If ucsi_init() fails for some reason (e.g. ucsi_register_port()
fails or general communication failure to the PPM), particularly at
any point after the GET_CAPABILITY command had been issued, this
results in unwinding the initialization and returning an error.
However the ucsi structure's ucsi_capability member retains its
current value, including likely a non-zero num_connectors.
And because ucsi_init() itself is done in a workqueue a UCSI
interface driver will be unaware that it failed and may think the
ucsi_register() call was completely successful. Later, if
ucsi_unregister() is called, due to this stale ucsi->cap value it
would try to access the items in the ucsi->connector array which
might not be in a proper state or not even allocated at all and
results in NULL or invalid pointer dereference.
Fix this by clearing the ucsi->cap value to 0 during the error
path of ucsi_init() in order to prevent a later ucsi_unregister()
from entering the connector cleanup loop.
Fixes: c1b0bc2dab ("usb: typec: Add support for UCSI interface")
Cc: stable@vger.kernel.org
Acked-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Mayank Rana <mrana@codeaurora.org>
Signed-off-by: Jack Pham <jackp@codeaurora.org>
Link: https://lore.kernel.org/r/20210609073535.5094-1-jackp@codeaurora.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The commit a390bef7db ("usb: gadget: fsl_mxc_udc: Remove the driver")
dropped the ARCH_MXC dependency from USB_FSL_USB2, leaving it depending
solely on FSL_SOC.
FSL_SOC is powerpc only; it was briefly available on ARM in 2014 but was
removed by commit cfd074ad86 ("ARM: imx: temporarily remove
CONFIG_SOC_FSL from LS1021A"). Therefore the driver can no longer be
enabled on ARM platforms.
This appears to be a mistake as arm64's ARCH_LAYERSCAPE and arm32
SOC_LS1021A SoCs use this symbol. It's enabled in these defconfigs:
arch/arm/configs/imx_v6_v7_defconfig:CONFIG_USB_FSL_USB2=y
arch/arm/configs/multi_v7_defconfig:CONFIG_USB_FSL_USB2=y
arch/powerpc/configs/mgcoge_defconfig:CONFIG_USB_FSL_USB2=y
arch/powerpc/configs/mpc512x_defconfig:CONFIG_USB_FSL_USB2=y
To fix, expand the dependencies so USB_FSL_USB2 can be enabled on the
ARM platforms, and with COMPILE_TEST.
Fixes: a390bef7db ("usb: gadget: fsl_mxc_udc: Remove the driver")
Signed-off-by: Joel Stanley <joel@jms.id.au>
Link: https://lore.kernel.org/r/20210610034957.93376-1-joel@jms.id.au
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
CP2102N revision A01 (firmware version <= 1.0.4) has a buggy
flow-control implementation that uses the ulXonLimit instead of
ulFlowReplace field of the flow-control settings structure (erratum
CP2102N_E104).
A recent change that set the input software flow-control limits
incidentally broke RTS control for these devices when CRTSCTS is not set
as the new limits would always enable hardware flow control.
Fix this by explicitly disabling flow control for the buggy firmware
versions and only updating the input software flow-control limits when
IXOFF is requested. This makes sure that the terminal settings matches
the default zero ulXonLimit (ulFlowReplace) for these devices.
Link: https://lore.kernel.org/r/20210609161509.9459-1-johan@kernel.org
Reported-by: David Frey <dpfrey@gmail.com>
Reported-by: Alex Villacís Lasso <a_villacis@palosanto.com>
Tested-by: Alex Villacís Lasso <a_villacis@palosanto.com>
Fixes: f61309d9c9 ("USB: serial: cp210x: set IXOFF thresholds")
Cc: stable@vger.kernel.org # 5.12
Signed-off-by: Johan Hovold <johan@kernel.org>
This adds interface between userspace and feedback endpoint to report real
feedback frequency to the Host.
Current implementation adds new userspace interface ALSA mixer control
"Capture Pitch 1000000" (similar to aloop driver's "PCM Rate Shift 100000"
mixer control)
Value in PPM is chosen to have correction value agnostic of the actual HW
rate, which the application is not necessarily dealing with, while still
retaining a good enough precision to allow smooth clock correction on the
playback side, if necessary.
Similar to sound/usb/endpoint.c, a slow down is allowed up to 25%. This
has no impact on the required bandwidth. Speedup correction has an impact
on the bandwidth reserved for the isochronous endpoint. The default
allowed speedup is 500ppm. This seems to be more than enough but, if
necessary, this is configurable through a module parameter. The reserved
bandwidth is rounded up to the next packet size.
Usage of this new control is easy to implement in existing userspace tools
like alsaloop from alsa-utils.
Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20210603220104.1216001-4-jbrunet@baylibre.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Current f_uac2 USB OUT (aka 'capture') synchronization
implements 'ASYNC' scenario which means USB Gadget has
it's own freerunning clock and can update Host about
real clock frequency through feedback endpoint so Host
can align number of samples sent to the USB gadget to
prevent overruns/underruns
In case if Gadget can has no it's internal clock and
can consume audio samples at any rate (for example,
on the Gadget side someone records audio directly to
a file, or audio samples are played through an
external DAC as soon as they arrive), UAC2 spec
suggests 'ADAPTIVE' synchronization type.
Change UAC2 driver to make it configurable through
additional 'c_sync' configfs file.
Default remains 'asynchronous' with possibility to
switch it to 'adaptive'
Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20210603220104.1216001-3-jbrunet@baylibre.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As per USB and UAC2 specs, asynchronous audio sink endpoint
requires explicit synchronization mechanism (Isochronous
Feedback Endpoint)
Implement feedback companion endpoint for ISO OUT endpoint
This patch adds all required infrastructure and USB requests
handling for feedback endpoint. Syncrhonization itself is
still dummy (feedback ep always reports 'nomimal frequency'
e.g. no adjustement is needed). This satisfies hosts that
require feedback endpoint (like Win10) and poll it periodically
Actual synchronization mechanism should be implemented
separately
Signed-off-by: Ruslan Bilovol <ruslan.bilovol@gmail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20210603220104.1216001-2-jbrunet@baylibre.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Every time the hub signals a reset while we (device) are hsotg->connected,
dwc2_hsotg_core_init_disconnected() is called, which in turn calls
dwc2_hs_phy_init().
GUSBCFG.USBTrdTim is cleared upon Core Soft Reset, so if
hsotg->params.phy_utmi_width is 8-bit, the value of GUSBCFG.USBTrdTim (the
default one: 0x5, corresponding to 16-bit) is always different from
hsotg->params.phy_utmi_width, thus dwc2_core_reset() is called every
time (usbcfg != usbcfg_old), which causes 2 issues:
1) The call to dwc2_core_reset() does another reset 300us after the initial
Chirp K of the first reset (which should last at least Tuch = 1ms), and
messes up the High-speed Detection Handshake: both hub and device drive
current into the D+ and D- lines at the same time.
2) GUSBCFG.USBTrdTim is cleared by the second reset, so its value is always
the default one (0x5).
Setting GUSBCFG.USBTrdTim after the potential call to dwc2_core_reset()
fixes both issues. It is now set even when select_phy is false because the
cost of the Core Soft Reset is removed.
Fixes: 1e868545f2 ("usb: dwc2: gadget: Move gadget phy init into core phy init")
Signed-off-by: Clément Lassieur <clement@lassieur.org>
Link: https://lore.kernel.org/r/20210603155921.940651-1-clement@lassieur.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, the dwc3 platform driver does not explicitly ask for
a DMA mask. This makes it fall back to the default 32-bit mask which
breaks the driver on systems that only have RAM starting above the
first 4G like the Apple M1 SoC.
Fix this by calling dma_set_mask_and_coherent with a 64bit mask.
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sven Peter <sven@svenpeter.dev>
Link: https://lore.kernel.org/r/20210607061751.89752-1-sven@svenpeter.dev
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is no validation of the index from dwc3_wIndex_to_dep() and we might
be referring a non-existing ep and trigger a NULL pointer exception. In
certain configurations we might use fewer eps and the index might wrongly
indicate a larger ep index than existing.
By adding this validation from the patch we can actually report a wrong
index back to the caller.
In our usecase we are using a composite device on an older kernel, but
upstream might use this fix also. Unfortunately, I cannot describe the
hardware for others to reproduce the issue as it is a proprietary
implementation.
[ 82.958261] Unable to handle kernel NULL pointer dereference at virtual address 00000000000000a4
[ 82.966891] Mem abort info:
[ 82.969663] ESR = 0x96000006
[ 82.972703] Exception class = DABT (current EL), IL = 32 bits
[ 82.978603] SET = 0, FnV = 0
[ 82.981642] EA = 0, S1PTW = 0
[ 82.984765] Data abort info:
[ 82.987631] ISV = 0, ISS = 0x00000006
[ 82.991449] CM = 0, WnR = 0
[ 82.994409] user pgtable: 4k pages, 39-bit VAs, pgdp = 00000000c6210ccc
[ 83.000999] [00000000000000a4] pgd=0000000053aa5003, pud=0000000053aa5003, pmd=0000000000000000
[ 83.009685] Internal error: Oops: 96000006 [#1] PREEMPT SMP
[ 83.026433] Process irq/62-dwc3 (pid: 303, stack limit = 0x000000003985154c)
[ 83.033470] CPU: 0 PID: 303 Comm: irq/62-dwc3 Not tainted 4.19.124 #1
[ 83.044836] pstate: 60000085 (nZCv daIf -PAN -UAO)
[ 83.049628] pc : dwc3_ep0_handle_feature+0x414/0x43c
[ 83.054558] lr : dwc3_ep0_interrupt+0x3b4/0xc94
...
[ 83.141788] Call trace:
[ 83.144227] dwc3_ep0_handle_feature+0x414/0x43c
[ 83.148823] dwc3_ep0_interrupt+0x3b4/0xc94
[ 83.181546] ---[ end trace aac6b5267d84c32f ]---
Signed-off-by: Marian-Cristian Rotariu <marian.c.rotariu@gmail.com>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210608162650.58426-1-marian.c.rotariu@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
usb_assign_descriptors() is called with 5 parameters,
the last 4 of which are the usb_descriptor_header for:
full-speed (USB1.1 - 12Mbps [including USB1.0 low-speed @ 1.5Mbps),
high-speed (USB2.0 - 480Mbps),
super-speed (USB3.0 - 5Gbps),
super-speed-plus (USB3.1 - 10Gbps).
The differences between full/high/super-speed descriptors are usually
substantial (due to changes in the maximum usb block size from 64 to 512
to 1024 bytes and other differences in the specs), while the difference
between 5 and 10Gbps descriptors may be as little as nothing
(in many cases the same tuning is simply good enough).
However if a gadget driver calls usb_assign_descriptors() with
a NULL descriptor for super-speed-plus and is then used on a max 10gbps
configuration, the kernel will crash with a null pointer dereference,
when a 10gbps capable device port + cable + host port combination shows up.
(This wouldn't happen if the gadget max-speed was set to 5gbps, but
it of course defaults to the maximum, and there's no real reason to
artificially limit it)
The fix is to simply use the 5gbps descriptor as the 10gbps descriptor,
if a 10gbps descriptor wasn't provided.
Obviously this won't fix the problem if the 5gbps descriptor is also
NULL, but such cases can't be so trivially solved (and any such gadgets
are unlikely to be used with USB3 ports any way).
Cc: Felipe Balbi <balbi@kernel.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210609024459.1126080-1-zenczykowski@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The reasoning for this change is that if we already had
a packet pending, then we also already had a pending timer,
and as such there is no need to reschedule it.
This also prevents packets getting delayed 60 ms worst case
under a tiny packet every 290us transmit load, by keeping the
timeout always relative to the first queued up packet.
(300us delay * 16KB max aggregation / 80 byte packet =~ 60 ms)
As such the first packet is now at most delayed by 300us.
Under low transmit load, this will simply result in us sending
a shorter aggregate, as originally intended.
This patch has the benefit of greatly reducing (by ~10 factor
with 1500 byte frames aggregated into 16 kiB) the number of
(potentially pretty costly) updates to the hrtimer.
Cc: Brooke Basile <brookebasile@gmail.com>
Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Cc: Felipe Balbi <balbi@kernel.org>
Cc: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Link: https://lore.kernel.org/r/20210608085438.813960-1-zenczykowski@gmail.com
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Peter writes:
Two bug fixes for cdns3 and cdnsp
* tag 'usb-v5.13-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/peter.chen/usb:
usb: cdnsp: Fix deadlock issue in cdnsp_thread_irq_handler
usb: cdns3: Enable TDL_CHK only for OUT ep
Jonah writes:
USB-serial fixes for 5.13-rc5
Here's a fix for some pipe-direction mismatches in the quatech2 driver,
and a couple of new device ids for ftdi_sio and omninet (and a related
trivial cleanup).
All but the ftdi_sio commit have been in linux-next, and with no
reported issues.
* tag 'usb-serial-5.13-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial:
USB: serial: ftdi_sio: add NovaTech OrionMX product ID
USB: serial: omninet: update driver description
USB: serial: omninet: add device id for Zyxel Omni 56K Plus
USB: serial: quatech2: fix control-request directions
This patch adds a udc_async_callbacks handler to the net2272 UDC
driver, which will prevent a theoretical race during gadget unbinding.
The net2272 driver is sufficiently complicated that I didn't want to
mess around with IRQ settings. Instead, the patch simply adds a new
flag to control async callbacks, and checks the flag before issuing
any of them.
Acked-by: Felipe Balbi <balbi@kernel.org>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20210520202206.GF1216852@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds a udc_async_callbacks handler to the net2280 UDC
driver, which will prevent a theoretical race during gadget unbinding.
The net2280 driver is sufficiently complicated that I didn't want to
mess around with IRQ settings. Instead, the patch simply adds a new
flag to control async callbacks, and checks the flag before issuing
any of them.
Acked-by: Felipe Balbi <balbi@kernel.org>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20210520202200.GE1216852@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds a udc_async_callbacks handler to the dummy-hcd UDC
driver, which will prevent a theoretical race during gadget unbinding.
The implementation is simple, since dummy-hcd already has a flag to
keep track of whether emulated IRQs are enabled. All the handler has
to do is store the enable value in the flag, and avoid setting the
flag prematurely.
Acked-by: Felipe Balbi <balbi@kernel.org>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20210520202152.GD1216852@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Gadget API has a theoretical race when a gadget driver is unbound.
Although the pull-up is turned off before the driver's ->unbind
callback runs, if the USB cable were to be unplugged at just the wrong
moment there would be nothing to prevent the UDC driver from invoking
the ->disconnect callback after the unbind has finished. In theory,
other asynchronous callbacks could also happen during the time before
the UDC driver's udc_stop routine is called, and the gadget driver
would not be prepared to handle any of them.
We need a way to tell UDC drivers to stop issuing asynchronous (that is,
->suspend, ->resume, ->disconnect, ->reset, or ->setup) callbacks at
some point after the pull-up has been turned off and before the
->unbind callback runs. This patch adds a new ->udc_async_callbacks
callback to the usb_gadget_ops structure for precisely this purpose,
and it adds the corresponding support to the UDC core.
Later patches in this series add support for udc_async_callbacks to
several UDC drivers.
Acked-by: Felipe Balbi <balbi@kernel.org>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20210520202144.GC1216852@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
During unbind, ffs_func_eps_disable() will be executed, resulting in
completion callbacks for any pending USB requests. When using AIO,
irrespective of the completion status, io_data work is queued to
io_completion_wq to evaluate and handle the completed requests. Since
work runs asynchronously to the unbind() routine, there can be a
scenario where the work runs after the USB gadget has been fully
removed, resulting in accessing of a resource which has been already
freed. (i.e. usb_ep_free_request() accessing the USB ep structure)
Explicitly drain the io_completion_wq, instead of relying on the
destroy_workqueue() (in ffs_data_put()) to make sure no pending
completion work items are running.
Signed-off-by: Wesley Cheng <wcheng@codeaurora.org>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/1621644261-1236-1-git-send-email-wcheng@codeaurora.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When receiving Alert Message, if it is not unexpected but is
unsupported for some reason, the port should return Not_Supported
Message response.
Also, according to PD3.0 Spec 6.5.2.1.4 Event Flags Field, the
OTP/OVP/OCP flags in the Event Flags field in Status Message no longer
require Get_PPS_Status Message to clear them. Thus remove it when
receiving Status Message with those flags being set.
In addition, add the missing AMS operations for Status Message.
Fixes: 64f7c494a3 ("typec: tcpm: Add support for sink PPS related messages")
Fixes: 0908c5aca3 ("usb: typec: tcpm: AMS and Collision Avoidance")
Signed-off-by: Kyle Tso <kyletso@google.com>
Link: https://lore.kernel.org/r/20210531164928.2368606-1-kyletso@google.com
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When only PHY1 is used (for example on Odroid-HC4), the regmap init code
uses the usb2 ports when doesn't initialize the PHY1 regmap entry.
This fixes:
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000020
...
pc : regmap_update_bits_base+0x40/0xa0
lr : dwc3_meson_g12a_usb2_init_phy+0x4c/0xf8
...
Call trace:
regmap_update_bits_base+0x40/0xa0
dwc3_meson_g12a_usb2_init_phy+0x4c/0xf8
dwc3_meson_g12a_usb2_init+0x7c/0xc8
dwc3_meson_g12a_usb_init+0x28/0x48
dwc3_meson_g12a_probe+0x298/0x540
platform_probe+0x70/0xe0
really_probe+0xf0/0x4d8
driver_probe_device+0xfc/0x168
...
Fixes: 013af227f5 ("usb: dwc3: meson-g12a: handle the phy and glue registers separately")
Reviewed-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210601084830.260196-1-narmstrong@baylibre.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tcpm_ams_start is used to initiate an AMS as well as checking Collision
Avoidance conditions but not for flagging passive AMS (initiated by the
port partner). Fix the misuses of tcpm_ams_start in tcpm_pd_svdm.
ATTENTION doesn't need responses so the AMS flag is not needed here.
Fixes: 0bc3ee9288 ("usb: typec: tcpm: Properly interrupt VDM AMS")
Signed-off-by: Kyle Tso <kyletso@google.com>
Link: https://lore.kernel.org/r/20210601123151.3441914-5-kyletso@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In USB PD Spec Rev 3.1 Ver 1.0, section "6.12.5 Applicability of
Structured VDM Commands", DFP is allowed and recommended to respond to
Discovery Identity with ACK. And in section "6.4.4.2.5.1 Commands other
than Attention", NAK should be returned only when receiving Messages
with invalid fields, Messages in wrong situation, or unrecognize
Messages.
Still keep the original design for SVDM Version 1.0 for backward
compatibilities.
Fixes: 193a68011f ("staging: typec: tcpm: Respond to Discover Identity commands")
Acked-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Kyle Tso <kyletso@google.com>
Link: https://lore.kernel.org/r/20210601123151.3441914-2-kyletso@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The DWC3 DebugFS directory and files are currently created once
during probe. This includes creation of subdirectories for each
of the gadget's endpoints. This works fine for peripheral-only
controllers, as dwc3_core_init_mode() calls dwc3_gadget_init()
just prior to calling dwc3_debugfs_init().
However, for dual-role controllers, dwc3_core_init_mode() will
instead call dwc3_drd_init() which is problematic in a few ways.
First, the initial state must be determined, then dwc3_set_mode()
will have to schedule drd_work and by then dwc3_debugfs_init()
could have already been invoked. Even if the initial mode is
peripheral, dwc3_gadget_init() happens after the DebugFS files
are created, and worse so if the initial state is host and the
controller switches to peripheral much later. And secondly,
even if the gadget endpoints' debug entries were successfully
created, if the controller exits peripheral mode, its dwc3_eps
are freed so the debug files would now hold stale references.
So it is best if the DebugFS endpoint entries are created and
removed dynamically at the same time the underlying dwc3_eps are.
Do this by calling dwc3_debugfs_create_endpoint_dir() as each
endpoint is created, and conversely remove the DebugFS entry when
the endpoint is freed.
Fixes: 41ce1456e1 ("usb: dwc3: core: make dwc3_set_mode() work properly")
Cc: stable <stable@vger.kernel.org>
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Signed-off-by: Jack Pham <jackp@codeaurora.org>
Link: https://lore.kernel.org/r/20210529192932.22912-1-jackp@codeaurora.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
