- Mar 11, 2021
-
-
Pierre-Louis Bossart authored
[ Upstream commit 3d09cf8d ] The quirk table is a mess, let's reorganize it by generation before making sure that the quirks are consistent for each generation. Signed-off-by:
Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by:
Guennadi Liakhovetski <guennadi.liakhovetski@intel.com> Reviewed-by:
Kai Vehmanen <kai.vehmanen@linux.intel.com> Link: https://lore.kernel.org/r/20210208233336.59449-2-pierre-louis.bossart@linux.intel.com Signed-off-by:
Mark Brown <broonie@kernel.org> Signed-off-by:
Sasha Levin <sashal@kernel.org>
-
Nadeem Athani authored
[ Upstream commit 4740b969 ] Cadence controller will not initiate autonomous speed change if strapped as Gen2. The Retrain Link bit is set as quirk to enable this speed change. Link: https://lore.kernel.org/r/20210209144622.26683-3-nadeem@cadence.com Signed-off-by:
Nadeem Athani <nadeem@cadence.com> Signed-off-by:
Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Signed-off-by:
-
Fabian Lesniak authored
[ Upstream commit fee03efc ] This commit adds mixer quirks for the Pioneer DJM-900NXS2 mixer. This device has 6 capture channels, 5 of them allow setting the signal source. This adds controls for these, similar to the DJM-250Mk2. However, playpack channels are not controllable via software like on the 250Mk2, as they can only be set manually on the mixing console. Read-only controls showing the currently selected playback channels are omitted. Signed-off-by:
Fabian Lesniak <fabian@lesniak-it.de> Link: https://lore.kernel.org/r/20210205215116.258724-2-fabian@lesniak-it.de Signed-off-by:
Takashi Iwai <tiwai@suse.de> Signed-off-by:
-
Olivia Mackintosh authored
[ Upstream commit a07df82c ] This allows for N different devices to use the pioneer mixer quirk for setting capture/record type and recording level. The impementation has not changed much with the exception of an additional mask on private_value to allow storing of a device index: DEVICE MASK 0xff000000 GROUP_MASK 0x00ff0000 VALUE_MASK 0x0000ffff This could be improved by changing the arrays of wValues for each channel to contain named definitions (e.g. SND_DJM_CAP_LINE). It would improve readability and perhaps would allow using the same array for multiple channels. The channel number can be specified on the control next to the wIndex. Feedback is very much appreciated as I'm not the most proficient C programmer but am learning as I go. Signed-off-by:
Olivia Mackintosh <livvy@base.nu> Link: https://lore.kernel.org/r/20210205184256.10201-2-livvy@base.nu Signed-off-by:
-
Hans de Goede authored
[ Upstream commit fc6a31b0 ] The ITE8568 EC on the Voyo Winpad A15 presents itself as an I2C-HID attached keyboard and mouse (which seems to never send any events). This needs the I2C_HID_QUIRK_NO_IRQ_AFTER_RESET quirk, otherwise we get the following errors: [ 3688.770850] i2c_hid i2c-ITE8568:00: failed to reset device. [ 3694.915865] i2c_hid i2c-ITE8568:00: failed to reset device. [ 3701.059717] i2c_hid i2c-ITE8568:00: failed to reset device. [ 3707.205944] i2c_hid i2c-ITE8568:00: failed to reset device. [ 3708.227940] i2c_hid i2c-ITE8568:00: can't add hid device: -61 [ 3708.236518] i2c_hid: probe of i2c-ITE8568:00 failed with error -61 Which leads to a significant boot delay. Signed-off-by:
Hans de Goede <hdegoede@redhat.com> Signed-off-by:
Jiri Kosina <jkosina@suse.cz> Signed-off-by:
-
Jisheng Zhang authored
[ Upstream commit 5f7dfda4 ] The SDHCI_PRESET_FOR_* registers are not set(all read as zeros), so set the quirk. Signed-off-by:
Jisheng Zhang <Jisheng.Zhang@synaptics.com> Link: https://lore.kernel.org/r/20201210165510.76b917e5@xhacker.debian Signed-off-by:
Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by:
-
AngeloGioacchino Del Regno authored
[ Upstream commit 8f03c30c ] The PC_DBG_ECO_CNTL register on the Adreno A5xx family gets programmed to some different values on a per-model basis. At least, this is what we intend to do here; Unfortunately, though, this register is being overwritten with a static magic number, right after applying the GPU-specific configuration (including the GPU-specific quirks) and that is effectively nullifying the efforts. Let's remove the redundant and wrong write to the PC_DBG_ECO_CNTL register in order to retain the wanted configuration for the target GPU. Signed-off-by:
AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org> Reviewed-by:
Jordan Crouse <jcrouse@codeaurora.org> Signed-off-by:
Rob Clark <robdclark@chromium.org> Signed-off-by:
-
Kiwoong Kim authored
[ Upstream commit f1ef9047 ] Exynos needs scatterlist entries aligned to page size because it isn't capable of transferring data contained in one DATA IN operation to seversal areas in memory. Link: https://lore.kernel.org/r/80d7e27d6ec537e650a6bd74897b6c60618efcdc.1611026909.git.kwmad.kim@samsung.com Signed-off-by:
Kiwoong Kim <kwmad.kim@samsung.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by:
-
Kiwoong Kim authored
[ Upstream commit a967ddb2 ] Set optimized values for the following timeouts: - FC0_PROTECTION_TIMER - TC0_REPLAY_TIMER - AFC0_REQUEST_TIMER Exynos doesn't yet use traffic class #1. Link: https://lore.kernel.org/r/a0ff44f665a4f31d2f945fd71de03571204c576c.1608513782.git.kwmad.kim@samsung.com Signed-off-by:
-
Kiwoong Kim authored
[ Upstream commit 2b2bfc8a ] Some SoCs require a single scatterlist entry for smaller than page size, i.e. 4KB. When dispatching commands with more than one scatterlist entry under 4KB in size the following behavior is observed: A command to read a block range is dispatched with two scatterlist entries that are named AAA and BBB. After dispatching, the host builds two PRDT entries and during transmission, device sends just one DATA IN because device doesn't care about host DMA. The host then transfers the combined amount of data from start address of the area named AAA. As a consequence, the area that follows AAA in memory would be corrupted. |<------------->| +-------+------------ +-------+ + AAA + (corrupted) ... + BBB + +-------+------------ +-------+ To avoid this we need to enforce page size alignment for sg entries. Link: https://lore.kernel.org/r/56dddef94f60bd9466fd77e69f64bbbd657ed2a1.1611026909.git.kwmad.kim@samsung.com Signed-off-by:
-
Aswath Govindraju authored
[ Upstream commit f6f1f8e6 ] A dummy zero bit is sent preceding the data during a read transfer by the Microchip 93LC46B eeprom (section 2.7 of[1]). This results in right shift of data during a read. In order to ignore this bit a quirk can be added to send an extra zero bit after the read address. Add a quirk to ignore the zero bit sent before data by adding a zero bit after the read address. [1] - https://www.mouser.com/datasheet/2/268/20001749K-277859.pdf Signed-off-by:
Aswath Govindraju <a-govindraju@ti.com> Link: https://lore.kernel.org/r/20210105105817.17644-3-a-govindraju@ti.com Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
-
Kiwoong Kim authored
[ Upstream commit b1d0d2eb ] The UniPro specification states that attribute IDs of the following parameters are vendor-specific so some SoCs could have no regions at the defined addresses: - DME_LocalFC0ProtectionTimeOutVal - DME_LocalTC0ReplayTimeOutVal - DME_LocalAFC0ReqTimeOutVal In addition, the following parameters should be set considering the compatibility between host and device. - PA_PWRMODEUSERDATA0 - PA_PWRMODEUSERDATA1 - PA_PWRMODEUSERDATA2 - PA_PWRMODEUSERDATA3 - PA_PWRMODEUSERDATA4 - PA_PWRMODEUSERDATA5 Introduce a quirk to allow vendor drivers to override the UniPro defaults. Link: https://lore.kernel.org/r/1fedd3dea0ccc980913a5995a10510d86a5b01b9.1608513782.git.kwmad.kim@samsung.com Acked-by:
Avri Altman <Avri.Altman@wdc.com> Signed-off-by:
-
Stanley Chu authored
[ Upstream commit 46ec9592 ] Flush during hibern8 is sufficient on MediaTek platforms, thus enable UFSHCI_QUIRK_SKIP_MANUAL_WB_FLUSH_CTRL to skip enabling fWriteBoosterBufferFlush during WriteBooster initialization. Link: https://lore.kernel.org/r/20201222072928.32328-1-stanley.chu@mediatek.com Reviewed-by:
Stanley Chu <stanley.chu@mediatek.com> Signed-off-by:
-
Pierre-Louis Bossart authored
[ Upstream commit 45c92ec3 ] We missed adding the TGL_HDMI quirk which is very much needed to expose the 4 display pipelines and will be required on TGL topologies. Fixes: 488cdbd8 ('ASoC: Intel: sof_sdw: add quirk for new TigerLake-SDCA device') Signed-off-by:
-
Vitaly Kuznetsov authored
[ Upstream commit 4683d758 ] Commit 7a873e45 ("KVM: selftests: Verify supported CR4 bits can be set before KVM_SET_CPUID2") reveals that KVM allows to set X86_CR4_PCIDE even when PCID support is missing: ==== Test Assertion Failure ==== x86_64/set_sregs_test.c:41: rc pid=6956 tid=6956 - Invalid argument 1 0x000000000040177d: test_cr4_feature_bit at set_sregs_test.c:41 2 0x00000000004014fc: main at set_sregs_test.c:119 3 0x00007f2d9346d041: ?? ??:0 4 0x000000000040164d: _start at ??:? KVM allowed unsupported CR4 bit (0x20000) Add X86_FEATURE_PCID feature check to __cr4_reserved_bits() to make kvm_is_valid_cr4() fail. Signed-off-by:
Vitaly Kuznetsov <vkuznets@redhat.com> Message-Id: <20210201142843.108190-1-vkuznets@redhat.com> Signed-off-by:
Paolo Bonzini <pbonzini@redhat.com> Signed-off-by:
-
Bjorn Helgaas authored
[ Upstream commit 05998379 ] Add function 1 DMA alias quirk for Marvell 88SS9215 PCIe SSD Controller. Link: https://bugzilla.kernel.org/show_bug.cgi?id=42679#c135 Link: https://lore.kernel.org/r/20201110220516.697934-1-helgaas@kernel.org Reported-by:
John Smith <LK7S2ED64JHGLKj75shg9klejHWG49h5hk@protonmail.com> Signed-off-by:
Bjorn Helgaas <bhelgaas@google.com> Signed-off-by:
-
Roger Quadros authored
[ Upstream commit 448373d9 ] Some platforms (e.g. TI) will not have any platform data which will lead to NULL pointer dereference if we don't check for NULL pdata. Fixes: 7cea9657 ("usb: cdns3: add quirk for enable runtime pm by default") Reported-by:
Nishanth Menon <nm@ti.com> Signed-off-by:
Roger Quadros <rogerq@ti.com> Acked-by:
Pawel Laszczak <pawell@cadence.com> Signed-off-by:
Peter Chen <peter.chen@nxp.com> Signed-off-by:
-
Peter Chen authored
[ Upstream commit 7cea9657 ] Some vendors (eg: NXP) may want to enable runtime pm by default for power saving, add one quirk for it. Reviewed-by:
Jun Li <jun.li@nxp.com> Signed-off-by:
-
Peter Chen authored
[ Upstream commit 68ed3f3d ] cdns3 manages PHY by own DRD driver, so skip the management by HCD core. Reviewed-by:
-
Peter Chen authored
[ Upstream commit ed227648 ] cdns3 has some special PM sequence between xhci_bus_suspend and xhci_suspend, add quirk to implement it. Reviewed-by:
-
Chris Chiu authored
[ Upstream commit 1bea2256 ] Tha ARCHOS Cesium 140 tablet has problem with the jack-sensing, thus the heaset functions are not working. Add quirk for this model to select the correct input map, jack-detect options and channel map to enable jack sensing and headset microphone. This device uses IN1 for its internal MIC and JD2 for jack-detect. Signed-off-by:
Chris Chiu <chiu@endlessos.org> Acked-by:
-
Jasper St. Pierre authored
[ Upstream commit 25417185 ] The GIGABYTE GB-BXBT-2807 is a mini-PC which uses off the shelf components, like an Intel GPU which is meant for mobile systems. As such, it, by default, has a backlight controller exposed. Unfortunately, the backlight controller only confuses userspace, which sees the existence of a backlight device node and has the unrealistic belief that there is actually a backlight there! Add a DMI quirk to force the backlight off on this system. Signed-off-by:
Jasper St. Pierre <jstpierre@mecheye.net> Reviewed-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by:
-
Daniel Lee Kruse authored
[ Upstream commit dbf0b3a7 ] On AMD Family 15h (Models 30h-3fh), I/O Memory Management Unit RiSC engine sometimes stalls, requiring a reset. As result, MythTV and w-scan won't scan channels on the AMD Kaveri APU with the Hauppauge QuadHD TV tuner card. For the solution I added the Input/Output Memory Management Unit's PCI Identity of 0x1423 to the broken_dev_id[] array, which is used by a quirks logic meant to fix similar problems with other AMD chipsets. Signed-off-by:
Daniel Lee Kruse <daniel.lee.kruse@protonmail.com> Signed-off-by:
Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by:
Mauro Carvalho Chehab <mchehab+huawei@kernel.org> Signed-off-by:
-
Ethan Warth authored
[ Upstream commit 1008230f ] Mayflash/Dragonrise seems to have yet another device ID for one of their Gamecube controller adapters. Previous to this commit, the adapter registered only one /dev/input/js* device, and all controller inputs (from any controller) were mapped to this device. This patch defines the 1846 USB device ID and enables the HID_QUIRK_MULTI_INPUT quirk for it, which fixes that (with the patch, four /dev/input/js* devices are created, one for each of the four controller ports). Signed-off-by:
Ethan Warth <redyoshi49q@gmail.com> Tested-by:
Wladimir J. van der Laan <laanwj@gmail.com> Signed-off-by:
-
Hans de Goede authored
[ Upstream commit bf753400 ] Add the Acer Aspire Switch 10E SW3-016 to the list of models which use the Acer Switch WMI interface for reporting SW_TABLET_MODE. Signed-off-by:
-
Hans de Goede authored
[ Upstream commit 5c54cb6c ] Add support for SW_TABLET_MODE on the Acer Switch 10 (SW5-012) and the acer Switch 10 (S1003) models. There is no way to detect if this is supported, so this uses DMI based quirks setting force_caps to ACER_CAP_KBD_DOCK (these devices have no other acer-wmi based functionality). The new SW_TABLET_MODE functionality can be tested on devices which are not in the DMI table by passing acer_wmi.force_caps=0x40 on the kernel commandline. Reviewed-by:
Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by:
-
Hans de Goede authored
[ Upstream commit 82cb8a5c ] Not all devices supporting WMID_GUID3 support the wmid3_set_function_mode() call, leading to errors like these: [ 60.138358] acer_wmi: Enabling RF Button failed: 0x1 - 0xff [ 60.140036] acer_wmi: Enabling Launch Manager failed: 0x1 - 0xff Add an ACER_CAP_SET_FUNCTION_MODE capability flag, so that these calls can be disabled through the new force_caps mechanism. Reviewed-by:
-
Hans de Goede authored
[ Upstream commit 39aa009b ] Add a new force_caps module parameter to allow overriding the drivers builtin capability detection mechanism. This can be used to for example: -Disable rfkill functionality on devices where there is an AA OEM DMI record advertising non functional rfkill switches -Force loading of the driver on devices with a missing AA OEM DMI record Note that force_caps is -1 when unset, this allows forcing the capability field to 0, which results in acer-wmi only providing WMI hotkey handling while disabling all other (led, rfkill, backlight) functionality. Reviewed-by:
-
Hans de Goede authored
[ Upstream commit 9feb0763 ] Cleanup accelerometer device handling: -Drop acer_wmi_accel_destroy instead directly call input_unregister_device. -The information tracked by the CAP_ACCEL flag mirrors acer_wmi_accel_dev being NULL. Drop the CAP flag, this is a preparation change for allowing users to override the capability flags. Dropping the flag stops users from causing a NULL pointer dereference by forcing the capability. Reviewed-by:
-
Hans de Goede authored
[ Upstream commit 7c936d8d ] Cleanup the ACER_CAP_FOO defines: -Switch to using BIT() macro. -The ACER_CAP_RFBTN flag is set, but it is never checked anywhere, drop it. -Drop the unused ACER_CAP_ANY define. Reviewed-by:
-
Tony Lindgren authored
[ Upstream commit cfeeea60 ] We need to enable no-reset-on-init quirk for GPMC if the config option for CONFIG_OMAP_GPMC_DEBUG is set. Otherwise the GPMC driver code is unable to show the bootloader configured timings. Signed-off-by:
Tony Lindgren <tony@atomide.com> Signed-off-by:
-
Pierre-Louis Bossart authored
[ Upstream commit 488cdbd8 ] Add quirks for jack detection, rt715 DAI and number of speakers. Signed-off-by:
Rander Wang <rander.wang@linux.intel.com> Link: https://lore.kernel.org/r/20201111214318.150529-2-pierre-louis.bossart@linux.intel.com Signed-off-by:
-
Tsuchiya Yuto authored
[ Upstream commit 4add4d98 ] If a reset is performed, but even the reset fails for some reasons (e.g., on Surface devices, the fw reset requires another quirks), cancel_work_sync() hangs in mwifiex_cleanup_pcie(). # firmware went into a bad state [...] [ 1608.281690] mwifiex_pcie 0000:03:00.0: info: shutdown mwifiex... [ 1608.282724] mwifiex_pcie 0000:03:00.0: rx_pending=0, tx_pending=1, cmd_pending=0 [ 1608.292400] mwifiex_pcie 0000:03:00.0: PREP_CMD: card is removed [ 1608.292405] mwifiex_pcie 0000:03:00.0: PREP_CMD: card is removed # reset performed after firmware went into a bad state [ 1609.394320] mwifiex_pcie 0000:03:00.0: WLAN FW already running! Skip FW dnld [ 1609.394335] mwifiex_pcie 0000:03:00.0: WLAN FW is active # but even the reset failed [ 1619.499049] mwifiex_pcie 0000:03:00.0: mwifiex_cmd_timeout_func: Timeout cmd id = 0xfa, act = 0xe000 [ 1619.499094] mwifiex_pcie 0000:03:00.0: num_data_h2c_failure = 0 [ 1619.499103] mwifiex_pcie 0000:03:00.0: num_cmd_h2c_failure = 0 [ 1619.499110] mwifiex_pcie 0000:03:00.0: is_cmd_timedout = 1 [ 1619.499117] mwifiex_pcie 0000:03:00.0: num_tx_timeout = 0 [ 1619.499124] mwifiex_pcie 0000:03:00.0: last_cmd_index = 0 [ 1619.499133] mwifiex_pcie 0000:03:00.0: last_cmd_id: fa 00 07 01 07 01 07 01 07 01 [ 1619.499140] mwifiex_pcie 0000:03:00.0: last_cmd_act: 00 e0 00 00 00 00 00 00 00 00 [ 1619.499147] mwifiex_pcie 0000:03:00.0: last_cmd_resp_index = 3 [ 1619.499155] mwifiex_pcie 0000:03:00.0: last_cmd_resp_id: 07 81 07 81 07 81 07 81 07 81 [ 1619.499162] mwifiex_pcie 0000:03:00.0: last_event_index = 2 [ 1619.499169] mwifiex_pcie 0000:03:00.0: last_event: 58 00 58 00 58 00 58 00 58 00 [ 1619.499177] mwifiex_pcie 0000:03:00.0: data_sent=0 cmd_sent=1 [ 1619.499185] mwifiex_pcie 0000:03:00.0: ps_mode=0 ps_state=0 [ 1619.499215] mwifiex_pcie 0000:03:00.0: info: _mwifiex_fw_dpc: unregister device # mwifiex_pcie_work hang happening [ 1823.233923] INFO: task kworker/3:1:44 blocked for more than 122 seconds. [ 1823.233932] Tainted: G WC OE 5.10.0-rc1-1-mainline #1 [ 1823.233935] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 1823.233940] task:kworker/3:1 state:D stack: 0 pid: 44 ppid: 2 flags:0x00004000 [ 1823.233960] Workqueue: events mwifiex_pcie_work [mwifiex_pcie] [ 1823.233965] Call Trace: [ 1823.233981] __schedule+0x292/0x820 [ 1823.233990] schedule+0x45/0xe0 [ 1823.233995] schedule_timeout+0x11c/0x160 [ 1823.234003] wait_for_completion+0x9e/0x100 [ 1823.234012] __flush_work.isra.0+0x156/0x210 [ 1823.234018] ? flush_workqueue_prep_pwqs+0x130/0x130 [ 1823.234026] __cancel_work_timer+0x11e/0x1a0 [ 1823.234035] mwifiex_cleanup_pcie+0x28/0xd0 [mwifiex_pcie] [ 1823.234049] mwifiex_free_adapter+0x24/0xe0 [mwifiex] [ 1823.234060] _mwifiex_fw_dpc+0x294/0x560 [mwifiex] [ 1823.234074] mwifiex_reinit_sw+0x15d/0x300 [mwifiex] [ 1823.234080] mwifiex_pcie_reset_done+0x50/0x80 [mwifiex_pcie] [ 1823.234087] pci_try_reset_function+0x5c/0x90 [ 1823.234094] process_one_work+0x1d6/0x3a0 [ 1823.234100] worker_thread+0x4d/0x3d0 [ 1823.234107] ? rescuer_thread+0x410/0x410 [ 1823.234112] kthread+0x142/0x160 [ 1823.234117] ? __kthread_bind_mask+0x60/0x60 [ 1823.234124] ret_from_fork+0x22/0x30 [...] This is a deadlock caused by calling cancel_work_sync() in mwifiex_cleanup_pcie(): - Device resets are done via mwifiex_pcie_card_reset() - which schedules card->work to call mwifiex_pcie_card_reset_work() - which calls pci_try_reset_function(). - This leads to mwifiex_pcie_reset_done() be called on the same workqueue, which in turn calls - mwifiex_reinit_sw() and that calls - _mwifiex_fw_dpc(). The problem is now that _mwifiex_fw_dpc() calls mwifiex_free_adapter() in case firmware initialization fails. That ends up calling mwifiex_cleanup_pcie(). Note that all those calls are still running on the workqueue. So when mwifiex_cleanup_pcie() now calls cancel_work_sync(), it's really waiting on itself to complete, causing a deadlock. This commit fixes the deadlock by skipping cancel_work_sync() on a reset failure path. After this commit, when reset fails, the following output is expected to be shown: kernel: mwifiex_pcie 0000:03:00.0: info: _mwifiex_fw_dpc: unregister device kernel: mwifiex: Failed to bring up adapter: -5 kernel: mwifiex_pcie 0000:03:00.0: reinit failed: -5 To reproduce this issue, for example, try putting the root port of wifi into D3 (replace "00:1d.3" with your setup). # put into D3 (root port) sudo setpci -v -s 00:1d.3 CAP_PM+4.b=0b Cc: Maximilian Luz <luzmaximilian@gmail.com> Signed-off-by:
Tsuchiya Yuto <kitakar@gmail.com> Signed-off-by:
Kalle Valo <kvalo@codeaurora.org> Link: https://lore.kernel.org/r/20201028142346.18355-1-kitakar@gmail.com Signed-off-by:
-
Abhishek Pandit-Subedi authored
[ Upstream commit 54780138 ] WCN3991 supports connectable advertisements so we need to add the valid le states quirk so the 'central-peripheral' role is exposed in userspace. Signed-off-by:
Abhishek Pandit-Subedi <abhishekpandit@chromium.org> Signed-off-by:
Marcel Holtmann <marcel@holtmann.org> Signed-off-by:
-
Andrey Ryabinin authored
commit 140456f9 upstream. increase_address_space() calls get_zeroed_page(gfp) under spin_lock with disabled interrupts. gfp flags passed to increase_address_space() may allow sleeping, so it comes to this: BUG: sleeping function called from invalid context at mm/page_alloc.c:4342 in_atomic(): 1, irqs_disabled(): 1, pid: 21555, name: epdcbbf1qnhbsd8 Call Trace: dump_stack+0x66/0x8b ___might_sleep+0xec/0x110 __alloc_pages_nodemask+0x104/0x300 get_zeroed_page+0x15/0x40 iommu_map_page+0xdd/0x3e0 amd_iommu_map+0x50/0x70 iommu_map+0x106/0x220 vfio_iommu_type1_ioctl+0x76e/0x950 [vfio_iommu_type1] do_vfs_ioctl+0xa3/0x6f0 ksys_ioctl+0x66/0x70 __x64_sys_ioctl+0x16/0x20 do_syscall_64+0x4e/0x100 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by moving get_zeroed_page() out of spin_lock/unlock section. Fixes: 754265bc ("iommu/amd: Fix race in increase_address_space()") Signed-off-by:
Andrey Ryabinin <arbn@yandex-team.com> Acked-by:
Will Deacon <will@kernel.org> Cc: <stable@vger.kernel.org> Link: https://lore.kernel.org/r/20210217143004.19165-1-arbn@yandex-team.com Signed-off-by:
Joerg Roedel <jroedel@suse.de> Signed-off-by:
-
Nikolay Borisov authored
commit 4d14c5cd upstream Calling btrfs_qgroup_reserve_meta_prealloc from btrfs_delayed_inode_reserve_metadata can result in flushing delalloc while holding a transaction and delayed node locks. This is deadlock prone. In the past multiple commits: * ae5e070e ("btrfs: qgroup: don't try to wait flushing if we're already holding a transaction") * 6f23277a ("btrfs: qgroup: don't commit transaction when we already hold the handle") Tried to solve various aspects of this but this was always a whack-a-mole game. Unfortunately those 2 fixes don't solve a deadlock scenario involving btrfs_delayed_node::mutex. Namely, one thread can call btrfs_dirty_inode as a result of reading a file and modifying its atime: PID: 6963 TASK: ffff8c7f3f94c000 CPU: 2 COMMAND: "test" #0 __schedule at ffffffffa529e07d #1 schedule at ffffffffa529e4ff #2 schedule_timeout at ffffffffa52a1bdd #3 wait_for_completion at ffffffffa529eeea <-- sleeps with delayed node mutex held #4 start_delalloc_inodes at ffffffffc0380db5 #5 btrfs_start_delalloc_snapshot at ffffffffc0393836 #6 try_flush_qgroup at ffffffffc03f04b2 #7 __btrfs_qgroup_reserve_meta at ffffffffc03f5bb6 <-- tries to reserve space and starts delalloc inodes. #8 btrfs_delayed_update_inode at ffffffffc03e31aa <-- acquires delayed node mutex #9 btrfs_update_inode at ffffffffc0385ba8 #10 btrfs_dirty_inode at ffffffffc038627b <-- TRANSACTIION OPENED #11 touch_atime at ffffffffa4cf0000 #12 generic_file_read_iter at ffffffffa4c1f123 #13 new_sync_read at ffffffffa4ccdc8a #14 vfs_read at ffffffffa4cd0849 #15 ksys_read at ffffffffa4cd0bd1 #16 do_syscall_64 at ffffffffa4a052eb #17 entry_SYSCALL_64_after_hwframe at ffffffffa540008c This will cause an asynchronous work to flush the delalloc inodes to happen which can try to acquire the same delayed_node mutex: PID: 455 TASK: ffff8c8085fa4000 CPU: 5 COMMAND: "kworker/u16:30" #0 __schedule at ffffffffa529e07d #1 schedule at ffffffffa529e4ff #2 schedule_preempt_disabled at ffffffffa529e80a #3 __mutex_lock at ffffffffa529fdcb <-- goes to sleep, never wakes up. #4 btrfs_delayed_update_inode at ffffffffc03e3143 <-- tries to acquire the mutex #5 btrfs_update_inode at ffffffffc0385ba8 <-- this is the same inode that pid 6963 is holding #6 cow_file_range_inline.constprop.78 at ffffffffc0386be7 #7 cow_file_range at ffffffffc03879c1 #8 btrfs_run_delalloc_range at ffffffffc038894c #9 writepage_delalloc at ffffffffc03a3c8f #10 __extent_writepage at ffffffffc03a4c01 #11 extent_write_cache_pages at ffffffffc03a500b #12 extent_writepages at ffffffffc03a6de2 #13 do_writepages at ffffffffa4c277eb #14 __filemap_fdatawrite_range at ffffffffa4c1e5bb #15 btrfs_run_delalloc_work at ffffffffc0380987 <-- starts running delayed nodes #16 normal_work_helper at ffffffffc03b706c #17 process_one_work at ffffffffa4aba4e4 #18 worker_thread at ffffffffa4aba6fd #19 kthread at ffffffffa4ac0a3d #20 ret_from_fork at ffffffffa54001ff To fully address those cases the complete fix is to never issue any flushing while holding the transaction or the delayed node lock. This patch achieves it by calling qgroup_reserve_meta directly which will either succeed without flushing or will fail and return -EDQUOT. In the latter case that return value is going to be propagated to btrfs_dirty_inode which will fallback to start a new transaction. That's fine as the majority of time we expect the inode will have BTRFS_DELAYED_NODE_INODE_DIRTY flag set which will result in directly copying the in-memory state. Fixes: c53e9653 ("btrfs: qgroup: try to flush qgroup space when we get -EDQUOT") CC: stable@vger.kernel.org # 5.10+ Reviewed-by:
Qu Wenruo <wqu@suse.com> Signed-off-by:
Nikolay Borisov <nborisov@suse.com> Signed-off-by:
David Sterba <dsterba@suse.com> [sudip: adjust context] Signed-off-by:
Sudip Mukherjee <sudipm.mukherjee@gmail.com> Signed-off-by:
-
Nikolay Borisov authored
commit 80e9baed upstream Signed-off-by:
-
Nathan Chancellor authored
commit e9c6deee upstream Similar to commit 28187dc8 ("ARM: 9025/1: Kconfig: CPU_BIG_ENDIAN depends on !LD_IS_LLD"), ld.lld prior to 13.0.0 does not properly support aarch64 big endian, leading to the following build error when CONFIG_CPU_BIG_ENDIAN is selected: ld.lld: error: unknown emulation: aarch64linuxb This has been resolved in LLVM 13. To avoid errors like this, only allow CONFIG_CPU_BIG_ENDIAN to be selected if using ld.bfd or ld.lld 13.0.0 and newer. While we are here, the indentation of this symbol used spaces since its introduction in commit a872013d ("arm64: kconfig: allow CPU_BIG_ENDIAN to be selected"). Change it to tabs to be consistent with kernel coding style. Link: https://github.com/ClangBuiltLinux/linux/issues/380 Link: https://github.com/ClangBuiltLinux/linux/issues/1288 Link: https://github.com/llvm/llvm-project/commit/7605a9a009b5fa3bdac07e3131c8d82f6d08feb7 Link: https://github.com/llvm/llvm-project/commit/eea34aae2e74e9b6fbdd5b95f479bc7f397bf387 Reported-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
Nathan Chancellor <nathan@kernel.org> Reviewed-by:
Nick Desaulniers <ndesaulniers@google.com> Link: https://lore.kernel.org/r/20210209005719.803608-1-nathan@kernel.org Signed-off-by:
-
Helge Deller authored
commit 778e45d7 upstream The kernel test robot reported multiple linkage problems like this: hppa64-linux-ld: init/main.o(.init.text+0x56c): cannot reach printk init/main.o: in function `unknown_bootoption': (.init.text+0x56c): relocation truncated to fit: R_PARISC_PCREL22F against symbol `printk' defined in .text.unlikely section in kernel/printk/printk.o There are two ways to solve it: a) Enable the -mlong-call compiler option (CONFIG_MLONGCALLS), b) Add long branch stub support in 64-bit linker. While b) is the long-term solution, this patch works around the issue by automatically enabling the CONFIG_MLONGCALLS option when CONFIG_COMPILE_TEST is set, which indicates that a non-production kernel (e.g. 0-day kernel) is built. Signed-off-by:
Helge Deller <deller@gmx.de> Reported-by:
kernel test robot <lkp@intel.com> Fixes: 00e35f2b ("parisc: Enable -mlong-calls gcc option by default when !CONFIG_MODULES") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by:
-
Zoltán Böszörményi authored
commit dc22c1c0 upstream My 2TB SKC2000 showed the exact same symptoms that were provided in 538e4a8c ("nvme-pci: avoid the deepest sleep state on Kingston A2000 SSDs"), i.e. a complete NVME lockup that needed cold boot to get it back. According to some sources, the A2000 is simply a rebadged SKC2000 with a slightly optimized firmware. Adding the SKC2000 PCI ID to the quirk list with the same workaround as the A2000 made my laptop survive a 5 hours long Yocto bootstrap buildfest which reliably triggered the SSD lockup previously. Signed-off-by:
Zoltán Böszörményi <zboszor@gmail.com> Signed-off-by:
Christoph Hellwig <hch@lst.de> [sudip: adjust context] Signed-off-by:
-
