Rather than allocating however much memory userspace asks for we only
allocate enough for a handful of states, and copy to userspace for each
completed row.
We start out with enough space for 16 states (per row), but grow that as
required. In most configurations we expect at most a handful of states
per row (more than that would have other negative effects on packet
processing performance).
Reviewed by: mjg
MFC after: 1 week
Sponsored by: Rubicon Communications, LLC ("Netgate")
Differential Revision: https://reviews.freebsd.org/D31111
Add a new version of the DIOCGETSTATES call, which extends the struct to
include the original interface information.
MFC after: 1 week
Sponsored by: Rubicon Communications, LLC ("Netgate")
Differential Revision: https://reviews.freebsd.org/D31097
Linux has had an "nconnect" NFS mount option for some time.
It specifies that N (up to 16) TCP connections are to created for a mount,
instead of just one TCP connection.
A discussion on freebsd-net@ indicated that this could improve
client<-->server network bandwidth, if either the client or server
have one of the following:
- multiple network ports aggregated to-gether with lagg/lacp.
- a fast NIC that is using multiple queues
It does result in using more IP port#s and might increase server
peak load for a client.
One difference from the Linux implementation is that this implementation
uses the first TCP connection for all RPCs composed of small messages
and uses the additional TCP connections for RPCs that normally have
large messages (Read/Readdir/Write). The Linux implementation spreads
all RPCs across all TCP connections in a round robin fashion, whereas
this implementation spreads Read/Readdir/Write across the additional
TCP connections in a round robin fashion.
Reviewed by: markj
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D30970
Rather than extending syr827 for syr828 (as initially done in D31103)
switch to the Fairchild Semiconductor Corporation fan53555 implementation
which is in-tree but was not attached to the build. The fan53555
implementation also supports syr827/syr8278 already. [1]
Update NOTES and the arm64 GENERIC configuration for the switch.
syr827 for now stays in the tree but is not used by any
kernel configuration.
Suggested by: mmel [1]
Reviewed by: mmel, manu
Differential Revision: https://reviews.freebsd.org/D31112
Only print this warning when boot verbose is enabled.
This can get pretty annoying (and useless) in some systems.
Reviewed by: kp
MFC after: 1 week
Sponsored by: Rubicon Communications, LLC ("Netgate")
The character between the E's was the letter O, however in the Arm
Documentation and XML the character is the number 0 (zero).
Sponsored by: The FreeBSD Foundation
This fixes the incorrect use of a sysctl add to u64. It
was for a useconds time, but on 32 bit platforms its
not a u64. Instead use the long directive.
Reviewed by: tuexen
Sponsored by: Netflix Inc.
Differential Revision: https://reviews.freebsd.org/D31107
Happily this wasn't a real bug, because pf_killstates() never fails, but
we should check the return value anyway, in case it does ever start
returning errors.
Reported by: clang --analyze
MFC after: 1 week
Sponsored by: Rubicon Communications, LLC ("Netgate")
pidx is never NULL, and is used unconditionally later on in the
function.
Add an assertion, as documentation for the requirement to provide an idx
pointer.
Reported by: clang --analyze
MFC after: 1 week
Sponsored by: Rubicon Communications, LLC ("Netgate")
Indicate that this is a kernel-only structure, and make it easier to
distinguish from others used to communicate with userspace.
Reviewed by: mjg
MFC after: 1 week
Sponsored by: Rubicon Communications, LLC ("Netgate")
Differential Revision: https://reviews.freebsd.org/D31096
The first release of an interrupt in a situation where the interrupt table
is full should schedule a full table check the next time an interrupt is
allocated. A full check is necessary to ensure maximum separation between
the order of allocation and the order of release.
Submitted by: ehem_freebsd@m5p.com (initial version)
Discussed in: https://reviews.freebsd.org/D29310
MFC after: 4 weeks
A user reported that when compiling without KERN_TLS, and
with -O0, the kernel failed to link due to ktls_disable_ifnet()
being undefined. Making the shim static works around this issue.
Reported by: Gary Jennejohn
Sponsored by: Netflix
When the initial fcmpset in pmap_promote_l2() fails, there is no need
to repeat the check for the physical address being 2MB aligned or for
the accessed bit being set. While the pmap is locked the hardware can
only transition the accessed bit from 0 to 1, and we have already
determined that it is 1 when the fcmpset fails.
MFC after: 1 week
When fixing another bug, I noticed that the alternate
TCP stacks do not build when various combinations of
ipv4 and ipv6 are disabled.
Reviewed by: rrs, tuexen
Differential Revision: https://reviews.freebsd.org/D31094
Sponsored by: Netflix
HPTS drives both rack and bbr, and yet there have been many complaints
about performance. This bit of work restructures hpts to help reduce CPU
overhead. It does this by now instead of relying on the timer/callout to
drive it instead use user return from a system call as well as lro flushes
to drive hpts. The timer becomes a backstop that dynamically adjusts
based on how "late" we are.
Reviewed by: tuexen, glebius
Sponsored by: Netflix Inc.
Differential Revision: https://reviews.freebsd.org/D31083
These ABIs do not use umtx at all, so there is nothing to clean.
Cloudabi references to umtx keys do not require any cleanups anyway.
Requested by: dchagin
Reviewed by: dchagin, markj
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D30987
Use sysentvec hooks to only call umtx_thread_exit/umtx_exec, which handle
robust mutexes, for native FreeBSD ABI. Similarly, there is no sense
in calling sigfastblock_clear() for non-native ABIs.
Requested by: dchagin
Reviewed by: dchagin, markj (previous version)
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D30987
after itimers are stopped. This makes it more usable for e.g. native FreeBSD
ABI sysentvecs.
Reviewed by: dchagin, markj
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D30987
Unlike sv_onexec(), it is called from the old (pre-exec) sysentvec structure.
The old vmspace for the process is still intact during the call.
Reviewed by: dchagin, markj
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D30987
The intent is to eliminate the MT_NOINIT flag and consequently a branch
from the constructor.
Reviewed by: gallatin
Sponsored by: Rubicon Communications, LLC ("Netgate")
Differential Revision: https://reviews.freebsd.org/D31080
Add the missing macros and decode all the fields as described in the
Arm Architecture System Registers XML corresponding to Armv8.5.
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D30983
This makes the ada(4) driver use UMA for its CCBs. While it's
da(4) counterpart needs some more testing, this one seems to be
safe now.
Please let me know via email if you notice any suspicious kernel
messages,
Reviewed By: imp
Sponsored by: NetApp, Inc.
Sponsored by: Klara, Inc.
Differential Revision: https://reviews.freebsd.org/D30567
During the removal of cam_sim_alloc_dev() in
aeb04e88f5 for sdhci.c and the
follow-up build-fix in a72af82e31
slot->dev and slot->bus got mixed up for MMCCAM; slot->dev is
only used in the !MMCCAM case so is uninitialised here leading to
a panic; switch back to slot->bus to return to the status quo.
Reviewed by: imp (ack on arm@)
X-Differential Revision: https://reviews.freebsd.org/D30857
There are several cases where we make a goodput measurement and we are running
out of data when we decide to make the measurement. In reality we should not make
such a measurement if there is no chance we can have "enough" data. There is also
some corner case TLP's that end up not registering as a TLP like they should, we
fix this by pushing the doing_tlp setup to the actual timeout that knows it did
a TLP. This makes it so we always have the appropriate flag on the sendmap
indicating a TLP being done as well as count correctly so we make no more
that two TLP's.
In addressing the goodput lets also add a "quality" metric that can be viewed via
blackbox logs so that a casual observer does not have to figure out how good
of a measurement it is. This is needed due to the fact that we may still make
a measurement that is of a poorer quality as we run out of data but still have
a minimal amount of data to make a measurement.
Reviewed by: tuexen
Sponsored by: Netflix Inc.
Differential Revision: https://reviews.freebsd.org/D31076
Ifnet (inline) hw kTLS NICs typically keep state within
a TLS record, so that when transmitting in-order,
they can continue encryption on each segment sent without
DMA'ing extra state from the host.
This breaks down when transmits are out of order (eg,
TCP retransmits). In this case, the NIC must re-DMA
the entire TLS record up to and including the segment
being retransmitted. This means that when re-transmitting
the last 1448 byte segment of a TLS record, the NIC will
have to re-DMA the entire 16KB TLS record. This can lead
to the NIC running out of PCIe bus bandwidth well before
it saturates the network link if a lot of TCP connections have
a high retransmoit rate.
This change introduces a new sysctl (kern.ipc.tls.ifnet_max_rexmit_pct),
where TCP connections with higher retransmit rate will be
switched to SW kTLS so as to conserve PCIe bandwidth.
Reviewed by: hselasky, markj, rrs
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D30908
No functional changes. Do not MFC this, it changes kernel ABI.
Sponsored by: NetApp, Inc.
Sponsored by: Klara, Inc.
Differential Revision: https://reviews.freebsd.org/D30698
Before UMA CCBs, all CCBs were of the same size, and could
be trivially copied using bcopy(9). Now we have to preserve
alloc_flags, otherwise we might end up attempting to free
stack-allocated CCB to UMA; we also need to take CCB size
into account.
This fixes kernel panic which would occur when trying to access
a stopped (as in, SCSI START STOP, also "ctladm stop") SCSI device.
Reported By: Gary Jennejohn <gljennjohn@gmail.com>
Tested By: Gary Jennejohn <gljennjohn@gmail.com>
Reviewed By: imp
Sponsored by: NetApp, Inc.
Sponsored by: Klara, Inc.
Differential Revision: https://reviews.freebsd.org/D31054
With the v5.13 device-tree update speed of the CPU switch port was
changed to 2.5G. Reflect that in the driver.
Submitted by: Kornel Duleba <mindal@semihalf.com>
Obtained from: Semihalf
Sponsored by: Alstom Group
New Samsung 980 SSDs report Namespace Preferred Write Alignment of
8 (4KB) and Namespace Preferred Write Granularity of 32 (16KB).
My quick tests show that 16KB is a minimal sequential write size
when the SSD reaches peak IOPS, so writing much less is very slow.
But writing slightly less or slightly more does not change much,
so it seems not so much a size granularity as minimum I/O size.
Thinking about different stripesize consumers:
- Partition alignment should be based on NPWA by definition.
- ZFS ashift in part of forcing alignment of all I/Os should also
be based on NPWA. In part of forcing size granularity, if really
needed, it may be set to NPWG, but too big value can make ZFS too
space-inefficient, and the 16KB is actually the biggest supported
value there now.
- ZFS recordsize/volblocksize could potentially be tuned up toward
NPWG to work as I/O size granularity, but enabled compression makes
it too fuzzy. And those are normally user-configurable things.
- ZFS I/O aggregation code could definitely use Optimal Write Size
value and may be NPWG, but we don't have fields in GEOM now to report
the minimal and optimal I/O sizes, and even maximal is not reported
outside GEOM DISK to be used by ZFS.
MFC after: 1 week
In a few places, on a failed compare-and-set, both the amd64 pmap and
the arm64 pmap repeat tests on bits that won't change state while the
pmap is locked. Eliminate some of these unnecessary tests.
Reviewed by: andrew, kib, markj
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D31014
We removed sysinstall(8) back in 2011, so this workaround should be long
since unnecessary. This workaround can end up breaking cases that are
hit in the real world, such as dd'ing a small pre-built disk image to a
large partition that you intend to grow on first boot and uses a UFS
disk label for / in its /etc/fstab (as the only reliable thing a raw UFS
image can reference).
Reviewed by: imp, mckusick
Differential Revision: https://reviews.freebsd.org/D30825
Since commit 2dd1bdf183 in 2016 the r_start and r_end fields have been
rman_res_t, which was briefly unsigned long, but commit da1b038af9
changed the typedef to be uintmax_t instead. C99 is also something we
assume these days.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D30808
Instead serialize against these operations with a dedicated lock.
Prior to the change, When pushing 17 mln pps of traffic, calling
DIOCRGETTSTATS in a loop would restrict throughput to about 7 mln. With
the change there is no slowdown.
Reviewed by: kp (previous version)
Sponsored by: Rubicon Communications, LLC ("Netgate")
Creating tables and zeroing their counters induces excessive IPIs (14
per table), which in turns kills single- and multi-threaded performance.
Work around the problem by extending per-CPU counters with a general
counter populated on "zeroing" requests -- it stores the currently found
sum. Then requests to report the current value are the sum of per-CPU
counters subtracted by the saved value.
Sample timings when loading a config with 100k tables on a 104-way box:
stock:
pfctl -f tables100000.conf 0.39s user 69.37s system 99% cpu 1:09.76 total
pfctl -f tables100000.conf 0.40s user 68.14s system 99% cpu 1:08.54 total
patched:
pfctl -f tables100000.conf 0.35s user 6.41s system 99% cpu 6.771 total
pfctl -f tables100000.conf 0.48s user 6.47s system 99% cpu 6.949 total
Reviewed by: kp (previous version)
Sponsored by: Rubicon Communications, LLC ("Netgate")
as a thin wrapper around native version found in sys/seqc.h.
This replaces out-of-base GPLv2-licensed code used by drm-kmod.
Reviewed by: hselasky
Differential revision: https://reviews.freebsd.org/D31006