Use a separate state for when a request to set RX_QUIESCE has been
sent but the resulting TCB reply has not been received. In
particular, this correctly handles the case where data has been
received and queued in the receive queue before the quiesce request
takes effect.
Reviewed by: np
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D44435
(cherry picked from commit 9978c6289d)
In particular, don't try to byteswap the values as 64-bit integers and
always print a non-empty version as a string.
Reviewed by: chuck, imp
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D44121
(cherry picked from commit 7485926e09)
This fixes a panic if a driver uses bus_set_resource to add a resource
that fails to reserve and then deletes the resource via
bus_delete_resource.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D43891
(cherry picked from commit e05436d577)
The rman description strings now match those used in the PCI-PCI
bridge driver. Using more specific names removes ambiguity in devinfo -u
output on systems with multiple host to PCI bridges.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D43890
(cherry picked from commit 51f8ac224f)
acpi_set_resource excludes certain types of resources for certain
devices. The intention of this is to avoid adding resource entries
for bogus resources enumerated via _CRS. However, this also prevents
drivers from adding those resources explicitly if needed. To fix
this, move the logic to exclude these resources into an ignore hook
used when parsing _CRS to create the initial set of resources for each
device.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D43892
(cherry picked from commit 055c1fe230)
No functional change, but it is cleaner to use the existing generic
wrappers rather than KOBJ methods directly.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D43686
(cherry picked from commit 1fb5445206)
These functions are not used by pcib subclasses so do not need to
be exposed in pcib_private.h.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D43688
(cherry picked from commit 03719c6514)
Most ULP modes in cxgbe's TOE are enabled on the fly when a protocol
is needed (e.g. ULP_MODE_ISCSI is enabled by cxgbei when offloading a
connection using iSCSI, and ULP_MODE_TLS is enabled when RX TLS keys
are programmed for a TOE connection). The one exception to this is
ULP_MODE_TCPDDP.
Currently the cxgbe driver enables ULP_MODE_TCPDDP when a TOE
connection is first created. However, since DDP connections cannot be
converted to other connection types, this requires some special
handling in the driver. For example, iSCSI daemons use the SO_NO_DDP
socket option to ensure TOE connections use ULP_MODE_NONE so they can
be converted to ULP_MODE_ISCSI. Similarly, using TLS receive offload
(ULP_MODE_TLS) requires disabling TCP DDP for new connections by
default.
This commit changes cxgbe to instead switch a connection from
ULP_MODE_NONE to ULP_MODE_TCPDDP when a connection first attempts to
use TCP DDP via aio_read(2). This permits connections to always start
as ULP_MODE_NONE and switch to a protocol-specific mode as needed.
Reviewed by: np
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D43670
(cherry picked from commit a5a965d759)
Previously this was only limited on T6 cards to support switching from
ULP_MODE_NONE to ULP_MODE_TLS. To support switching to
ULP_MODE_TCPDDP, enable this for all adapters.
Reviewed by: np
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D43669
(cherry picked from commit afaefb7737)
Update the I2C controller logic to be more consistent with the
newer version of the controller reference manual.
This makes it work better on modern LS/LX platforms and avoids
unnecessary delays. Also fixes a lock leak.
Tested by: bz (LS1088a FDT), Pierre-Luc Drouin (Honeycomb, ACPI)
(cherry picked from commit 44847114bd)
Move the code from the arm specific to the iicbus controller directory.
Split up between general logic and bus attachment code.
Add support for ACPI attachment in addition to FDT.
Tested by: bz (LS1088a FDT), Pierre-Luc Drouin (Honeycomb, ACPI)
Based on: D24917 by Val Packett (initial early version)
(cherry picked from commit 5ca8e32633)
Add support for the 4 channel I2C switch from NXP by adding a new
description struct and the list entries. Compared to x=[2345] which
require code to support the INT, for this one no further changes are
needed.
Tested on: WHLE-LS1088A using a SPF+
Reviewed by: imp
(cherry picked from commit c81df1c112)
Enabling 11n for ath(4) so far was handled by a kernel option, which
was only enabled for certain kernel configurations.
In order to allow loading ath(4) as a module with 11n support on
all platforms, remove the kernel option and unconditionally enable
11n in ath(4).
Reported by: pkubaj
Reviewed by: adrian, imp
(cherry picked from commit 89c1e54a71)
Setting media to autoselect would always return EOPNOTSUPP.
As autoselect is the only valid media, this change now returns
success instead.
PR: 264253
Reported by: Prakash Shiva <prakashs0234@gmail.com>
Reviewed by: Dexuan Cui <decui microsoft com>, whu
Approved by: whu
MFC after: 2 weeks
(cherry picked from commit 63a7c4be4a)
Audio devices with more than 8 channels need bitperfect mode to operate,
the vchan processing chain is limited to 8 channels. For these devices,
let applications properly select a certain number of channels supported
by the driver, instead of mapping the request to a vchan format.
Reviewed by: emaste
Pull Request: https://github.com/freebsd/freebsd-src/pull/914
(cherry picked from commit 61c8316796)
Without appropriate load-synchronization to pair with store barriers in
wg_encrypt() and wg_decrypt(), the compiler and hardware are often
allowed to reorder these loads in wg_deliver_out() and wg_deliver_in()
such that we end up with a garbage or intermediate mbuf that we try to
pass on. The issue is particularly prevalent with the weaker
memory models of !x86 platforms.
Switch from the big-hammer wmb() to more explicit acq/rel atomics to
both make it obvious what we're syncing up with, and to avoid somewhat
hefty fences on platforms that don't necessarily need this.
With this patch, my dual-iperf3 reproducer is dramatically more stable
than it is without on aarch64.
PR: 264115
Reviewed by: andrew, zlei
(cherry picked from commit 3705d679a6)
Priority of psci_reboot set so that it is run after shutdown_panic is
executed. This is to provide uniform experience with other platforms.
(cherry picked from commit 0f354b2b58)
A syscon_power instance can handle either poweroff or reboot, but not
both. If the instance handles reboot then set its priority to be after
shutdown_panic.
This is to provide uniform experience with other platforms.
(cherry picked from commit 5f7312a0d7)
Previously, the function would return early if RB_POWERCYCLE was
specified without RB_POWEROFF. Those flags are exclusive at the moment,
that is, they are never set together.
Søren Schmidt (sos) uses a similar but extended patch locally.
(cherry picked from commit 8b408fc6f2)
(cherry picked from commit a044cf60bd)
Previosuly, USB_IFACE_DRIVER_ACTIVE would report that the driver is
active even after it detached. That's because a device(9) still
remains.
So, add device_is_alive(9) check for more accurate reporting.
Reviewed by: markj
(cherry picked from commit 8f374fa528)
Despite still being in production the device appeared not able to use
memory above BUS_SPACE_MAXADDR_32BIT, and if your desktop has a lot of
memory there is a high chance driver would allocate inaccessible memory.
Submitted by: wulf
Reviewed by: mav
(cherry picked from commit 89189224ad)
The msi address contains apic id. The code in vmbus_pcib_map_msi()
treats it as cpu id, which could cause mis-configuration of msix
IRQs, leading to missing interrupts for SRIOV devices. This happens
when apic id is not the same as cpu id on certain large VM sizes
with multiple numa domains in Azure. Fix this issue by correctly
mapping apic ids to cpu ids.
On vPCI version before 1.4, it only supports up to 64 vcpus
for msi/msix interrupt. This change also adds a check and returns
error if the vcpu_id is greater than 63.
Reported by: NetApp
Tested by: whu
Sponsored by: Microsoft
(cherry picked from commit 999174ba03)
The sample rate selection of snd_uaudio(4) at runtime was implicitly
relying on a specific order in the device config list. In case a default
was set through the hw.usb.uaudio.default_rate sysctl tunable, commit
42fdcd9fd9 removed a duplicate sample rate entry from that list, which
inadvertently broke sample rate selection at runtime. Implement sample
rate selection in a way that works for any order in the device config
list.
Reported by: Lexi Winter <lexi@le-fay.org>
MFC after: 1 week
Reviewed by: christos
Differential Revision: https://reviews.freebsd.org/D44051
(cherry picked from commit a9341f0f0a)
Fixing the error messages when detaching the mana gdma devices
showed in dmesg: "Device leaked memory resources".
Reported by: NetApp
MFC after: 3 days
Sponsored by: Microsoft
(cherry picked from commit 47e99e5bc5)
snd_uaudio(4) selects the first maching rate/channel/bit/format/buffer
configuration for use during attach, even though it will print the rest
of the supported configurations detected. To make this clear, mark the
selected playback and recording configurations with a "selected" string.
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Reviewed by: dev_submerge.ch
Differential Revision: https://reviews.freebsd.org/D43766
(cherry picked from commit f60e99f45e)
Let the USB audio descriptor iteration detect configurations with more
channels and larger sample size, even when the following global sysctl
tunables are set to a lower value:
hw.usb.uaudio.default_channels hw.usb.uaudio.default_bits
This improves utility and is closer to the meaning of default.
Also, do not create duplicate sample rate entries.
MFC after: 2 weeks
Reviewed by: christos
Differential Revision: https://reviews.freebsd.org/D43679
(cherry picked from commit 42fdcd9fd9)
Adapt the length of the driver side audio buffer to the USB transfer
interval, which is adjustable through the buffer_ms tunable. This
eliminates unnecessary latency in USB audio playback.
To reduce power consumption caused by frequent CPU wakeups, increase the
default buffer_ms value to 4ms. In combination with adaptive buffer
length, this still results in less roundtrip latency compared to the
previous 2ms default.
Extend the buffer_ms value range to 1ms for low latency applications.
MFC after: 2 weeks
Reviewed by: christos
Differential Revision: https://reviews.freebsd.org/D41942
(cherry picked from commit b2e97edffd)
The goal of reserving firmware-assigned resources is to ensure that
"wildcard" resource allocation requests will not claim an address
range that is actually in use even if no attached driver is actively
using that range. However, the current approach can break in some
cases.
In particular, ACPI can enumerate devices behind PCI bridges that
don't show up in a normal PCI scan, but those device_t objects can end
up as direct children of acpi0. Reserving resources for those devices
directly from acpi0 ends up conflicting with later attempts to reserve
the PCI bridge windows.
As a workaround, defer reserving unclaimed resources until after the
initial probe and attach scan. Eventually this pass of reserving
unclaimed resources can be moved earlier, but it requires changes to
other drivers in the tree to permit enumerating devices and reserving
firmware-assigned resources in a depth-first traversal before
attaching devices whose drivers request wildcard allocations.
PR: 272507
Reported by: Justin Tocci <justin@tocci.org>
Reported by: john@feith.com, many others
Tested by: Oleg Sidorkin <osidorkin@gmail.com>, dch
(cherry picked from commit f2fcb68074)
While we only support 4-byte registers in the uart code the physical
access may be to an 8-byte register. Support this as an option on
non-i386. On i386 we lack the needed 8-byte bus_space functions.
ACPI has an option for 8-byte register io width, and FDT can be given
any size. Support these sizes, even if we don't expect to see hardware
with an 8-byte io width.
Reviewed by: imp
Sponsored by: Arm Ltd
Differential Revision: https://reviews.freebsd.org/D43374
(cherry picked from commit a9fc9d6d15)
Tell bus_dma(9) which NUMA domain the PCI driver is closest to so it
can allocate memory from there when possible.
Reviewed by: markj, jhb
Sponsored by: Arm Ltd
Differential Revision: https://reviews.freebsd.org/D42186
(cherry picked from commit 7098f3c756)
Some platforms require an adjustment of the ethernet hearders. Rather
than make this be on __NO_STRICT_ALIGNMENT being defined, define
VTNET_ETHER_ALIGN to be either 0 or ETHER_ALIGN (aka 2). Add a test to
the if statements to only do them when != 0. This eliminates the #ifdef
sprinkled in the code, still communicates the intent and gives the same
compiled results.
Sponsored by: Netflix
Reviewed by: bz, bryanv
Differential Revision: https://reviews.freebsd.org/D43654
(cherry picked from commit 0ea4b40848)
While we account for the padding in the length of the mbuf we use, we do
not account for it when we 'guess' the size of the mbuf to allocate
based in the MTU of the device. This leads to a situation where we might
fail if the mtu is close to a bucket size (say 2018) such that the added
padding would push us over the edge for a full-sized packet. mtu of 2018
is super rare (2016 and 2020 would both work), but fix it none-the-less.
It's a shame we can't just set VTNET_RX_HEADER_PAD to 2 in this case. The 4
seems hard-coded somewhere I've not found documented (I think it's in the
protocol given the comments about VIRTIO_F_ANY_LAYOUT).
Sponsored by: Netflix
Reviewed by: bz
Differential Revision: https://reviews.freebsd.org/D43656
(cherry picked from commit d9e0e42627)
