment variable
Reviewed by:	bcr
Approved by:	bcr (mentor)
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D22478
2019-11-21 15:54:56 +00:00
Niclas Zeising
40c0b3b187 ObsoleteFiles.inc: add sio(4) leftovers
Add the manual page for sio(4) to ObsoleteFiles.inc, so that make delete-all
will remove it.  The manual page was removed together with sio(4) in
r354929.

Approved by:		emaste
Differential Revision:	https://reviews.freebsd.org/D22477
2019-11-21 15:38:27 +00:00
Ed Maste
8b3c5418f4 arch.7: document only FreeBSD 11.0 and later
FreeBSD 11 is the earliest supported version and there is no longer a
need to include information about differences present in earlier
releases.
2019-11-21 14:23:16 +00:00
Kyle Evans
28b1b80e0e bcm2835_sdhci: only inspect interrupts we handle
We'll write the value we read back to ack pending interrupts, but we should
at least make it clear to ourselves that we only want to ack pending
transfer interrupts.
2019-11-21 14:01:44 +00:00
Andrew Turner
a27ac4644a Disable KCSAN within a panic.
The kernel is single threaded at this point and the panic is more
important.

Sponsored by:	DARPA, AFRL
2019-11-21 13:59:01 +00:00
Ed Maste
58e70aa0cf revert contrib/amd changes accidentally committed with r354950 2019-11-21 13:57:30 +00:00
Ed Maste
1da495d063 revert r354935 and apply fix for cleandir failure
This reapplies the RISC-V GNU ld workaround from r354896, r354899, and
354900, along with a fix for the build failure during cleandir.

LINKER_TYPE was not being set during cleandir, resulting in
Malformed conditional (${LINKER_TYPE} == "bfd" && ${MACHINE} == "riscv")
from Cirrus-CI.

PR:		242109
Sponsored by:	The FreeBSD Foundation
2019-11-21 13:56:16 +00:00
Andriy Gapon
8491540808 MFV r354383: 10592 misc. metaslab and vdev related ZoL bug fixes
illumos/illumos-gate@555d674d5d
555d674d5d

https://www.illumos.org/issues/10592
  This is a collection of recent fixes from ZoL:
  8eef997679 Error path in metaslab_load_impl() forgets to drop ms_sync_lock
  928e8ad47d Introduce auxiliary metaslab histograms
  425d3237ee Get rid of space_map_update() for ms_synced_length
  6c926f426a Simplify log vdev removal code
  21e7cf5da8 zdb -L should skip leak detection altogether
  df72b8bebe Rename range_tree_verify to range_tree_verify_not_present
  75058f3303 Remove unused vdev_t fields

Portions contributed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Author: Serapheim Dimitropoulos <serapheim@delphix.com>
MFC after:	4 weeks
2019-11-21 13:35:43 +00:00
Andrew Turner
68cad68149 Add kcsan_md_unsupported from NetBSD.
It's used to ignore virtual addresses that may have a different physical
address depending on the CPU.

Sponsored by:	DARPA, AFRL
2019-11-21 13:22:23 +00:00
Andrew Turner
bba0065f0d Fix the bus_space functions with KCSAN on arm64.
Arm64 doesn't define the bus_space_set_multi_stream and
bus_space_set_region_stream functions. Don't try to define them there.

Sponsored by:	DARPA, AFRL
2019-11-21 13:12:58 +00:00
Andrew Turner
1b8c58f283 Fix for style(9): use parentheses around return statements.
Reported by:	kib
Sponsored by:	DARPA, AFRL
2019-11-21 12:29:20 +00:00
Andrew Turner
849aef496d Port the NetBSD KCSAN runtime to FreeBSD.
Update the NetBSD Kernel Concurrency Sanitizer (KCSAN) runtime to work in
the FreeBSD kernel. It is a useful tool for finding data races between
threads executing on different CPUs.

This can be enabled by enabling KCSAN in the kernel config, or by using the
GENERIC-KCSAN amd64 kernel. It works on amd64 and arm64, however the later
needs a compiler change to allow -fsanitize=thread that KCSAN uses.

Sponsored by:	DARPA, AFRL
Differential Revision:	https://reviews.freebsd.org/D22315
2019-11-21 11:22:08 +00:00
Andriy Gapon
489912da7b MFV r354382,r354385: 10601 10757 Pool allocation classes
illumos/illumos-gate@663207adb1
663207adb1

10601 Pool allocation classes
https://www.illumos.org/issues/10601
  illumos port of ZoL Pool allocation classes. Includes at least these two
  commits:
  441709695 Pool allocation classes misplacing small file blocks
  cc99f275a Pool allocation classes

10757 Add -gLp to zpool subcommands for alt vdev names
https://www.illumos.org/issues/10757
  Port from ZoL of
  d2f3e292d Add -gLp to zpool subcommands for alt vdev names
  Note that a subsequent ZoL commit changed -p to -P
  a77f29f93 Change full path subcommand flag from -p to -P

Portions contributed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Portions contributed by: Håkan Johansson <f96hajo@chalmers.se>
Portions contributed by: Richard Yao <ryao@gentoo.org>
Portions contributed by: Chunwei Chen <david.chen@nutanix.com>
Portions contributed by: loli10K <ezomori.nozomu@gmail.com>
Author: Don Brady <don.brady@delphix.com>

11541 allocation_classes feature must be enabled to add log device

illumos/illumos-gate@c1064fd7ce
c1064fd7ce

https://www.illumos.org/issues/11541
  After the allocation_classes feature was integrated, one can no longer add a
  log device to a pool unless that feature is enabled. There is an explicit check
  for this, but it is unnecessary in the case of log devices, so we should handle
  this better instead of forcing the feature to be enabled.

Author: Jerry Jelinek <jerry.jelinek@joyent.com>

FreeBSD notes.
I faithfully added the new -g, -L, -P flags, but only -g does something:
vdev GUIDs are displayed instead of device names.  -L, resolve symlinks,
and -P, display full disk paths, do nothing at the moment.
The use of special vdevs is backward compatible for read-only access, so
root pools should be bootable, but exercise caution.

MFC after:	4 weeks
2019-11-21 08:20:05 +00:00
Warner Losh
c81e4a699b Include MACHINE in the which variable selection to use section. 2019-11-21 06:48:30 +00:00
Warner Losh
7b692b8e0f Add table for MACHINE_CPUARCH
Add table and also some additional verbage of which one to use.
2019-11-21 06:44:19 +00:00
Glen Barber
fb1c294d47 Revert r354896, r354899, r354900:
Fix build.

Sponsored by:	Rubicon Communications, LLC (netgate.com)
2019-11-21 04:12:08 +00:00
Ed Maste
21ad8c5eb0 mark arm.arm (v4/v5) kernels as NO_UNIVERSE for now
r354290 removed arm.arm from universe, but arm.arm kernels were still
found and built during the kernel stage.  I'm not aware of a better way
to address this at the moment, but since there aren't many arm.arm
kernels anyhow just add an explicit NO_UNIVERSE to them.

Reported by:	rpokala
2019-11-21 03:10:02 +00:00
Kyle Evans
a8761a2a55 bcm2835_sdhci: clean up DMA segments in error handling path
Later parts assume that this would've been done if interrupts are enabled,
but this is the only case in which that wouldn't have been true. This commit
also reorders operations such that we're done touching slot/slot->intmask
before we call back into the SDHCI framework and exit.
2019-11-21 02:49:41 +00:00
Kyle Evans
b61ac06706 bcm2835_sdhci: roll back r354823
r354823 kicked DATA_END handling out of the DMA interrupt path "to make
things easy", but this was likely a mistake -- if we know we're done after
we've finished pending DMA operations, we should go ahead and acknowledge
it rather than waiting for the controller to finalize it. If it's not ready,
we'll simply re-enable interrupts and wait for it anyways, to be re-entered
in sdhci_data_intr.
2019-11-21 02:47:55 +00:00
Kyle Evans
e68976d877 Revert r354930: wrong diff, right message. 2019-11-21 02:44:05 +00:00
Kyle Evans
7af945c666 bcm2835_sdhci: clean up DMA segments in error handling path
Later parts assume that this would've been done if interrupts are enabled,
but this is the only case in which that wouldn't have been true. This commit
also reorders operations such that we're done touching slot/slot->intmask
before we call back into the SDHCI framework and exit.
2019-11-21 02:41:22 +00:00
Gleb Smirnoff
71f0077631 Remove sio(4).
It had been disconnected from build in r181233 in 2008.

Reviewed by:	imp
2019-11-21 01:24:49 +00:00
Mark Johnston
e7485a436b Avoid relying on pollution from libcasper.h.
Reported by:	sjg
Sponsored by:	The FreeBSD Foundation
2019-11-21 00:17:14 +00:00
Warner Losh
1cdb8eb8fe Add --esp/-E argument to print the currently booted ESP
Add code to decode the BootCurrent and BootXXXX variable it points at
to deduce the ESP used to boot the system. By default, it prints the
path to that device. With --unix-path (-p) it will instead print the
current mount point for the ESP, if any (or an error). With
--device-path (-d) it wil print the UEFI device path for the ESP.

Note: This is the best guess based on the UEFI variables. If the ESP
is part of a gmirror, etc, that won't be reported. If by some weird
chance there was a complicated series of chain boots, this may not be
what you want. For setups that don't add layers on top of the raw
devices, it is accurate.

Differential Revision: https://reviews.freebsd.org/D22432
2019-11-20 23:58:36 +00:00
Pedro F. Giffuni
9af6c78cd4 MFV 354917, 354918, 354919
openresolv: update to version 3.9.2

MFC after:	3 weeks
2019-11-20 23:56:20 +00:00
Warner Losh
6467506baf Create /etc/os-release file.
Each boot, regenerate /var/run/os-release based on the currently running
system. Create a /etc/os-release symlink pointing to this file (so that this
doesn't create a new reason /etc can not be mounted read-only).

This is compatible with what other systems do and is what the sysutil/os-release
port attempted to do, but in an incomplete way. Linux, Solaris and DragonFly all
implement this natively as well. The complete standard can be found at
https://www.freedesktop.org/software/systemd/man/os-release.html

Moving this to the base solves both the non-standard location problem with the
port, as well as the lack of update of this file on system update.

Bump __FreeBSD_version to 1300060

PR: 238953
Differential Revision:  https://reviews.freebsd.org/D22271
2019-11-20 23:45:31 +00:00
Pedro F. Giffuni
2d823aa1c7 Import openresolv 3.9.0
More information at: https://roy.marples.name/projects/openresolv
2019-11-20 22:13:14 +00:00
Warner Losh
599d032a6d Standardize EFI's ESP mount point.
Mount the UEFI ESP on /boot/efi. No current system uses this by default, but
there are many ad-hoc schemes that do this in /efi or /esp or /uefi and adding a
new directory at the top-level would have a much higher likelihood of
collision. Document this in /etc/mtree/BSD.root.mtree and create EFIDIR and
related variables in bsd.own.mk.

Differential Revision: https://reviews.freebsd.org/D21344
2019-11-20 21:06:29 +00:00
Alexander Motin
cc453b2272 Set handling for some "Logical unit not ready" errors.
MFC after:	2 weeks
Sponsored by:	iXsystems, Inc.
2019-11-20 20:00:03 +00:00
Conrad Meyer
c41faf5591 random/ivy: Trivial refactoring
It is clearer to me to return success/error (true/false) instead of some
retry count linked to the inline assembly implementation.

No functional change.

Approved by:	core(csprng) => csprng(markm)
Differential Revision:	https://reviews.freebsd.org/D22454
2019-11-20 19:55:43 +00:00
Conrad Meyer
61287be181 Re-apply fixed r354847
unifdef(1): Improve worst-case bound on symbol resolution

Use RB_TREE to make some algorithms O(lg N) and O(N lg N) instead of O(N)
and O(N^2).

While here, remove arbitrarily limit on number of macros understood.

Reverts r354877 and r354878, which disabled the (correct) test.

PR:		242095
Reported by:	lwhsu
2019-11-20 19:43:34 +00:00
Conrad Meyer
0877992e6b Revert r354847 for now
It was broken.

PR:		242095
Reported by:	lwhsu
2019-11-20 19:07:22 +00:00
Dimitry Andric
1d91c857d7 Fix typo: deprected -> deprecated. 2019-11-20 19:07:08 +00:00
Brooks Davis
303d2dd8f1 Make the warning for deprecated NO_ variables an error.
Support for NO_CTF, NO_DEBUG_FILES, NO_INSTALLLIB, NO_MAN, NO_PROFILE,
and NO_WARNS as deprecated in 2014 with a warning added for each one
found. Turn these into error in preperation for removal of compatability
support before FreeBSD 13.

Reviewed by:	imp
Relnotes:	yes
Sponsored by:	DARPA, AFRL
Differential Revision:	https://reviews.freebsd.org/D22448
2019-11-20 18:36:58 +00:00
Dimitry Andric
a1aaa66be8 Add explanatory comments for the different SRCS_xxx variables used in
the Makefiles for libllvm and libclang.  While here, cleanup a commented
out SRCS entry in libllvmminimal's Makefile.

MFC after:	3 days
2019-11-20 18:12:01 +00:00
Andrew Turner
09a65f9ff5 As with r354905 use uint16_t to store aflags on the stack and as function
arguments as the aflags size in vm_page_t has increased.

Sponsored by:	DARPA, AFRL
2019-11-20 18:00:43 +00:00
Ed Maste
f3cdc62295 src.conf.5: regen for several recent changes
r354289 armv6: Switch to LLD by default
r354290 Take arm.arm (armv5) out of universe
r354348 armv6, armv7: Switch to llvm-libunwind by default
r354660 Enable the RISC-V LLVM backend by default.

as well as lib32 changes

Sponsored by:	The FreeBSD Foundation
2019-11-20 17:57:46 +00:00
Andrew Turner
ad216bc10d Use atomic_load_16 to load aflags as it's a uint16_t after r354820.
Sponsored by:	DARPA, AFRL
2019-11-20 17:49:58 +00:00
Ed Maste
63b149ff63 src.conf.5: regen after r354902, WITHOUT_AMD by default 2019-11-20 17:49:32 +00:00