Fix 'cast to restricted' sparse warnings reported by kernel test robot
in https://lore.kernel.org/all/202211021607.ssjymlKi-lkp@intel.com/
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
When writing into a slow device like an EEPROM chip, the
controller may exit the busy state before the device releases
the bus. In this case, the ft260_xfer_status returns success
before the data transfer completion.
The patch fixes it by returning from the ft260_xfer_status()
with the "-EAGAIN" on both controller and bus busy status when
appropriate.
It does not apply to the i2c combined transactions when after
the write IO, the controller keeps the bus busy until the read
IO and then between reading IOs to ensure an atomic operation.
Co-developed-by: Germain Hebert <germain.hebert@ca.abb.com>
Signed-off-by: Germain Hebert <germain.hebert@ca.abb.com>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The zero-length passed into the ft260_i2c_write() triggered the
NULL pointer dereference in the debug message on data[0] access.
Since the controller does not support a write of zero length,
let's not allow it.
Before:
$ sudo i2ctransfer -y 13 w0@0x51
Killed
After:
$ sudo i2ctransfer -y 13 w0@0x51
Error: Sending messages failed: Invalid argument
Reported-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The FT260 can enter a power saving mode after being idle for longer
than 5 seconds.
When being woken up from power saving mode by an I2C write request,
a possible NACK is not correctly reported by the controller. As a
workaround, the driver will issue an I2C status report two times in
ft260_xfer_status() after the chip has been idle for more than 5s.
Co-developed-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The FT260 controller does not return NACK when performing a big
read (of multiple hid reports size) from a non-existing device
or from the device responding with NACK when it is not ready
to serve the request. However, it responds correctly with NACK
to a read of up to a single hid report size.
To overcome this issue, we split the muli-report read request
into a read of a single HID report of 60 bytes size and a
multi-report read.
Big read of 256 bytes with first read of 60 bytes:
$ sudo ./i2cperf -d 2 -o 2 -s 256 -r 0-0xff 1 0x50 -S
[ +5.633280] ft260_i2c_write_read: off 0x0 rlen 255 wlen 2
[ +0.000006] ft260_i2c_write: rep 0xd0 addr 0x50 off 0 len 2 wlen 2 flag 0x2 d[0] 0x0
[ +0.013205] ft260_xfer_status: bus_status 0x20, clock 100
[ +0.000007] ft260_i2c_read: rep 0xc2 addr 0x50 len 255 rlen 60 flag 0x3
[ +0.010932] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.004733] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000006] ft260_i2c_read: rep 0xc2 addr 0x50 len 195 rlen 128 flag 0x0
[ +0.012572] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.005789] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.003189] ft260_raw_event: i2c resp: rep 0xd1 len 8
[ +0.004092] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000010] ft260_i2c_read: rep 0xc2 addr 0x50 len 67 rlen 67 flag 0x4
[ +0.011688] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.004700] ft260_raw_event: i2c resp: rep 0xd1 len 7
[ +0.004858] ft260_xfer_status: bus_status 0x20, clock 100
Read from non-existing device at address 8. The first 60 read responded
with NACK.
$ sudo ./i2cperf -d 2 -o 2 -s 256 -r 0-0xff 1 0x8 -S
[Oct19 15:37] ft260_i2c_write_read: off 0x0 rlen 255 wlen 2
[ +0.000007] ft260_i2c_write: rep 0xd0 addr 0x8 off 0 len 2 wlen 2 flag 0x2 d[0] 0x0
[ +0.022820] ft260_xfer_status: bus_status 0x20, clock 100
[ +0.000007] ft260_i2c_read: rep 0xc2 addr 0x8 len 255 rlen 60 flag 0x3
[ +0.010658] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.005965] ft260_xfer_status: bus_status 0x46, clock 100 <-- NACK
[ +0.000009] ft260 0003:0403:6030.0004: i2c bus error: 0x46
[ +0.007784] ft260_i2c_reset: done
Co-developed-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The i2cdetect uses the SMBus Quick command by default to scan devices
on the I2C bus. The FT260 implements an I2C bus controller. The SMBus
is derived from I2C, but there are several differences between the
specifications of the two buses in the areas of timing, protocols,
operation modes, and electrical characteristics.
One of the differences is that the I2C devices allow the slave not
to ACK its slave address, but SMBus requires it to always ACK it as
a mechanism to detect a detachable device’s presence on the bus.
Since FT260 is the I2C bus controller, it does not acknowledge the
SMBus Quick write command, which sends a single bit to the device at
the place of the RD/WR bit.
The ft260 driver attempted to mimic the SMBus Quick Write functionality
by writing a single byte as the SMBus Byte Write command does.
Usually, one byte in the SMBus Quick Write will be fine. However, it may
cause problems with devices with a control register at offset 0, like
i2c muxes, for example, when scanned with the i2cdetect utility.
The i2cdetect with the "-r" option uses the SMBus Read Byte command,
which is a reasonable workaround. To prevent the I2C bus from locking
at write-only devices (most notably clock chips at address 0x69), use
the "-r" option in conjunction with scanning range parameters.
This patch removes the SMBus Quick command support.
$ sudo i2cdetect -y 13
Warning: Can't use SMBus Quick Write command, will skip some addresses
0 1 2 3 4 5 6 7 8 9 a b c d e f
00:
10:
20:
30: -- -- -- -- -- -- -- --
40:
50: 50 51 -- -- -- -- -- -- -- -- -- -- -- -- -- --
60:
70:
$ sudo i2cdetect -y -r 13
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
50: 50 51 -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
70: -- -- -- -- -- -- -- --
Reported-by: Vince Asbridge <VAsbridge@sanblaze.com>
Reported-by: Stephen Shirron <SShirron@sanblaze.com>
Reported-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The FT260 is not supposed to generate unexpected HID reports. However,
in theory, the unsolicited HID Input reports can be issued by a specially
crafted malicious USB device masquerading as FT260 when the attacker has
physical access to the USB port. In this case, the read_buf pointer points
to the final data portion of the previous I2C Read transfer, and the memcpy
invoked in the ft260_raw_event() will try copying the content of the
unexpected report into the wrong location.
This commit sets the Read buffer pointer to NULL on the I2C Read
transaction completion and checks it in the ft260_raw_event() to detect
and skip the unsolicited Input report.
Reported-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Do not populate the /dev/hidraw on ft260 interfaces when the hid-ft260
driver is loaded.
$ sudo insmod hid-ft260.ko
$ ls /dev/hidraw*
/dev/hidraw0
$ sudo rmmod hid-ft260.ko
$ ls /dev/hidraw*
/dev/hidraw0 /dev/hidraw1 /dev/hidraw2
Reported-by: Enrik Berkhan <Enrik.Berkhan@inka.de>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
A random i2c read operation in EEPROM devices is implemented as a dummy
write operation, followed by a current address read operation. The dummy
write operation is used to load the target byte or word address (a.k.a
offset) into the offset counter, from which the subsequent read operation
then reads.
To support longer than one HID report size random read, the ft260 driver
issues multiple pairs of i2c write offset + read data transactions of HID
report size so that the EEPROM device sees many i2c random read requests
from different offsets.
Two issues with the current implementation:
- This approach suffers from extra overhead caused by writing offset
requests.
- Necessity to handle offset per HID report in big-endian representation
as EEPROM devices expect. The current implementation does not do it and
correctly handles the reads up to 60 bytes only.
This patch addresses both issues by implementing more efficient approach.
It issues a single i2c read request of up to the EEPROM page size and then
waits for the data to arrive in multiple HID reports. For example, to read
the 256 bytes from a 24LC512 chip, which has 128 bytes page size, the old
method performs six ft260_i2c_write_read transactions while the new - two
only.
Before:
$ sudo ./i2cperf -d 2 -o 2 -s 128 -r 0-0xff 13 0x51 -S
Read block via i2ctransfer by chunks
-------------------------------------------------------------------
data rate(bps) efficiency(%) data size(B) total IOs IO size(B)
-------------------------------------------------------------------
40803 85 256 2 128
Kernel log of a single 128 bytes read request:
[ +2.376308] ft260_i2c_write_read: read_off 0x0 left_len 128 len 60
[ +0.000002] ft260_i2c_write: rep 0xd0 addr 0x51 off 0 len 2 wlen 2 flag 0x2 d[0] 0x0
[ +0.000707] ft260_xfer_status: bus_status 0x41, clock 100
[ +0.000173] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000001] ft260_i2c_read: rep 0xc2 addr 0x51 len 60
[ +0.008660] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.000156] ft260_xfer_status: bus_status 0x20, clock 100
[ +0.000001] ft260_i2c_write_read: read_off 0x3c left_len 68 len 60
[ +0.000001] ft260_i2c_write: rep 0xd0 addr 0x51 off 0 len 2 wlen 2 flag 0x2 d[0] 0x3c
[ +0.001034] ft260_xfer_status: bus_status 0x41, clock 100
[ +0.000191] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000001] ft260_i2c_read: rep 0xc2 addr 0x51 len 60
[ +0.008614] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.000203] ft260_xfer_status: bus_status 0x20, clock 100
[ +0.000001] ft260_i2c_write_read: read_off 0x78 left_len 8 len 8
[ +0.000001] ft260_i2c_write: rep 0xd0 addr 0x51 off 0 len 2 wlen 2 flag 0x2 d[0] 0x78
[ +0.000987] ft260_xfer_status: bus_status 0x41, clock 100
[ +0.000192] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000001] ft260_i2c_read: rep 0xc2 addr 0x51 len 8
[ +0.002614] ft260_raw_event: i2c resp: rep 0xd1 len 8
[ +0.000200] ft260_xfer_status: bus_status 0x20, clock 100
After:
$ sudo ./i2cperf -d 2 -o 2 -s 128 -r 0-0xff 13 0x51 -S
Read block via i2ctransfer by chunks
-------------------------------------------------------------------
data rate(bps) efficiency(%) data size(B) total IOs IO size(B)
-------------------------------------------------------------------
43990 85 256 2 128
Kernel log of a single 128 bytes read request:
[ +1.464346] ft260_i2c_write_read: off 0x0 rlen 128 wlen 2
[ +0.000002] ft260_i2c_write: rep 0xd0 addr 0x51 off 0 len 2 wlen 2 flag 0x2 d[0] 0x0
[ +0.001653] ft260_xfer_status: bus_status 0x41, clock 100
[ +0.000188] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000002] ft260_i2c_read: rep 0xc2 addr 0x51 len 128 rlen 60 flag 0x3
[ +0.008609] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.000157] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000002] ft260_i2c_read: rep 0xc2 addr 0x51 len 68 rlen 60 flag 0x0
[ +0.008840] ft260_raw_event: i2c resp: rep 0xde len 60
[ +0.000203] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000002] ft260_i2c_read: rep 0xc2 addr 0x51 len 8 rlen 8 flag 0x4
[ +0.002794] ft260_raw_event: i2c resp: rep 0xd1 len 8
[ +0.000201] ft260_xfer_status: bus_status 0x20, clock 100
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Tested-by: Guillaume Champagne <champagne.guillaume.c@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
To support longer than one HID report size write, the driver splits a
single i2c message data payload into multiple i2c messages of HID report
size. However, it does not replicate the offset bytes within the EEPROM
chip in every consequent HID report because it is not and should not be
aware of the EEPROM type. It breaks the i2c write message integrity and
causes the EEPROM device not to acknowledge the second HID report keeping
the i2c bus busy until the ft260 controller reports failure.
This patch preserves the i2c write message integrity by manipulating the
i2c flag bits across multiple HID reports to be seen by the EEPROM device
as a single i2c write transfer.
Before:
$ sudo ./i2cperf -f 2 -o 2 -s 64 -r 0-0xff 13 0x51 -S
Error: Sending messages failed: Input/output error
[ +3.667741] ft260_i2c_write: rep 0xde addr 0x51 off 0 len 60 d[0] 0x0
[ +0.007330] ft260_hid_output_report_check_status: wait 6400 usec, len 64
[ +0.000203] ft260_xfer_status: bus_status 0x40, clock 100
[ +0.000001] ft260_i2c_write: rep 0xd1 addr 0x51 off 60 len 6 d[0] 0x0
[ +0.002337] ft260_hid_output_report_check_status: wait 1000 usec, len 10
[ +0.000157] ft260_xfer_status: bus_status 0x2e, clock 100
[ +0.000241] ft260_i2c_reset: done
[ +0.000003] ft260_i2c_write: failed to start transfer, ret -5
After:
$ sudo ./i2cperf -f 2 -o 2 -s 128 -r 0-0xff 13 0x51 -S
Fill block with increment via i2ctransfer by chunks
-------------------------------------------------------------------
data rate(bps) efficiency(%) data size(B) total IOs IO size(B)
-------------------------------------------------------------------
71260 86 256 2 128
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Tested-by: Guillaume Champagne <champagne.guillaume.c@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
After clarifying with FTDI's support, it turned out that the error
condition (bit 1) in byte 1 of the i2c status HID report is a status
bit reflecting all error conditions. When bits 2, 3, or 4 are raised
to 1, bit 1 is set to 1 also. Since the ft260_xfer_status routine tests
the error condition bit and exits in the case of an error, the program
flow never reaches the conditional expressions for 2, 3, and 4 bits when
any of them indicates an error state. Though these expressions are never
evaluated to true, they are checked several times per IO, increasing the
ft260_xfer_status polling cycle duration.
The patch removes the conditional expressions for 2, 3, and 4 bits in
byte 1 of the i2c status HID report.
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Tested-by: Guillaume Champagne <champagne.guillaume.c@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This patch adds support for the DualShock4 dongle in a very similar
way we contributed to hid-sony before.
The dongle is a USB to Bluetooth bridge and uses the same HID reports
as a USB device. It reports data through the DS4's main USB input
report independent on whether a Bluetooth controller is connected.
For this reason there is custom dongle report parsing code to
detect controller hotplug and kick of calibration work until we
are ready to process actual input reports.
The logic also incorporates a workaround needed for Steam in which
hid-playstation and Steam using hidraw can fight.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The poll interval for DualShock4 in Bluetooth mode is adjustable
through the main output report. Configure it to 4ms, which is
similar to USB.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Add support for DualShock4 in Bluetooth mode. In Bluetooth, the device
is a bit strange in that after 'calibration' it switches sending all its
input data from a basic report (only containing buttons/sticks) to an
extended report, which also contains touchpad, motion sensors and other
data. The overall design of this code is similar to the DualSense code.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This patch adds a parameter to ps_get_report to ignore CRC checks.
This prepares for DualShock4, which has some HID reports, which lack CRC.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Expose the lightbar LEDs in the same manner as hid-sony through
individual LEDs for backwards compatibility reasons. There is a
slight change in LED naming to use the input device name as opposed
to the MAC address like hid-sony did. This is expected to not
cause any issues and should make the naming more compliant.
In addition set a default lightbar color based on player ID.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Make the max_brightness adjustable through ps_led_info struct. This
paves the way for a next DualShock4 patch to allow larger brightness
values.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This patch implements DualShock4 rumble support in a similar manner
as the DualSense implementation. It adds an output worker with
granular control of different features of the main DualShock4 output
report.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Support accelerometer and gyroscope as separate input devices similar
how DualSense and hid-sony do it.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Support the DualShock4 touchpad as a separate input device. The code
describes the touchpad input reports through structures similar a bit
to the DualSense code.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Provide DualShock4 battery support through powersupply framework.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Report DualShock4 hardware and firmware version info through sysfs.
It uses the same sysfs nodes as the DualSense did (and hid-sony).
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Add basic support for DualShock4 USB controller with buttons and sticks.
Signed-off-by: Roderick Colenbrander <roderick.colenbrander@sony.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This reverts commit 9866d141a0.
The csky rseq support has been merged without ever notifying the rseq
maintainers, and without any of the required asssembler glue in the rseq
selftests, which means it is entirely untested.
It is also derived from a non-upstream riscv patch which has known bugs.
The assembly part of this revert should be carefully reviewed by the
architecture maintainer because it touches code which has changed since
the merge of the reverted patch.
The rseq selftests assembly glue should be introduced at the same time
as the architecture rseq support. Without the presence of any test, I
recommend reverting rseq support from csky for now.
Link: https://lore.kernel.org/lkml/1257037909.25426.1626705790861.JavaMail.zimbra@efficios.com/
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Guo Ren <guoren@kernel.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-csky@vger.kernel.org
This reverts commit f36e0aab6f.
The csky rseq support has been merged without ever notifying the rseq
maintainers, and without any of the required asssembler glue in the rseq
selftests, which means it is entirely untested.
It is also derived from a non-upstream riscv patch which has known bugs.
The assembly part of this revert should be carefully reviewed by the
architecture maintainer because it touches code which has changed since
the merge of the reverted patch.
The rseq selftests assembly glue should be introduced at the same time
as the architecture rseq support. Without the presence of any test, I
recommend reverting rseq support from csky for now.
Link: https://lore.kernel.org/lkml/1257037909.25426.1626705790861.JavaMail.zimbra@efficios.com/
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Guo Ren <guoren@kernel.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-csky@vger.kernel.org
Add a RDMA VF driver for Microsoft Azure Network Adapter (MANA).
Co-developed-by: Ajay Sharma <sharmaajay@microsoft.com>
Signed-off-by: Ajay Sharma <sharmaajay@microsoft.com>
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-13-git-send-email-longli@linuxonhyperv.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Long Li says:
====================
Introduce Microsoft Azure Network Adapter (MANA) RDMA driver [netdev prep]
The first 11 patches which modify the MANA Ethernet driver to support
RDMA driver.
* 'mana-shared-6.2' of https://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
net: mana: Define data structures for protection domain and memory registration
net: mana: Define data structures for allocating doorbell page from GDMA
net: mana: Define and process GDMA response code GDMA_STATUS_MORE_ENTRIES
net: mana: Define max values for SGL entries
net: mana: Move header files to a common location
net: mana: Record port number in netdev
net: mana: Export Work Queue functions for use by RDMA driver
net: mana: Set the DMA device max segment size
net: mana: Handle vport sharing between devices
net: mana: Record the physical address for doorbell page region
net: mana: Add support for auxiliary device
====================
Link: https://lore.kernel.org/all/1667502990-2559-1-git-send-email-longli@linuxonhyperv.com/
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The driver for XillyUSB devices maintains a kref reference count on each
xillyusb_dev structure, which represents a physical device. This reference
count reaches zero when the device has been disconnected and there are no
open file descriptors that are related to the device. When this occurs,
kref_put() calls cleanup_dev(), which clears up the device's data,
including the structure itself.
However, when xillyusb_open() is called, this reference count becomes
tricky: This function needs to obtain the xillyusb_dev structure that
relates to the inode's major and minor (as there can be several such).
xillybus_find_inode() (which is defined in xillybus_class.c) is called
for this purpose. xillybus_find_inode() holds a mutex that is global in
xillybus_class.c to protect the list of devices, and releases this
mutex before returning. As a result, nothing protects the xillyusb_dev's
reference counter from being decremented to zero before xillyusb_open()
increments it on its own behalf. Hence the structure can be freed
due to a rare race condition.
To solve this, a mutex is added. It is locked by xillyusb_open() before
the call to xillybus_find_inode() and is released only after the kref
counter has been incremented on behalf of the newly opened inode. This
protects the kref reference counters of all xillyusb_dev structs from
being decremented by xillyusb_disconnect() during this time segment, as
the call to kref_put() in this function is done with the same lock held.
There is no need to hold the lock on other calls to kref_put(), because
if xillybus_find_inode() finds a struct, xillyusb_disconnect() has not
made the call to remove it, and hence not made its call to kref_put(),
which takes place afterwards. Hence preventing xillyusb_disconnect's
call to kref_put() is enough to ensure that the reference doesn't reach
zero before it's incremented by xillyusb_open().
It would have been more natural to increment the reference count in
xillybus_find_inode() of course, however this function is also called by
Xillybus' driver for PCIe / OF, which registers a completely different
structure. Therefore, xillybus_find_inode() treats these structures as
void pointers, and accordingly can't make any changes.
Reported-by: Hyunwoo Kim <imv4bel@gmail.com>
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Eli Billauer <eli.billauer@gmail.com>
Link: https://lore.kernel.org/r/20221030094209.65916-1-eli.billauer@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix v_recv_cmd_submit() to use PIPE_BULK define instead of hard coded
values. This also fixes the following signed integer overflow error
reported by cppcheck. This is not an issue since pipe is unsigned int.
However, this change improves the code to use proper define.
drivers/usb/usbip/vudc_rx.c:152:26: error: Signed integer overflow for expression '3<<30'. [integerOverflow]
urb_p->urb->pipe &= ~(3 << 30);
In addition, add a build time check for PIPE_BULK != 3 as the code path
depends on PIPE_BULK = 3.
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Link: https://lore.kernel.org/r/20221110194738.38514-1-skhan@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
NO_IRQ is a relic from the old days. It is not used anymore in core
functions. By the way, function irq_of_parse_and_map() returns value 0
on error.
In some drivers, NO_IRQ is erroneously used to check the return of
irq_of_parse_and_map().
It is not a real bug today because the only architectures using the
drivers being fixed by this patch define NO_IRQ as 0, but there are
architectures which define NO_IRQ as -1. If one day those
architectures start using the non fixed drivers, there will be a
problem.
Long time ago Linus advocated for not using NO_IRQ, see
https://lkml.org/lkml/2005/11/21/221 . He re-iterated the same view
recently in https://lkml.org/lkml/2022/10/12/622
So test !irq instead of tesing irq == NO_IRQ.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/13feefdf6b240817944e6441e26a8ddc1d81ced1.1668102802.git.christophe.leroy@csgroup.eu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code in the FOTG210 driver isn't entirely endianness-agnostic
as reported by the kernel robot sparse testing. This came to
the surface while moving the files around.
The driver is only used on little-endian systems, so this causes
no real-world regression, but it is nice to be strict and have
some compile coverage also on big endian machines, so fix it
up with the right LE accessors.
Fixes: b84a8dee23 ("usb: gadget: add Faraday fotg210_udc driver")
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/linux-usb/202211110910.0dJ7nZCn-lkp@intel.com/
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20221111090317.94228-1-linus.walleij@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Cortina Systems Gemini has bolted on a PHY inside the
silicon that can be handled by six bits in a MISC register in
the system controller.
If we are running on Gemini, look up a syscon regmap through
a phandle and enable VBUS and optionally the Mini-B connector.
If the device is flagged as "wakeup-source" using the standard
DT bindings, we also enable this in the global controller for
respective port.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20221109200554.1957185-1-linus.walleij@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This arrange device tree nodes in alphabetical order.
Signed-off-by: Sam Shih <sam.shih@mediatek.com>
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20221106085034.12582-2-linux@fw-web.de
Signed-off-by: Matthias Brugger <matthias.bgg@gmail.com>
Enable missing configs in the arm64 defconfig to get all devices probing
on the mt8183-kukui-jacuzzi-juniper machine.
The devices enabled are: ATH10K SDIO wireless adapter, Elan touchscreen,
cr50 TPM, MediaTek SPI controller, JPEG video decoder, ANX7625 DSI/DPI
to DP bridge (used for the internal display), MT8183 sound cards, SCP
co-processor, MediaTek Global Command Engine (controlled by CMDQ
driver), MediaTek Smart Voltage Scaling (SVS) engine, CCI frequency
and voltage scaling, AUXADC thermal sensors.
All symbols are enabled as modules with the exception of SPI, which is
enabled as builtin since on some platforms like mt8195-cherry, the
ChromeOS Embedded Controller is connected through SPI and it is
responsible for the regulators powering the MMC controller used for the
SD card, and thus SPI support is required for booting.
By enabling the support for all of this machine's devices on the
defconfig we make it effortless to test the relevant hardware both by
developers as well as CI systems like KernelCI.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20221109195012.1231059-1-nfraprado@collabora.com
Signed-off-by: Matthias Brugger <matthias.bgg@gmail.com>
In ata_tdev_add(), the return value of transport_add_device() is
not checked. As a result, it causes null-ptr-deref while removing
the module, because transport_remove_device() is called to remove
the device that was not added.
Unable to handle kernel NULL pointer dereference at virtual address 00000000000000d0
CPU: 13 PID: 13603 Comm: rmmod Kdump: loaded Tainted: G W 6.1.0-rc3+ #36
pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : device_del+0x48/0x3a0
lr : device_del+0x44/0x3a0
Call trace:
device_del+0x48/0x3a0
attribute_container_class_device_del+0x28/0x40
transport_remove_classdev+0x60/0x7c
attribute_container_device_trigger+0x118/0x120
transport_remove_device+0x20/0x30
ata_tdev_delete+0x24/0x50 [libata]
ata_tlink_delete+0x40/0xa0 [libata]
ata_tport_delete+0x2c/0x60 [libata]
ata_port_detach+0x148/0x1b0 [libata]
ata_pci_remove_one+0x50/0x80 [libata]
ahci_remove_one+0x4c/0x8c [ahci]
Fix this by checking and handling return value of transport_add_device()
in ata_tdev_add(). In the error path, device_del() is called to delete
the device which was added earlier in this function, and ata_tdev_free()
is called to free ata_dev.
Fixes: d9027470b8 ("[libata] Add ATA transport class")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
In ata_tlink_add(), the return value of transport_add_device() is
not checked. As a result, it causes null-ptr-deref while removing
the module, because transport_remove_device() is called to remove
the device that was not added.
Unable to handle kernel NULL pointer dereference at virtual address 00000000000000d0
CPU: 33 PID: 13850 Comm: rmmod Kdump: loaded Tainted: G W 6.1.0-rc3+ #12
pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : device_del+0x48/0x39c
lr : device_del+0x44/0x39c
Call trace:
device_del+0x48/0x39c
attribute_container_class_device_del+0x28/0x40
transport_remove_classdev+0x60/0x7c
attribute_container_device_trigger+0x118/0x120
transport_remove_device+0x20/0x30
ata_tlink_delete+0x88/0xb0 [libata]
ata_tport_delete+0x2c/0x60 [libata]
ata_port_detach+0x148/0x1b0 [libata]
ata_pci_remove_one+0x50/0x80 [libata]
ahci_remove_one+0x4c/0x8c [ahci]
Fix this by checking and handling return value of transport_add_device()
in ata_tlink_add().
Fixes: d9027470b8 ("[libata] Add ATA transport class")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
In ata_tport_add(), the return value of transport_add_device() is
not checked. As a result, it causes null-ptr-deref while removing
the module, because transport_remove_device() is called to remove
the device that was not added.
Unable to handle kernel NULL pointer dereference at virtual address 00000000000000d0
CPU: 12 PID: 13605 Comm: rmmod Kdump: loaded Tainted: G W 6.1.0-rc3+ #8
pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : device_del+0x48/0x39c
lr : device_del+0x44/0x39c
Call trace:
device_del+0x48/0x39c
attribute_container_class_device_del+0x28/0x40
transport_remove_classdev+0x60/0x7c
attribute_container_device_trigger+0x118/0x120
transport_remove_device+0x20/0x30
ata_tport_delete+0x34/0x60 [libata]
ata_port_detach+0x148/0x1b0 [libata]
ata_pci_remove_one+0x50/0x80 [libata]
ahci_remove_one+0x4c/0x8c [ahci]
Fix this by checking and handling return value of transport_add_device()
in ata_tport_add().
Fixes: d9027470b8 ("[libata] Add ATA transport class")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
In the error path in ata_tport_add(), when calling put_device(),
ata_tport_release() is called, it will put the refcount of 'ap->host'.
And then ata_host_put() is called again, the refcount is decreased
to 0, ata_host_release() is called, all ports are freed and set to
null.
When unbinding the device after failure, ata_host_stop() is called
to release the resources, it leads a null-ptr-deref(), because all
the ports all freed and null.
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008
CPU: 7 PID: 18671 Comm: modprobe Kdump: loaded Tainted: G E 6.1.0-rc3+ #8
pstate: 80400009 (Nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : ata_host_stop+0x3c/0x84 [libata]
lr : release_nodes+0x64/0xd0
Call trace:
ata_host_stop+0x3c/0x84 [libata]
release_nodes+0x64/0xd0
devres_release_all+0xbc/0x1b0
device_unbind_cleanup+0x20/0x70
really_probe+0x158/0x320
__driver_probe_device+0x84/0x120
driver_probe_device+0x44/0x120
__driver_attach+0xb4/0x220
bus_for_each_dev+0x78/0xdc
driver_attach+0x2c/0x40
bus_add_driver+0x184/0x240
driver_register+0x80/0x13c
__pci_register_driver+0x4c/0x60
ahci_pci_driver_init+0x30/0x1000 [ahci]
Fix this by removing redundant ata_host_put() in the error path.
Fixes: 2623c7a5f2 ("libata: add refcounting to ata_host")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Nobody seems to call ata_sff_busy_sleep(), so we can get rid of it...
Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Lorenzo Bianconi says:
====================
introduce WED RX support to MT7986 SoC
Similar to TX counterpart available on MT7622 and MT7986, introduce
RX Wireless Ethernet Dispatch available on MT7986 SoC in order to
offload traffic received by wlan nic to the wired interfaces (lan/wan).
Changes since v3:
- remove reset property in ethsys dts node
- rely on readx_poll_timeout in wo mcu code
- fix typos
- move wo-ccif binding in soc folder
- use reserved-memory for wo-dlm
- improve wo-ccif binding
Changes since v2:
- rely on of_reserved_mem APIs in mcu code
- add some dts fixes
- rename {tx,rx}_wdma in {rx,tx}_wdma
- update entry in maintainers file
Changes since v1:
- fix sparse warnings
- rely on memory-region property in mt7622-wed.yaml
- some more binding fixes
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Update MEDIATEK ETHERNET driver maintainer file enty adding myself to
maintainers list
Acked-by: Felix Fietkau <nbd@nbd.name>
Acked-by: Sean Wang <sean.wang@mediatek.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce WED RX MIB counters support available on MT7986a SoC.
Tested-by: Daniel Golle <daniel@makrotopia.org>
Co-developed-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Enable RX Wireless Ethernet Dispatch available on MT7986 Soc.
Tested-by: Daniel Golle <daniel@makrotopia.org>
Co-developed-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
