This reverts commit e835c0a4e2.
Don't omit soft-reset. During initialization, the driver may need to
perform a soft reset to ensure the phy is ready when the controller
updates the GCTL.PRTCAPDIR or other settings by issuing phy soft-reset.
Many platforms often have access to DCTL register for soft-reset despite
being host-only. If there are actual reported issues from the platforms
that don't expose DCTL registers, then we will need to revisit (perhaps
to teach dwc3 to perform xhci's soft-reset USBCMD.HCRST).
Cc: <stable@vger.kernel.org>
Fixes: e835c0a4e2 ("usb: dwc3: don't reset device side if dwc3 was configured as host-only")
Signed-off-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/7668ab11a48f260820825274976eb41fec7f54d1.1703282469.git.Thinh.Nguyen@synopsys.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Mika writes:
thunderbolt: Changes for v6.8 merge window
This includes following USB4/Thunderbolt changes for the v6.8 merge
window:
- Intel Lunar Lake support
- PCIe tunneling improvements
- DisplayPort tunneling improvements
- Asymmetric switching improvements
- Couple of minor fixes and cleanups.
All these have been in linux-next with no reported issues.
* tag 'thunderbolt-for-v6.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt:
thunderbolt: Reduce retry timeout to speed up boot for some devices
thunderbolt: Keep link as asymmetric if preferred by hardware
thunderbolt: Add support for Intel Lunar Lake
thunderbolt: Disable PCIe extended encapsulation upon teardown properly
thunderbolt: Make PCIe tunnel setup and teardown follow CM guide
thunderbolt: Improve logging when DisplayPort resource is added due to hotplug
thunderbolt: Use tb_dp_read_cap() to read DP_COMMON_CAP as well
thunderbolt: Disable CL states only when actually needed
thunderbolt: Transition link to asymmetric only when both sides support it
thunderbolt: Log XDomain link speed and width
thunderbolt: Move width_name() helper to tb.h
thunderbolt: Handle lane bonding of Gen 4 XDomain links properly
thunderbolt: Unwind TMU configuration if tb_switch_set_tmu_mode_params() fails
thunderbolt: Remove duplicated re-assignment of pointer 'out'
This is a followup to "thunderbolt: Workaround an IOMMU fault on certain
systems with Intel Maple Ridge".
It seems like the timeout can be reduced to 250ms. This reduces the overall
delay caused by the retires to ~1s. This is about the time other things
being initialized in parallel need anyway*, so like this the effective boot
time is no longer compromised.
*I only had a single device available for my measurements: A Clevo X170KM-G
desktop replacement notebook.
Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
In case of the link is brought up as asymmetric (due to hardware preference), we
honor that and don't transition it to symmetric, unless a router with symmetric
link got plugged below, in the topology (and a bandwidth allows transition to
symmetric).
Signed-off-by: Gil Fine <gil.fine@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
During device enumeration usb core resets endpoint 0 if the max packet
size value differs from the one read from the device descriptor.
usb core will additionally reset endpoint 0 during S4 resume, before
re-enumerating the device, if the device has a reset-resume flag set.
In this case the xhci device representation vdev may be lost due to
xHC restore error and re-initialization during S4 resume.
Make sure slot_id and vdev are valid before trying to re-configure max
packet size during endpoint 0 reset.
max packet size will be re-configured later during re-enumeration.
This fixes commit e34900f46c ("xhci: Reconfigure endpoint 0 max packet
size only during endpoint reset") which is currently in usb-next,
on its way to 6.8
Fixes: e34900f46c ("xhci: Reconfigure endpoint 0 max packet size only during endpoint reset")
Tested-by: Wendy Wang <wendy.wang@intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231215125707.1732989-2-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 588b9e8560 ("usb: gadget: uvc: add v4l2 enumeration api calls")
has rendered the precomposed (aka legacy) webcam gadget unloadable.
uvc_alloc() since then has depended on certain config groups being
available in configfs tree related to the UVC function. However, legacy
gadgets do not create anything in configfs, so uvc_alloc() must fail
with -ENOENT no matter what.
This patch mimics the required configfs hierarchy to satisfy the code which
inspects formats and frames found in uvcg_streaming_header.
This has been tested with guvcview on the host side, using vivid as a
source of video stream on the device side and using the userspace program
found at https://gitlab.freedesktop.org/camera/uvc-gadget.git.
Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@collabora.com>
Fixes: 588b9e8560 ("usb: gadget: uvc: add v4l2 enumeration api calls")
Link: https://lore.kernel.org/r/20231215131614.29132-1-andrzej.p@collabora.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When compiling with gcc version 14.0.0 20231206 (experimental)
and CONFIG_FORTIFY_SOURCE=y, I've noticed the following warning:
...
In function 'fortify_memcpy_chk',
inlined from '__ffs_func_bind_do_os_desc' at drivers/usb/gadget/function/f_fs.c:2934:3:
./include/linux/fortify-string.h:588:25: warning: call to '__read_overflow2_field'
declared with attribute warning: detected read beyond size of field (2nd parameter);
maybe use struct_group()? [-Wattribute-warning]
588 | __read_overflow2_field(q_size_field, size);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This call to 'memcpy()' is interpreted as an attempt to copy both
'CompatibleID' and 'SubCompatibleID' of 'struct usb_ext_compat_desc'
from an address of the first one, which causes an overread warning.
Since we actually want to copy both of them at once, use the
convenient 'struct_group()' and 'sizeof_field()' here.
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
Link: https://lore.kernel.org/r/20231214090428.27292-1-dmantipov@yandex.ru
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a general misunderstanding amongst engineers that {v}snprintf()
returns the length of the data *actually* encoded into the destination
array. However, as per the C99 standard {v}snprintf() really returns
the length of the data that *would have been* written if there were
enough space for it. This misunderstanding has led to buffer-overruns
in the past. It's generally considered safer to use the {v}scnprintf()
variants in their place (or even sprintf() in simple cases). So let's
do that.
Link: https://lwn.net/Articles/69419/
Link: https://github.com/KSPP/linux/issues/105
Signed-off-by: Lee Jones <lee@kernel.org>
Link: https://lore.kernel.org/r/20231213164246.1021885-11-lee@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a general misunderstanding amongst engineers that {v}snprintf()
returns the length of the data *actually* encoded into the destination
array. However, as per the C99 standard {v}snprintf() really returns
the length of the data that *would have been* written if there were
enough space for it. This misunderstanding has led to buffer-overruns
in the past. It's generally considered safer to use the {v}scnprintf()
variants in their place (or even sprintf() in simple cases). So let's
do that.
Link: https://lwn.net/Articles/69419/
Link: https://github.com/KSPP/linux/issues/105
Signed-off-by: Lee Jones <lee@kernel.org>
Link: https://lore.kernel.org/r/20231213164246.1021885-10-lee@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a general misunderstanding amongst engineers that {v}snprintf()
returns the length of the data *actually* encoded into the destination
array. However, as per the C99 standard {v}snprintf() really returns
the length of the data that *would have been* written if there were
enough space for it. This misunderstanding has led to buffer-overruns
in the past. It's generally considered safer to use the {v}scnprintf()
variants in their place (or even sprintf() in simple cases). So let's
do that.
Whilst we're at it, let's define some magic numbers to increase
readability and ease of maintenance.
Link: https://lwn.net/Articles/69419/
Link: https://github.com/KSPP/linux/issues/105
Cc: Tomoki Sekiyama <tomoki.sekiyama@gmail.com>
Signed-off-by: Lee Jones <lee@kernel.org>
Link: https://lore.kernel.org/r/20231213164246.1021885-9-lee@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a general misunderstanding amongst engineers that {v}snprintf()
returns the length of the data *actually* encoded into the destination
array. However, as per the C99 standard {v}snprintf() really returns
the length of the data that *would have been* written if there were
enough space for it. This misunderstanding has led to buffer-overruns
in the past. It's generally considered safer to use the {v}scnprintf()
variants in their place (or even sprintf() in simple cases). So let's
do that.
Link: https://lwn.net/Articles/69419/
Link: https://github.com/KSPP/linux/issues/105
Cc: Pawel Laszczak <pawell@cadence.com>
Signed-off-by: Lee Jones <lee@kernel.org>
Link: https://lore.kernel.org/r/20231213164246.1021885-7-lee@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a general misunderstanding amongst engineers that {v}snprintf()
returns the length of the data *actually* encoded into the destination
array. However, as per the C99 standard {v}snprintf() really returns
the length of the data that *would have been* written if there were
enough space for it. This misunderstanding has led to buffer-overruns
in the past. It's generally considered safer to use the {v}scnprintf()
variants in their place (or even sprintf() in simple cases). So let's
do that.
Link: https://lwn.net/Articles/69419/
Link: https://github.com/KSPP/linux/issues/105
Signed-off-by: Lee Jones <lee@kernel.org>
Link: https://lore.kernel.org/r/20231213164246.1021885-2-lee@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Qualcomm WCD9390/WCD9395 is a standalone Hi-Fi audio codec IC with a
functionally separate USB SubSystem for Altmode/Analog Audio Switch
accessible over an I2C interface.
It provides switching USB-C USB2.0 lines between USB and Audio Headphones
speaker lines, and the USB-C SBU lines between DisplayPort AUX and Audio
Headphones Microphone/Ground.
The Audio Headphone and Microphone data path between the Codec and the
USB-C Mux subsystems are external to the IC, thus requiring DT
port-endpoint graph description to handle USB-C altmode & orientation
switching for Audio Accessory Mode.
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://lore.kernel.org/r/20231212-topic-sm8650-upstream-wcd939x-usbss-v2-2-38961fea5867@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It's possible that usb_choose_configuration() can get called when a
USB device has no driver. In this case the recent commit a87b8e3be9
("usb: core: Allow subclassed USB drivers to override
usb_choose_configuration()") can cause a crash since it dereferenced
the driver structure without checking for NULL. Let's add a check.
A USB device with no driver is an anomaly, so make
usb_choose_configuration() return immediately if there is no driver.
This was seen in the real world when usbguard got ahold of a r8152
device at the wrong time. It can also be simulated via this on a
computer with one r8152-based USB Ethernet adapter:
cd /sys/bus/usb/drivers/r8152-cfgselector
to_unbind="$(ls -d *-*)"
real_dir="$(readlink -f "${to_unbind}")"
echo "${to_unbind}" > unbind
cd "${real_dir}"
echo 0 > authorized
echo 1 > authorized
Fixes: a87b8e3be9 ("usb: core: Allow subclassed USB drivers to override usb_choose_configuration()")
Reviewed-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20231211073237.v3.1.If27eb3bf7812f91ab83810f232292f032f4203e0@changeid
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In ACM support for sending breaks to devices is optional.
If a device says that it doenot support sending breaks,
the host must respect that.
Given the number of optional features providing tty operations
for each combination is not practical and errors need to be
returned dynamically if unsupported features are requested.
In case a device does not support break, we want the tty layer
to treat that like it treats drivers that statically cannot
support sending a break. It ignores the inability and does nothing.
This patch uses EOPNOTSUPP to indicate that.
Signed-off-by: Oliver Neukum <oneukum@suse.com>
Fixes: 9e98966c7b ("tty: rework break handling")
Link: https://lore.kernel.org/r/20231207132639.18250-1-oneukum@suse.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Intel Lunar Lake has similar integrated Thunderbolt/USB4 controller as
Intel Meteor Lake with some small differences in the host router (it has
3 DP IN adapters for instance). Add the Intel Lunar Lake PCI IDs to the
driver list of supported devices.
Tested-by: Pengfei Xu <pengfei.xu@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
In case of PCIe tunnel teardown (including if caused by router unplug),
PCIe extended encapsulation bit should be cleared in downstream and
upstream routers accordingly.
Signed-off-by: Gil Fine <gil.fine@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The USB4 Connection Manager guide suggests that the PCIe paths are
enabled from the upstream adapter to the downstream adapter and vice
versa on disable so make the driver follows this sequence.
Signed-off-by: Gil Fine <gil.fine@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
To allow us differentiate how DisplayPort resource is added to the
DisplayPort resources list make the debug log to append "hotplug" when
this was due to an actual hotplug.
Signed-off-by: Gil Fine <gil.fine@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
There is no point doing this separately as the register layout is the
same. For this reason rename tb_dp_read_dprx() to tb_dp_wait_dprx() and
call tb_dp_read_cap() instead.
While there add debug log if the DPRX capability read times out.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
If there is not going to be an actual transition to asymmetric or
symmetric, there is no point to disable and re-enable CL states either.
So instead disable them only when we know that an actual transition is
going to take place.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
