- 13 Dec, 2022 3 commits
-
-
Jiri Kosina authored
- fixes and performance improvements to the hid-ft260 driver (Michael Zaidman)
-
Jiri Kosina authored
- removal of superfluous hid_hw_stop() calls for drivers with default .remove callback (Marcus Folkesson)
-
Jiri Kosina authored
- new quirks for select Apple keyboards (Kerem Karabay, Aditya Garg)
-
- 08 Dec, 2022 2 commits
-
-
Benjamin Tissoires authored
This reverts commit 532223c8. As reported in [0], hid-logitech-hidpp now binds on all bluetooth mice, but there are corner cases where hid-logitech-hidpp just gives up on the mouse. This leads the end user with a dead mouse. Given that we are at -rc8, we are definitively too late to find a proper fix. We already identified 2 issues less than 24 hours after the bug report. One in that ->match() was never designed to be used anywhere else than in hid-generic, and the other that hid-logitech-hidpp has corner cases where it gives up on devices it is not supposed to. So we have no choice but postpone this patch to the next kernel release. [0] https://lore.kernel.org/linux-input/CAJZ5v0g-_o4AqMgNwihCb0jrwrcJZfRrX=jv8aH54WNKO7QB8A@mail.gmail.com/Reported-by:
Rafael J . Wysocki <rjw@rjwysocki.net> Signed-off-by:
Benjamin Tissoires <benjamin.tissoires@redhat.com> Signed-off-by:
Jiri Kosina <jkosina@suse.cz>
-
Benjamin Tissoires authored
This reverts commit 8544c812. We need to revert commit 532223c8 ("HID: logitech-hidpp: Enable HID++ for all the Logitech Bluetooth devices") because that commit might make hid-logitech-hidpp bind on mice that are not well enough supported by hid-logitech-hidpp, and the end result is that the probe of those mice is now returning -ENODEV, leaving the end user with a dead mouse. Given that commit 8544c812 ("HID: logitech-hidpp: Remove special-casing of Bluetooth devices") is a direct dependency of 532223c8, revert it too. Note that this also adapt according to commit 908d325e ("HID: logitech-hidpp: Detect hi-res scrolling support") to re-add support of the devices that were removed from that commit too. I have locally an MX Master and I tested this device with that revert, ensuring we still have high-res scrolling. Reported-by:
-
- 22 Nov, 2022 1 commit
-
-
Ankit Patel authored
Some additional USB mouse devices are needing ALWAYS_POLL quirk without which they disconnect and reconnect every 60s. Add below devices to the known quirk list. CHERRY VID 0x046a, PID 0x000c MICROSOFT VID 0x045e, PID 0x0783 PRIMAX VID 0x0461, PID 0x4e2a Signed-off-by:
Ankit Patel <anpatel@nvidia.com> Signed-off-by:
Haotien Hsu <haotienh@nvidia.com> Signed-off-by:
-
- 21 Nov, 2022 4 commits
-
-
Marcus Folkesson authored
hid_device_remove() will call hid_hw_stop() as default .remove function if no function is specified. Signed-off-by:
Marcus Folkesson <marcus.folkesson@gmail.com> Signed-off-by:
-
Marcus Folkesson authored
hid_device_remove() will call hid_hw_stop() as default .remove function if no function is specified. Signed-off-by:
-
ZhangPeng authored
Syzbot reported shift-out-of-bounds in hid_report_raw_event. microsoft 0003:045E:07DA.0001: hid_field_extract() called with n (128) > 32! (swapper/0) ====================================================================== UBSAN: shift-out-of-bounds in drivers/hid/hid-core.c:1323:20 shift exponent 127 is too large for 32-bit type 'int' CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.1.0-rc4-syzkaller-00159-g4bbf3422 #0 Hardware name: Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022 Call Trace: <IRQ> __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0x1e3/0x2cb lib/dump_stack.c:106 ubsan_epilogue lib/ubsan.c:151 [inline] __ubsan_handle_shift_out_of_bounds+0x3a6/0x420 lib/ubsan.c:322 snto32 drivers/hid/hid-core.c:1323 [inline] hid_input_fetch_field drivers/hid/hid-core.c:1572 [inline] hid_process_report drivers/hid/hid-core.c:1665 [inline] hid_report_raw_event+0xd56/0x18b0 drivers/hid/hid-core.c:1998 hid_input_report+0x408/0x4f0 drivers/hid/hid-core.c:2066 hid_irq_in+0x459/0x690 drivers/hid/usbhid/hid-core.c:284 __usb_hcd_giveback_urb+0x369/0x530 drivers/usb/core/hcd.c:1671 dummy_timer+0x86b/0x3110 drivers/usb/gadget/udc/dummy_hcd.c:1988 call_timer_fn+0xf5/0x210 kernel/time/timer.c:1474 expire_timers kernel/time/timer.c:1519 [inline] __run_timers+0x76a/0x980 kernel/time/timer.c:1790 run_timer_softirq+0x63/0xf0 kernel/time/timer.c:1803 __do_softirq+0x277/0x75b kernel/softirq.c:571 __irq_exit_rcu+0xec/0x170 kernel/softirq.c:650 irq_exit_rcu+0x5/0x20 kernel/softirq.c:662 sysvec_apic_timer_interrupt+0x91/0xb0 arch/x86/kernel/apic/apic.c:1107 ====================================================================== If the size of the integer (unsigned n) is bigger than 32 in snto32(), shift exponent will be too large for 32-bit type 'int', resulting in a shift-out-of-bounds bug. Fix this by adding a check on the size of the integer (unsigned n) in snto32(). To add support for n greater than 32 bits, set n to 32, if n is greater than 32. Reported-by: syzbot+8b1641d2f14732407e23@syzkaller.appspotmail.com Fixes: dde5845a ("[PATCH] Generic HID layer - code split") Signed-off-by:
ZhangPeng <zhangpeng362@huawei.com> Signed-off-by:
-
José Expósito authored
Commit f7d8e387 ("HID: uclogic: Switch to Digitizer usage for styluses") changed the usage used in UCLogic from "Pen" to "Digitizer". However, the IS_INPUT_APPLICATION() macro evaluates to false for HID_DG_DIGITIZER causing issues with the XP-Pen Star G640 tablet. Add the HID_QUIRK_HIDINPUT_FORCE quirk to bypass the IS_INPUT_APPLICATION() check. Reported-by:
Torge Matthies <openglfreak@googlemail.com> Reported-by:
Alexander Zhang <alex@alexyzhang.dev> Tested-by:
José Expósito <jose.exposito89@gmail.com> Signed-off-by:
-
- 15 Nov, 2022 1 commit
-
-
Benjamin Tissoires authored
When I2C_HID_OF_ELAN is set, we need to turn on I2C_HID_CORE to ensure we get all the HID requirements. Fixes: bd3cba00 ("HID: i2c-hid: elan: Add support for Elan eKTH6915 i2c-hid touchscreens") Reported-by:
kernel test robot <lkp@intel.com> Signed-off-by:
-
- 14 Nov, 2022 4 commits
-
-
Anastasia Belova authored
If an empty buf is received, lbuf is also empty. So lbuf is accessed by index -1. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: f31a2de3 ("HID: hid-lg4ff: Allow switching of Logitech gaming wheels between compatibility modes") Signed-off-by:
Anastasia Belova <abelova@astralinux.ru> Signed-off-by:
-
Hans de Goede authored
The Acer Aspire Switch V 10 (SW5-017)'s keyboard-dock uses the same ITE controller setup as other Acer Switch 2-in-1's. This needs special handling for the wifi on/off toggle hotkey as well as to properly report touchpad on/off keypresses. Add the USB-ids for the SW5-017's keyboard-dock with a quirk setting of QUIRK_TOUCHPAD_ON_OFF_REPORT to fix both issues. Cc: Rudolf Polzer <rpolzer@google.com> Signed-off-by:
Hans de Goede <hdegoede@redhat.com> Signed-off-by:
-
Aditya Garg authored
This patch allows users to swap the control and command keys. This can be useful for the Mac users who are used to using Command instead of Control in macOS for various commonly used shortcuts. Signed-off-by:
Aditya Garg <gargaditya08@live.com> Signed-off-by:
-
José Expósito authored
When parsing a frame template with a placeholder indicating the number of buttons present on the frame its value was incorrectly set on big endian architectures due to double little endian conversion. In order to reproduce the issue and verify the fix, run the HID KUnit tests on the PowerPC architecture: $ ./tools/testing/kunit/kunit.py run --kunitconfig=drivers/hid \ --arch=powerpc --cross_compile=powerpc64-linux-gnu- Fixes: 867c8925 ("HID: uclogic: Allow to generate frame templates") Signed-off-by:
-
- 11 Nov, 2022 21 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/hid/hidLinus Torvalds authored
Pull HID fixes from Jiri Kosina: - fix for memory leak (on error path) in Hyper-V driver (Yang Yingliang) - regression fix for handling 3rd barrel switch emulation in Wacom driver (Jason Gerecke) * tag 'for-linus-2022111101' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid: HID: wacom: Fix logic used for 3rd barrel switch emulation HID: hyperv: fix possible memory leak in mousevsc_probe() HID: asus: Remove unused variable in asus_report_tool_width()
-
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/soundLinus Torvalds authored
Pull sound fixes from Takashi Iwai: "Things look calming down, as this contains only a few small fixes: - Fix for a corner-case bug with SG-buffer page allocation helper - A regression fix for Roland USB-audio device probe - A potential memory leak fix at the error path - Handful quirks and device-specific fixes for HD- and USB-audio" * tag 'sound-6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: ALSA: hda: fix potential memleak in 'add_widget_node' ALSA: memalloc: Don't fall back for SG-buffer with IOMMU ALSA: usb-audio: add quirk to fix Hamedal C20 disconnect issue ALSA: hda/realtek: Add Positivo C6300 model quirk ALSA: usb-audio: Add DSD support for Accuphase DAC-60 ALSA: usb-audio: Add quirk entry for M-Audio Micro ALSA: hda/hdmi - enable runtime pm for more AMD display audio ALSA: usb-audio: Remove redundant workaround for Roland quirk ALSA: usb-audio: Yet more regression for for the delayed card registration ALSA: hda/ca0132: add quirk for EVGA Z390 DARK ALSA: hda: clarify comments on SCF changes ALSA: arm: pxa: pxa2xx-ac97-lib: fix return value check of platform_get_irq() ALSA: hda/realtek: Add quirk for ASUS Zenbook using CS35L41
-
git://anongit.freedesktop.org/drm/drmLinus Torvalds authored
Pull drm fixes from Dave Airlie: "Weekly pull request for graphics, mostly amdgpu and i915, with a couple of fixes for vc4 and panfrost, panel quirks and a kconfig change for rcar-du. Nothing seems to be too strange at this stage. amdgpu: - Fix s/r in amdgpu_vram_mgr_new - SMU 13.0.4 update - GPUVM TLB race fix - DCN 3.1.4 fixes - DCN 3.2.x fixes - Vega10 fan fix - BACO fix for Beige Goby board - PSR fix - GPU VM PT locking fixes amdkfd: - CRIU fixes vc4: - HDMI fixes to vc4. panfrost: - Make panfrost's uapi header compile with C++. - Handle 1 gb boundary correctly in panfrost mmu code. panel: - Add rotation quirks for 2 panels. rcar-du: - DSI Kconfig fix i915: - Fix sg_table handling in map_dma_buf - Send PSR update also on invalidate - Do not set cache_dirty for DGFX - Restore userptr probe_range behaviour" * tag 'drm-fixes-2022-11-11' of git://anongit.freedesktop.org/drm/drm: (29 commits) drm/amd/display: only fill dirty rectangles when PSR is enabled drm/amdgpu: disable BACO on special BEIGE_GOBY card drm/amdgpu: Drop eviction lock when allocating PT BO drm/amdgpu: Unlock bo_list_mutex after error handling Revert "drm/amdgpu: Revert "drm/amdgpu: getting fan speed pwm for vega10 properly"" drm/amd/display: Enforce minimum prefetch time for low memclk on DCN32 drm/amd/display: Fix gpio port mapping issue drm/amd/display: Fix reg timeout in enc314_enable_fifo drm/amd/display: Fix FCLK deviation and tool compile issues drm/amd/display: Zeromem mypipe heap struct before using it drm/amd/display: Update SR watermarks for DCN314 drm/amdgpu: workaround for TLB seq race drm/amdkfd: Fix error handling in criu_checkpoint drm/amdkfd: Fix error handling in kfd_criu_restore_events drm/amd/pm: update SMU IP v13.0.4 msg interface header drm: rcar-du: Fix Kconfig dependency between RCAR_DU and RCAR_MIPI_DSI drm/panfrost: Split io-pgtable requests properly drm/amdgpu: Fix the lpfn checking condition in drm buddy drm: panel-orientation-quirks: Add quirk for Acer Switch V 10 (SW5-017) drm: panel-orientation-quirks: Add quirk for Nanote UMPC-01 ...
-
Michael Zaidman authored
Fix 'cast to restricted' sparse warnings reported by kernel test robot in https://lore.kernel.org/all/202211021607.ssjymlKi-lkp@intel.com/Reported-by:
Michael Zaidman <michael.zaidman@gmail.com> Signed-off-by:
-
Michael Zaidman authored
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:
-
Michael Zaidman authored
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 authored
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:
-
Michael Zaidman authored
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:
-
Michael Zaidman authored
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:
-
Michael Zaidman authored
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:
-
Michael Zaidman authored
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:
-
Michael Zaidman authored
The patch increases the read buffer size to 180 bytes. It reduces the number of ft260_i2c_read() calls by three, improving the big reads performance. $ sudo i2ctransfer -y -f 13 w2@0x51 0x0 0x0 r180 Before: [ +4.071878] ft260_i2c_write_read: off 0x0 rlen 180 wlen 2 [ +0.000005] ft260_i2c_write: rep 0xd0 addr 0x51 off 0 len 2 wlen 2 flag 0x2 d[0] 0x0 [ +0.001097] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000175] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.000004] ft260_i2c_read: rep 0xc2 addr 0x51 len 180 rlen 60 flag 0x3 [ +0.008579] ft260_raw_event: i2c resp: rep 0xde len 60 [ +0.000208] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.000001] ft260_i2c_read: rep 0xc2 addr 0x51 len 120 rlen 60 flag 0x0 [ +0.008794] ft260_raw_event: i2c resp: rep 0xde len 60 [ +0.000181] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.000002] ft260_i2c_read: rep 0xc2 addr 0x51 len 60 rlen 60 flag 0x4 [ +0.008817] ft260_raw_event: i2c resp: rep 0xde len 60 [ +0.000223] ft260_xfer_status: bus_status 0x20, clock 100 After: [ +11.611642] ft260_i2c_write_read: off 0x0 rlen 180 wlen 2 [ +0.000005] ft260_i2c_write: rep 0xd0 addr 0x51 off 0 len 2 wlen 2 flag 0x2 d[0] 0x0 [ +0.008001] ft260_xfer_status: bus_status 0x20, clock 100 [ +0.000001] ft260_i2c_read: rep 0xc2 addr 0x51 len 180 rlen 180 flag 0x7 [ +0.008994] ft260_raw_event: i2c resp: rep 0xde len 60 [ +0.007987] ft260_raw_event: i2c resp: rep 0xde len 60 [ +0.007992] ft260_raw_event: i2c resp: rep 0xde len 60 [ +0.000206] ft260_xfer_status: bus_status 0x20, clock 100 Suggested-by:
-
Michael Zaidman authored
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:
Guillaume Champagne <champagne.guillaume.c@gmail.com> Signed-off-by:
-
Michael Zaidman authored
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 authored
The patch improves the I2C write performance by 20 - 30 percent by revising the sleep time in the ft260_hid_output_report_check_status() in the following ways: 1. Reduce the wait time and start to poll earlier. Sending a large amount of data at a low I2C clock rate saturates the internal FT260 buffer and causes hiccups in status readiness, as shown below in the log fragment. Aligning the status check wait time to the worst case significantly reduces the write performance. [Oct22 10:28] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 [ +0.005296] ft260_xfer_status: bus_status 0x20, clock 100 [ +0.013460] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 [ +0.003244] ft260_hid_output_report_check_status: wait 1920 usec, len 38 [ +0.000190] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.015324] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 [ +0.003491] ft260_hid_output_report_check_status: wait 1920 usec, len 38 [ +0.000202] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.016047] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 [ +0.002768] ft260_hid_output_report_check_status: wait 1920 usec, len 38 [ +0.000150] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.011389] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 [ +0.003467] ft260_hid_output_report_check_status: wait 1920 usec, len 38 [ +0.000191] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000172] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000131] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000241] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000233] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000190] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000196] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.011314] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 [ +0.003334] ft260_hid_output_report_check_status: wait 1920 usec, len 38 [ +0.000227] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000204] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000198] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000147] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.011060] ft260_i2c_write: rep 0xd8 addr 0x51 off 0 len 34 d[0] 0x0 Before: $ sudo ./i2cperf -f 2 -o 2 -s 32 -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) ------------------------------------------------------------------- 40510 80 256 8 32 After: $ sudo ./i2cperf -f 2 -o 2 -s 32 -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) ------------------------------------------------------------------- 52584 80 256 8 32 2. Do not sleep if the estimated I2C transfer time is below 2 ms since the first xfer status query frequently takes around 1.5 ms, and the following status queries take about 200us on average. So we usually return from the routine after the first 1 - 3 status checks. [Oct22 11:14] ft260_i2c_write: rep 0xd4 addr 0x51 off 0 len 18 d[0] 0x0 [ +0.004270] ft260_xfer_status: bus_status 0x20, clock 100 [ +0.013889] ft260_i2c_write: rep 0xd4 addr 0x51 off 0 len 18 d[0] 0x0 [ +0.000856] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000138] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.013352] ft260_i2c_write: rep 0xd4 addr 0x51 off 0 len 18 d[0] 0x0 [ +0.001501] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000177] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.014477] ft260_i2c_write: rep 0xd4 addr 0x51 off 0 len 18 d[0] 0x0 [ +0.001377] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000233] ft260_xfer_status: bus_status 0x41, clock 100 [ +0.000191] ft260_xfer_status: bus_status 0x40, clock 100 [ +0.013197] ft260_i2c_write: rep 0xd4 addr 0x51 off 0 len 18 d[0] 0x0 Before: $ sudo ./i2cperf -f 2 -o 2 -s 16 -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) ------------------------------------------------------------------- 28826 73 256 16 16 After: $ sudo ./i2cperf -f 2 -o 2 -s 16 -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) ------------------------------------------------------------------- 45138 73 256 16 16 Signed-off-by: -
Michael Zaidman authored
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:
-
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netLinus Torvalds authored
Pull networking fixes from Jakub Kicinski: "Including fixes from netfilter, wifi, can and bpf. Current release - new code bugs: - can: af_can: can_exit(): add missing dev_remove_pack() of canxl_packet Previous releases - regressions: - bpf, sockmap: fix the sk->sk_forward_alloc warning - wifi: mac80211: fix general-protection-fault in ieee80211_subif_start_xmit() - can: af_can: fix NULL pointer dereference in can_rx_register() - can: dev: fix skb drop check, avoid o-o-b access - nfnetlink: fix potential dead lock in nfnetlink_rcv_msg() Previous releases - always broken: - bpf: fix wrong reg type conversion in release_reference() - gso: fix panic on frag_list with mixed head alloc types - wifi: brcmfmac: fix buffer overflow in brcmf_fweh_event_worker() - wifi: mac80211: set TWT Information Frame Disabled bit as 1 - eth: macsec offload related fixes, make sure to clear the keys from memory - tun: fix memory leaks in the use of napi_get_frags - tun: call napi_schedule_prep() to ensure we own a napi - tcp: prohibit TCP_REPAIR_OPTIONS if data was already sent - ipv6: addrlabel: fix infoleak when sending struct ifaddrlblmsg to network - tipc: fix a msg->req tlv length check - sctp: clear out_curr if all frag chunks of current msg are pruned, avoid list corruption - mctp: fix an error handling path in mctp_init(), avoid leaks" * tag 'net-6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (101 commits) eth: sp7021: drop free_netdev() from spl2sw_init_netdev() MAINTAINERS: Move Vivien to CREDITS net: macvlan: fix memory leaks of macvlan_common_newlink ethernet: tundra: free irq when alloc ring failed in tsi108_open() net: mv643xx_eth: disable napi when init rxq or txq failed in mv643xx_eth_open() ethernet: s2io: disable napi when start nic failed in s2io_card_up() net: atlantic: macsec: clear encryption keys from the stack net: phy: mscc: macsec: clear encryption keys when freeing a flow stmmac: dwmac-loongson: fix missing of_node_put() while module exiting stmmac: dwmac-loongson: fix missing pci_disable_device() in loongson_dwmac_probe() stmmac: dwmac-loongson: fix missing pci_disable_msi() while module exiting cxgb4vf: shut down the adapter when t4vf_update_port_info() failed in cxgb4vf_open() mctp: Fix an error handling path in mctp_init() stmmac: intel: Update PCH PTP clock rate from 200MHz to 204.8MHz net: cxgb3_main: disable napi when bind qsets failed in cxgb_up() net: cpsw: disable napi in cpsw_ndo_open() iavf: Fix VF driver counting VLAN 0 filters ice: Fix spurious interrupt during removal of trusted VF net/mlx5e: TC, Fix slab-out-of-bounds in parse_tc_actions net/mlx5e: E-Switch, Fix comparing termination table instance ... -
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linuxJakub Kicinski authored
Saeed Mahameed says: ==================== mlx5 fixes 2022-11-02 This series provides bug fixes to mlx5 driver. * tag 'mlx5-fixes-2022-11-09' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux: net/mlx5e: TC, Fix slab-out-of-bounds in parse_tc_actions net/mlx5e: E-Switch, Fix comparing termination table instance net/mlx5e: TC, Fix wrong rejection of packet-per-second policing net/mlx5e: Fix tc acts array not to be dependent on enum order net/mlx5e: Fix usage of DMA sync API net/mlx5e: Add missing sanity checks for max TX WQE size net/mlx5: fw_reset: Don't try to load device in case PCI isn't working net/mlx5: E-switch, Set to legacy mode if failed to change switchdev mode net/mlx5: Allow async trigger completion execution on single CPU systems net/mlx5: Bridge, verify LAG state when adding bond to bridge ==================== Link: https://lore.kernel.org/r/20221109184050.108379-1-saeed@kernel.orgSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queueJakub Kicinski authored
Tony Nguyen says: ==================== Intel Wired LAN Driver Updates 2022-11-09 (ice, iavf) This series contains updates to ice and iavf drivers. Norbert stops disabling VF queues that are not enabled for ice driver. Michal stops accounting of VLAN 0 filter to match expectations of PF driver for iavf. * '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue: iavf: Fix VF driver counting VLAN 0 filters ice: Fix spurious interrupt during removal of trusted VF ==================== Link: https://lore.kernel.org/r/20221110003744.201414-1-anthony.l.nguyen@intel.comSigned-off-by:
-
Wei Yongjun authored
It's not necessary to free netdev allocated with devm_alloc_etherdev() and using free_netdev() leads to double free. Fixes: fd3040b9 ("net: ethernet: Add driver for Sunplus SP7021") Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Link: https://lore.kernel.org/r/20221109150116.2988194-1-weiyongjun@huaweicloud.comSigned-off-by:
-
Dave Airlie authored
Merge tag 'drm-intel-fixes-2022-11-10' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes - Fix sg_table handling in map_dma_buf (Matthew Auld) - Send PSR update also on invalidate (Jouni Högander) - Do not set cache_dirty for DGFX (Niranjana Vishwanathapura) - Restore userptr probe_range behaviour (Matthew Auld) Signed-off-by:
Dave Airlie <airlied@redhat.com> From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/Y2zCy5q85qE9W0J8@tursulin-desk
-
- 10 Nov, 2022 4 commits
-
-
git://linuxtv.org/pinchartl/mediaDave Airlie authored
R-Car DSI Kconfig dependency fix Signed-off-by:
-
Florian Fainelli authored
Last patch from Vivien was nearly 3 years ago and he has not reviewed or responded to DSA patches since then, move to CREDITS. Signed-off-by:
Florian Fainelli <f.fainelli@gmail.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/20221109231907.621678-1-f.fainelli@gmail.comSigned-off-by:
-
git://anongit.freedesktop.org/drm/drm-miscDave Airlie authored
drm-misc-fixes for v6.1-rc5: - HDMI fixes to vc4. - Make panfrost's uapi header compile with C++. - Add rotation quirks for 2 panels. - Fix s/r in amdgpu_vram_mgr_new - Handle 1 gb boundary correctly in panfrost mmu code. Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linuxLinus Torvalds authored
Pull btrfs fixes from David Sterba: - revert memory optimization for scrub blocks, this misses errors in 2nd and following blocks - add exception for ENOMEM as reason for transaction abort to not print stack trace, syzbot has reported many - zoned fixes: - fix locking imbalance during scrub - initialize zones for seeding device - initialize zones for cloned device structures - when looking up device, change assertion to a real check as some of the search parameters can be passed by ioctl, reported by syzbot - fix error pointer check in self tests * tag 'for-6.1-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: btrfs: zoned: fix locking imbalance on scrub btrfs: zoned: initialize device's zone info for seeding btrfs: zoned: clone zoned device info when cloning a device Revert "btrfs: scrub: use larger block size for data extent scrub" btrfs: don't print stack trace when transaction is aborted due to ENOMEM btrfs: selftests: fix wrong error check in btrfs_free_dummy_root() btrfs: fix match incorrectly in dev_args_match_device
-
