Alan Somers [Mon, 29 Nov 2021 01:53:31 +0000 (18:53 -0700)]
fusefs: update atime on reads when using cached attributes
When using cached attributes, whether or not the data cache is enabled,
fusefs must update a file's atime whenever it reads from it, so long as
it wasn't mounted with -o noatime. Update it in-kernel, and flush it to
the server on close or during the next setattr operation.
The downside is that close() will now frequently trigger a FUSE_SETATTR
upcall. But if you care about performance, you should be using
-o noatime anyway.
Alan Somers [Sat, 25 Sep 2021 16:16:20 +0000 (10:16 -0600)]
fusefs: fix intermittency in the dev_fuse_poll test
The DevFusePoll::access/select test would occasionally segfault. The
cause was a file descriptor that was shared between two threads. The
first thread would kill the second and close the file descriptor. But
it was possible that the second would read the file descriptor before it
shut down. That did not cause problems for kqueue, poll, or blocking
operation, but it triggered segfaults in select's macros.
Andrew Turner [Mon, 22 Nov 2021 10:25:01 +0000 (10:25 +0000)]
Move the arm64 identify_cpu SYSINIT earlier
It is used by late ifunc resolvers so needs to be at an earlier stage
of the boot. Previously it was at the same stage so may not have run
before the ifunc resolvers.
Andrew Turner [Tue, 2 Nov 2021 11:31:17 +0000 (11:31 +0000)]
Use a builtin where possible in msun
Some of the functions in msun can be implemented using a compiler
builtin function to generate a small number of instructions. Implement
this support in fma, fmax, fmin, and sqrt on arm64.
Care must be taken as the builtin can be implemented as a function
call on some architectures that lack direct support. In these cases
we need to use the original code path.
As we don't set errno on failure build with -fno-math-errno so the
toolchain doesn't convert a builtin into a function call when it
detects a failure, e.g. gcc will add a call to sqrt when the input
is negative leading to an infinite loop.
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32801
Andrew Turner [Mon, 1 Nov 2021 13:06:56 +0000 (13:06 +0000)]
Switch to Arm Optimized Routines for mem* & str*
These are the updated version of the older Cortex Strings Library we
previously used. The Arm Optimized Routines also support CPU features
that are currently in development on FreeBSD, e.g. Branch Target
Identification (BTI). Rather than add BTI support to the old code it's
easier to just use the maintained version.
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32774
Andrew Turner [Fri, 12 Nov 2021 18:28:18 +0000 (18:28 +0000)]
Move KHELP_DECLARE_MOD_UMA later in the boot
Both KHELP_DECLARE_MOD_UMA and the kernel linker SYSINIT to find
in-kernel modules run at SI_SUB_KLD, SI_ORDER_ANY. As the former
depends on the latter running first move it later in the boot,
to the new SI_SUB_KHELP. This ensures KHELP_DECLARE_MOD_UMA
module SYSINIT functions will be after the kernel linker.
Previously we may have received a panic similar to the following if
the order was incorrect:
panic: module_register_init: module named ertt not found
Reported by: bob prohaska <fbsd AT www.zefox.net>
Discussed with: imp, jhb
Sponsored by: The FreeBSD Foundation
Andrew Turner [Mon, 1 Nov 2021 11:19:57 +0000 (11:19 +0000)]
Print the correct register for the arm64 elr
In 7ec86b6609912 ("Also print symbols when printing arm64 registers")
a new function was created to print most registers. Unfortunately the
Link Register (LR) was being printed when we should have printed the
Exception Link Register (ELR).
Andrew Turner [Wed, 20 Oct 2021 09:19:55 +0000 (10:19 +0100)]
Only change the fadt revision in acpidump on i386
There is a workaround in acpidump for an old IBM laptop. This is being
hit on a modern arm64 system causing acpidump to use the wrong field
when reading a later address.
As the laptop has a 32-bit CPU the only FreeBSD support for it is i386.
As such limit the workaround to that architecture.
Reviewed by: emaste, jkim, jhb (all earlier version)
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32573
Andrew Turner [Mon, 18 Oct 2021 09:04:06 +0000 (10:04 +0100)]
Pass the ACPI ID when reading the ACPI domain
The ACPI ID may not be the same as the FreeBSD CPU id. Use the former
when finding the CPU domain as there is no requirement for it to be
identical to the latter.
Reported by: dch, kevans
Reviewed by: kevans
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D32546
Andrew Turner [Tue, 12 Oct 2021 11:39:14 +0000 (12:39 +0100)]
Stop reading the arm64 domain when it's known
There is no need to read the domain on arm64 when there is only one
in the ACPI tables. This can also happen when the table is missing
as it is unneeded.
Reported by: dch
Sponsored by: The FreeBSD Foundation
Reviewed-by: Tony Hutter <hutter2@llnl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Alexander Motin <mav@FreeBSD.org> Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org>
Closes #12828
Reviewed-by: Tony Hutter <hutter2@llnl.gov> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Alexander Motin <mav@FreeBSD.org> Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org>
Closes #12828
Mark Johnston [Fri, 19 Nov 2021 22:26:39 +0000 (17:26 -0500)]
Fix several bugs in the FreeBSD rename VOP implementation
- To avoid a use-after-free, zfsvfs->z_log needs to be loaded after the
teardown lock is acquired with ZFS_ENTER().
- Avoid leaking vnode locks in zfs_rename_relock() and zfs_rename_()
when the ZFS_ENTER() macros forces an early return.
Refactor the rename implementation so that ZFS_ENTER() can be used
safely. As a bonus, this lets us use the ZFS_VERIFY_ZP() macro instead
of open-coding its implementation.
Reported-by: Peter Holm <pho@FreeBSD.org> Tested-by: Peter Holm <pho@FreeBSD.org> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com> Signed-off-by: Mark Johnston <markj@FreeBSD.org> Sponsored-by: The FreeBSD Foundation
Closes #12717
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Signed-off-by: Pawel Jakub Dawidek <pawel@dawidek.net>
Closes #12748
Mark Johnston [Thu, 28 Oct 2021 17:25:26 +0000 (13:25 -0400)]
Exit the teardown section later in rename on FreeBSD
We have to hold the teardown lock while dereferencing zfsvfs->z_os and,
I believe, when committing to the ZIL.
Note that jumping to the "out" label, "error" is always non-zero.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes #12704
Mark Johnston [Thu, 28 Oct 2021 15:58:57 +0000 (11:58 -0400)]
Fix potential use-after-frees in FreeBSD getpages and setattr VOPs
The objset object is reallocated during certain dataset operations, such
as rollbacks, so the objset pointer must be loaded after acquiring the
teardown lock.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes #12704
Kristof Provost [Thu, 2 Dec 2021 17:47:40 +0000 (18:47 +0100)]
pf tests: more thorough pfsync defer test
Add a somewhat more extensive pfsync defer mode test. Ensure that pfsync
actually delays the state creating packet until after it has sent the
pfsync update and given the peer time to create the state.
Ideally the test should validate the pfsync state update and generate an
ack message, but to keep the test simple we rely on the timeout of the
deferred packet instead.
Kristof Provost [Thu, 2 Dec 2021 17:39:23 +0000 (18:39 +0100)]
pfsync: fix incorrect enabling of defer mode
When we exposed the PFSYNCF_OK flag to userspace in 5f5bf88949d we
unintentionally caused defer mode to always be enabled.
The ioctl check only looked for nonzero, not for the PFSYNCF_DEFER flag.
Kristof Provost [Thu, 2 Dec 2021 16:42:56 +0000 (17:42 +0100)]
pfsync: locking fixes
* Ensure we unlock the pfsync lock in pfsync_defer()
* We must hold the bucket lock when calling pfsync_push()
* The pfsync_defer_tmo() callout locks the bucket lock, not the pfsync
lock
Kristof Provost [Thu, 2 Dec 2021 16:30:36 +0000 (17:30 +0100)]
pfsync: fix defer timeout
Don't use a fixed number of ticks, but take hz into account so we have a
consistent timeout, regardless of what hz is set up.
Use a 20ms timeout, becaues that's what OpenBSD uses.
That commit changed libc to use the MI pdfork implementation, but with
an incremental build the object file for the pdfork.S stub lingers and
causes a linker error.
Cleaning the depend file is not enouch, so modify clean_deps() to remove
object files as well, and add a call to ensure that pdfork.*o is
cleaned. The new file is _pdfork.o.
Reported by: jhb
Reviewed by: emaste
Fixes: cbdec8db18b5 ("libc: Add pdfork to the list of interposed system calls")
Sponsored by: The FreeBSD Foundation
Mark Johnston [Sun, 5 Dec 2021 15:45:12 +0000 (10:45 -0500)]
ng_ubt: Avoid attaching to several newer Intel controllers
Like other Intel controllers, these require firmware to be loaded, and
generic ng_ubt attach causes them to lock up until a power cycle.
However, their firmware interface for querying version info and loading
operational firmware is different from that implemented by ng_ubt_intel
and iwmbtfw, so they are not usable yet. Just disable attach for now to
avoid stalls during USB device enumeration.
Andreas Wetzel [Mon, 6 Dec 2021 07:21:38 +0000 (09:21 +0200)]
rtwn/usb: add product ID for Asus USB N10 Nano Rev. B1
According to information found on the internet the following products
use exactly the same hardware but probably different USB IDs:
- Edimax EW-7811Un V2 (v2)
- Edimax EW-7811GLN 2.0A (v2)
I am not adding them as I cannot verify.
Ed Maste [Sun, 28 Nov 2021 17:50:13 +0000 (12:50 -0500)]
sbin: build ping if at least one of INET & INET6 is enabled
It does not build (and serves no purpose) if neither is true (i.e.,
building WITHOUT_INET and WITHOUT_INET6). Also add an explicit error
in ping to make this case clear.
Alan Somers [Wed, 6 Oct 2021 22:54:59 +0000 (16:54 -0600)]
ping: fix parsing of options including '4' and '6'
ping uses a two-pass option parser. The first pass determines whether
ipv4 or ipv6 is desired, and the second parses the rest of the options.
But the first pass wrongly detects a '4' or '6' in an option's value as
a request to use ipv6 or ipv6 respectively, for example in an invocation
like "ping -c6 1.2.3.4".
Fix this confusion by including all options in the first round of
parsing, but ignoring those unrelated to ipv4/ipv6 selection.
Ed Maste [Mon, 6 Dec 2021 21:27:48 +0000 (16:27 -0500)]
Remove FREEBSD-vendor files
These files were intended to track version and perhaps maintainership
information for contrib software. However, they were never used beyond
bzip2, netcat, and OpenSSH, and generally haven't been kept up to date
recently (my OpenSSH 8.7p1 update notwithstanding). Just remove them to
avoid having confusing or outdated information.
Suggested by: des
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Ed Maste [Tue, 2 Mar 2021 01:57:36 +0000 (20:57 -0500)]
libc: Use musl's optimized strchr and strchrnul
Parentheses added to HASZERO macro to avoid a GCC warning, and formatted
with clang-format as we have adopted these and don't consider them
'contrib' code.
Obtained from: musl (snapshot at commit 4d0a82170a25)
Reviewed by: kib (libc integration), mjg (both earlier)
MFC after: 1 month
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D17630
Ed Maste [Thu, 25 Jun 2020 00:42:10 +0000 (20:42 -0400)]
bxe: tag files to skip clang-format formatting
bxe contains three files which are sets of constants or other data, and
might be auto-generated or have an upstream. They are rather large
files and clang-format takes quite some time when run against them, so
just skip formatting.
Kristof Provost [Thu, 2 Dec 2021 07:22:34 +0000 (08:22 +0100)]
if_pflog: fix packet length
There were two issues with the new pflog packet length.
The first is that the length is expected to be a multiple of
sizeof(long), but we'd assumed it had to be a multiple of
sizeof(uint32_t).
The second is that there's some broken software out there (such as
Wireshark) that makes incorrect assumptions about the amount of padding.
That is, Wireshark assumes there's always three bytes of padding, rather
than however much is needed to get to a multiple of sizeof(long).
Fix this by adding extra padding, and a fake field to maintain
Wireshark's assumption.
Rick Macklem [Sat, 27 Nov 2021 23:02:04 +0000 (15:02 -0800)]
nfscl: Sanity check the Sequence slotid in reply
The slotid in the Sequence reply must be the same as
in the request. Check that it is the same and log
a console message if it is not, plus set it to the
correct value.
Colin Percival [Tue, 23 Nov 2021 04:16:34 +0000 (20:16 -0800)]
EC2: Turn off IPv6 DAD
Disable Duplicate Address Detection in EC2 instances. The networking
configuration in EC2, with IPv6 addresses assigned by DHCPv6 and
host egress filtering, makes "duplicate addresses" impossible.
This speeds up the boot process in EC2 by 2 seconds.
Dimitry Andric [Sun, 5 Dec 2021 17:54:13 +0000 (18:54 +0100)]
Apply fix for clang crashing on invalid -Wa,-march= values
Merge commit df08b2fe8b35 from llvm git (by Dimitry Andric):
[AArch64] Avoid crashing on invalid -Wa,-march= values
As reported in https://bugs.freebsd.org/260078, the gnutls Makefiles
pass -Wa,-march=all to compile a number of assembly files. Clang does
not support this -march value, but because of a mistake in handling
the arguments, an unitialized Arg pointer is dereferenced, which can
cause a segfault.
Work around this by adding a check if the local WaMArch variable is
initialized, and if so, using its value in the diagnostic message.
Mitchell Horne [Mon, 6 Dec 2021 14:09:54 +0000 (10:09 -0400)]
ucom: s/sio/ucom/
Seems like a copy-paste error, or at least this made more sense when the
sio(4) driver still existed. This modifies the debug port name displayed
at boot, but otherwise has no functional change.
Reviewed by: hselasky
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D33278
Mitchell Horne [Sun, 5 Dec 2021 15:11:55 +0000 (11:11 -0400)]
x86: remove unused T_USER flag
It stopped being used in 3c256f5395aa, when trap() was reorganized to
have separate switch statements for user and kernel traps. Remove the
two leftover references and the flag itself.
Reviewed by: kib
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D33253
When no mask is supplied to the ioctl adding an Internet interface
address, revert to using the historical class mask rather than a
single default. Similarly for the NFS bootp code.
Mike Karels [Thu, 28 Oct 2021 23:39:43 +0000 (18:39 -0500)]
systat: clean up code assuming network classes
Similar to netstat, clean up code that uses inet_lnaof() to check for
binding to "host 0" (lowest host on network) as a "network" bind.
Such things don't happen, and current networks are seldom if ever
found in /etc/networks.
Mike Karels [Wed, 27 Oct 2021 03:12:24 +0000 (22:12 -0500)]
sockstat: change check for wildcard sockets to avoid historical classes
sockstat was checking whether a bound address was "host 0", the lowest
host on a network, using inet_lnaof(). This only works for class A/B/C.
However, it isn't useful to bind such an address unless it is really
the unspecified address INADDR_ANY. Change the check to to use that.
Mike Karels [Wed, 27 Oct 2021 03:39:10 +0000 (22:39 -0500)]
netstat: reduce use of historical Internet classes
When attempting to characterize bound addresses, netstat was checking
for host 0 on a (historical) net using inet_lnaof(). Such addresses
are not normally bound, as they would not work, with the exception
of the unspecified address, INADDR_ANY. Check for that explicitly.
Similarly, don't check bound addresses for a match to a network name.
Mike Karels [Wed, 27 Oct 2021 03:48:23 +0000 (22:48 -0500)]
mountd: deprecate exports to a network without mask
The exports file format allows export to a network using an explicit
mask or prefix length (CIDR). It also allows a network with just
a dotted address, in which case the historical mask was used.
Deprecate this usage, and warn when it is used. Document that this
is deprecated.
Mike Karels [Wed, 27 Oct 2021 03:25:09 +0000 (22:25 -0500)]
man pages: deprecate Internet Class A/B/C
Mark functions inet_netof(), inet_lnaof(), and inet_makeaddr() as
deprecated, as they assume Class A/B/C. inet_makeaddr() mostly works
when networks are a multiple of 8 bits, but warn for anything other
than historical classes. Reduce other mentions of network classes.
Mike Karels [Thu, 28 Oct 2021 14:32:31 +0000 (09:32 -0500)]
ifconfig: warn if setting an Internet address without mask
Add a postproc function for af_inet, and add interface flags as a
parameter. Check there if setting an address without a mask unless
the interface is loopback or point-to-point, where mask is not really
meaningful; warn if so. This will hopefully be an error in the future.
Mike Karels [Wed, 27 Oct 2021 03:01:09 +0000 (22:01 -0500)]
kernel: deprecate Internet Class A/B/C
Hide historical Class A/B/C macros unless IN_HISTORICAL_NETS is defined;
define it for user level. Define IN_MULTICAST separately from IN_CLASSD,
and use it in pf instead of IN_CLASSD. Stop using class for setting
default masks when not specified; instead, define new default mask
(24 bits). Warn when an Internet address is set without a mask.
Andriy Gapon [Fri, 19 Nov 2021 07:56:30 +0000 (09:56 +0200)]
iflib_stop: drain rx tasks to prevent any data races
iflib_stop modifies iflib data structures that are used by _task_fn_rx,
most prominently the free lists. So, iflib_stop has to ensure that the
rx task threads are not active.
This should help to fix a crash seen when iflib_if_ioctl (e.g.,
SIOCSIFCAP) is called while there is already traffic flowing.
The crash has been seen on VMWare guests with vmxnet3 driver.
My guess is that on physical hardware the couple of 1ms delays that
iflib_stop has after disabling interrupts are enough for the queued work
to be completed before any iflib state is touched.
But on busy hypervisors the guests might not get enough CPU time to
complete the work, thus there can be a race between the taskqueue
threads and the work done to handle an ioctl, specifically in iflib_stop
and iflib_init_locked.
Andriy Gapon [Fri, 26 Nov 2021 07:44:30 +0000 (09:44 +0200)]
twsi: unify error handling, explicitly handle more conditions
twsi_error() is a new function that stops the current transfer and sets
up softc when an error condition is detected.
TWSI_STATUS_DATA_WR_NACK, TWSI_STATUS_BUS_ERROR and
TWSI_STATUS_ARBITRATION_LOST are now handled explicitly rather than
via the catch-all unknown status.
Also, twsi_intr() now calls wakeup() in a single place when the
transfer is finished.
Andriy Gapon [Fri, 26 Nov 2021 07:25:01 +0000 (09:25 +0200)]
twsi: protect interaction between twsi_transfer and twsi_intr
All accesses to softc are now done under a mutex to prevent data races
between the open context and the interrupt handler.
Additionally, the wait time in twsi_transfer is bounded now.
Previously we could get stuck there forever if an interrupt got lost.
Michael Tuexen [Sat, 4 Dec 2021 14:00:05 +0000 (15:00 +0100)]
tcpdrop: allow TCP connections to be filtered by cc-algo
In addition to filtering by stack and state, allow filtering
by the congestion control module used. Choose the command line
options to be consistent with the ones of sockstat.