aboutsummaryrefslogtreecommitdiff
path: root/drivers
AgeCommit message (Collapse)AuthorFilesLines
2021-12-15igb: Fix removal of unicast MAC filters of VFsKaren Sornek1-14/+14
Move checking condition of VF MAC filter before clearing or adding MAC filter to VF to prevent potential blackout caused by removal of necessary and working VF's MAC filter. Fixes: 1b8b062a99dc ("igb: add VF trust infrastructure") Signed-off-by: Karen Sornek <[email protected]> Tested-by: Konrad Jankowski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-12-15Merge tag 'hyperv-fixes-signed-20211214' of ↵Linus Torvalds1-0/+1
git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux Pull hyperv fix from Wei Liu: "Build fix from Randy Dunlap" * tag 'hyperv-fixes-signed-20211214' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux: hv: utils: add PTP_1588_CLOCK to Kconfig to fix build
2021-12-15usb: dwc2: fix STM ID/VBUS detection startup delay in dwc2_driver_probeAmelie Delaunay1-0/+3
When activate_stm_id_vb_detection is enabled, ID and Vbus detection relies on sensing comparators. This detection needs time to stabilize. A delay was already applied in dwc2_resume() when reactivating the detection, but it wasn't done in dwc2_probe(). This patch adds delay after enabling STM ID/VBUS detection. Then, ID state is good when initializing gadget and host, and avoid to get a wrong Connector ID Status Change interrupt. Fixes: a415083a11cc ("usb: dwc2: add support for STM32MP15 SoCs USB OTG HS and FS") Cc: stable <[email protected]> Acked-by: Minas Harutyunyan <[email protected]> Signed-off-by: Amelie Delaunay <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-15USB: gadget: bRequestType is a bitfield, not a enumGreg Kroah-Hartman3-9/+9
Szymon rightly pointed out that the previous check for the endpoint direction in bRequestType was not looking at only the bit involved, but rather the whole value. Normally this is ok, but for some request types, bits other than bit 8 could be set and the check for the endpoint length could not stall correctly. Fix that up by only checking the single bit. Fixes: 153a2d7e3350 ("USB: gadget: detect too-big endpoint 0 requests") Cc: Felipe Balbi <[email protected]> Reported-by: Szymon Heidrich <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-15soc/tegra: fuse: Fix bitwise vs. logical OR warningNathan Chancellor2-2/+2
A new warning in clang points out two instances where boolean expressions are being used with a bitwise OR instead of logical OR: drivers/soc/tegra/fuse/speedo-tegra20.c:72:9: warning: use of bitwise '|' with boolean operands [-Wbitwise-instead-of-logical] reg = tegra_fuse_read_spare(i) | ^~~~~~~~~~~~~~~~~~~~~~~~~~ || drivers/soc/tegra/fuse/speedo-tegra20.c:72:9: note: cast one or both operands to int to silence this warning drivers/soc/tegra/fuse/speedo-tegra20.c:87:9: warning: use of bitwise '|' with boolean operands [-Wbitwise-instead-of-logical] reg = tegra_fuse_read_spare(i) | ^~~~~~~~~~~~~~~~~~~~~~~~~~ || drivers/soc/tegra/fuse/speedo-tegra20.c:87:9: note: cast one or both operands to int to silence this warning 2 warnings generated. The motivation for the warning is that logical operations short circuit while bitwise operations do not. In this instance, tegra_fuse_read_spare() is not semantically returning a boolean, it is returning a bit value. Use u32 for its return type so that it can be used with either bitwise or boolean operators without any warnings. Fixes: 25cd5a391478 ("ARM: tegra: Add speedo-based process identification") Link: https://github.com/ClangBuiltLinux/linux/issues/1488 Suggested-by: Michał Mirosław <[email protected]> Signed-off-by: Nathan Chancellor <[email protected]> Reviewed-by: Nick Desaulniers <[email protected]> Signed-off-by: Thierry Reding <[email protected]>
2021-12-15Merge tag 'wireless-drivers-2021-12-15' of ↵David S. Miller7-13/+18
git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers Kalle Valo says: ==================== wireless-drivers fixes for v5.16 Second set of fixes for v5.16, hopefully also the last one. I changed my email in MAINTAINERS, one crash fix in iwlwifi and some build problems fixed. iwlwifi * fix crash caused by a warning * fix LED linking problem brcmsmac * rework LED dependencies for being consistent with other drivers mt76 * mt7921: fix build regression ==================== Signed-off-by: David S. Miller <[email protected]>
2021-12-14bcache: fix NULL pointer reference in cached_dev_detach_finishLin Feng1-1/+2
Commit 0259d4498ba4 ("bcache: move calc_cached_dev_sectors to proper place on backing device detach") tries to fix calc_cached_dev_sectors when bcache device detaches, but now we have: cached_dev_detach_finish ... bcache_device_detach(&dc->disk); ... closure_put(&d->c->caching); d->c = NULL; [*explicitly set dc->disk.c to NULL*] list_move(&dc->list, &uncached_devices); calc_cached_dev_sectors(dc->disk.c); [*passing a NULL pointer*] ... Upper codeflows shows how bug happens, this patch fix the problem by caching dc->disk.c beforehand, and cache_set won't be freed under us because c->caching closure at least holds a reference count and closure callback __cache_set_unregister only being called by bch_cache_set_stop which using closure_queue(&c->caching), that means c->caching closure callback for destroying cache_set won't be trigger by previous closure_put(&d->c->caching). So at this stage(while cached_dev_detach_finish is calling) it's safe to access cache_set dc->disk.c. Fixes: 0259d4498ba4 ("bcache: move calc_cached_dev_sectors to proper place on backing device detach") Signed-off-by: Lin Feng <[email protected]> Signed-off-by: Coly Li <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jens Axboe <[email protected]>
2021-12-14RDMA/hns: Replace kfree() with kvfree()Jiacheng Shi1-1/+1
Variables allocated by kvmalloc_array() should not be freed by kfree. Because they may be allocated by vmalloc. So we replace kfree() with kvfree() here. Fixes: 6fd610c5733d ("RDMA/hns: Support 0 hop addressing for SRQ buffer") Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jiacheng Shi <[email protected]> Acked-by: Wenpeng Liang <[email protected]> Signed-off-by: Jason Gunthorpe <[email protected]>
2021-12-14IB/qib: Fix memory leak in qib_user_sdma_queue_pkts()José Expósito1-1/+1
The wrong goto label was used for the error case and missed cleanup of the pkt allocation. Fixes: d39bf40e55e6 ("IB/qib: Protect from buffer overflow in struct qib_user_sdma_pkt fields") Link: https://lore.kernel.org/r/[email protected] Addresses-Coverity-ID: 1493352 ("Resource leak") Signed-off-by: José Expósito <[email protected]> Acked-by: Mike Marciniszyn <[email protected]> Signed-off-by: Jason Gunthorpe <[email protected]>
2021-12-14RDMA/hns: Fix RNR retransmission issue for HIP08Yangyang Li2-7/+65
Due to the discrete nature of the HIP08 timer unit, a requester might finish the timeout period sooner, in elapsed real time, than its responder does, even when both sides share the identical RNR timeout length included in the RNR Nak packet and the responder indeed starts the timing prior to the requester. Furthermore, if a 'providential' resend packet arrived before the responder's timeout period expired, the responder is certainly entitled to drop the packet silently in the light of IB protocol. To address this problem, our team made good use of certain hardware facts: 1) The timing resolution regards the transmission arrangements is 1 microsecond, e.g. if cq_period field is set to 3, it would be interpreted as 3 microsecond by hardware 2) A QPC field shall inform the hardware how many timing unit (ticks) constitutes a full microsecond, which, by default, is 1000 3) It takes 14ns for the processor to handle a packet in the buffer, so the RNR timeout length of 10ns would ensure our processing mechanism is disabled during the entire timeout period and the packet won't be dropped silently To achieve (3), we permanently set the QPC field mentioned in (2) to zero which nominally indicates every time tick is equivalent to a microsecond in wall-clock time; now, a RNR timeout period at face value of 10 would only last 10 ticks, which is 10ns in wall-clock time. It's worth noting that we adapt the driver by magnifying certain configuration parameters(cq_period, eq_period and ack_timeout)by 1000 given the user assumes the configuring timing unit to be microseconds. Also, this particular improvisation is only deployed on HIP08 since other hardware has already solved this issue. Fixes: cfc85f3e4b7f ("RDMA/hns: Add profile support for hip08 driver") Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Yangyang Li <[email protected]> Signed-off-by: Wenpeng Liang <[email protected]> Signed-off-by: Jason Gunthorpe <[email protected]>
2021-12-14drm/amd/pm: fix a potential gpu_metrics_table memory leakLang Yu1-0/+3
Memory is allocated for gpu_metrics_table in renoir_init_smc_tables(), but not freed in int smu_v12_0_fini_smc_tables(). Free it! Fixes: 95868b85764a ("drm/amd/powerplay: add Renoir support for gpu metrics export") Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Lijo Lazar <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2021-12-14drm/amdgpu: correct the wrong cached state for GMC on PICASSOEvan Quan3-5/+18
Pair the operations did in GMC ->hw_init and ->hw_fini. That can help to maintain correct cached state for GMC and avoid unintention gate operation dropping due to wrong cached state. BugLink: https://gitlab.freedesktop.org/drm/amd/-/issues/1828 Signed-off-by: Evan Quan <[email protected]> Acked-by: Guchun Chen <[email protected]> Reviewed-by: Mario Limonciello <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2021-12-14drm/amd/display: Reset DMCUB before HW initNicholas Kazlauskas1-0/+5
[Why] If the firmware wasn't reset by PSP or HW and is currently running then the firmware will hang or perform underfined behavior when we modify its firmware state underneath it. [How] Reset DMCUB before setting up cache windows and performing HW init. Reviewed-by: Aurabindo Jayamohanan Pillai <[email protected]> Acked-by: Pavle Kotarac <[email protected]> Tested-by: Daniel Wheeler <[email protected]> Signed-off-by: Nicholas Kazlauskas <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2021-12-14drm/amd/display: Set exit_optimized_pwr_state for DCN31Nicholas Kazlauskas1-0/+1
[Why] SMU now respects the PHY refclk disable request from driver. This causes a hang during hotplug when PHY refclk was disabled because it's not being re-enabled and the transmitter control starts on dc_link_detect. [How] We normally would re-enable the clk with exit_optimized_pwr_state but this is only set on DCN21 and DCN301. Set it for dcn31 as well. This fixes DMCUB timeouts in the PHY. Fixes: 64b1d0e8d500 ("drm/amd/display: Add DCN3.1 HWSEQ") Reviewed-by: Eric Yang <[email protected]> Acked-by: Pavle Kotarac <[email protected]> Tested-by: Daniel Wheeler <[email protected]> Signed-off-by: Nicholas Kazlauskas <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2021-12-14drm/amd/pm: fix reading SMU FW version from amdgpu_firmware_info on YCMario Limonciello1-0/+3
This value does not get cached into adev->pm.fw_version during startup for smu13 like it does for other SMU like smu12. Signed-off-by: Mario Limonciello <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]> Cc: [email protected]
2021-12-14drm/amdgpu: don't override default ECO_BITs settingHawking Zhang8-9/+0
Leave this bit as hardware default setting Signed-off-by: Hawking Zhang <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]> Cc: [email protected]
2021-12-14drm/amdgpu: correct register access for RLC_JUMP_TABLE_RESTORELe Ma1-2/+2
should count on GC IP base address Signed-off-by: Le Ma <[email protected]> Signed-off-by: Hawking Zhang <[email protected]> Reviewed-by: Hawking Zhang <[email protected]> Signed-off-by: Alex Deucher <[email protected]> Cc: [email protected]
2021-12-14ice: Don't put stale timestamps in the skbKarol Kolacinski2-7/+10
The driver has to check if it does not accidentally put the timestamp in the SKB before previous timestamp gets overwritten. Timestamp values in the PHY are read only and do not get cleared except at hardware reset or when a new timestamp value is captured. The cached_tstamp field is used to detect the case where a new timestamp has not yet been captured, ensuring that we avoid sending stale timestamp data to the stack. Fixes: ea9b847cda64 ("ice: enable transmit timestamps for E810 devices") Signed-off-by: Karol Kolacinski <[email protected]> Tested-by: Gurucharan G <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-12-14ice: Use div64_u64 instead of div_u64 in adjfineKarol Kolacinski1-1/+1
Change the division in ice_ptp_adjfine from div_u64 to div64_u64. div_u64 is used when the divisor is 32 bit but in this case incval is 64 bit and it caused incorrect calculations and incval adjustments. Fixes: 06c16d89d2cb ("ice: register 1588 PTP clock device object for E810 devices") Signed-off-by: Karol Kolacinski <[email protected]> Tested-by: Gurucharan G <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-12-14mlxsw: spectrum_router: Consolidate MAC profiles when possibleDanielle Ratson1-1/+2
Currently, when setting a router interface (RIF) MAC address while the MAC profile is not shared with other RIFs, the profile is edited so that the new MAC address is assigned to it. This does not take into account a situation in which the new MAC address already matches an existing MAC profile. In that situation, two MAC profiles will be occupied even though they hold MAC addresses from the same profile. In order to prevent that, add a check to ensure that editing a MAC profile takes place only when the new MAC address does not match an existing profile. Fixes: 605d25cd782a6 ("mlxsw: spectrum_router: Add RIF MAC profiles support") Reported-by: Maksym Yaremchuk <[email protected]> Tested-by: Maksym Yaremchuk <[email protected]> Signed-off-by: Danielle Ratson <[email protected]> Signed-off-by: Ido Schimmel <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-14Merge branch '40GbE' of ↵David S. Miller1-3/+2
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue Tony Nguyen says: ==================== Intel Wired LAN Driver Updates 2021-12-13 This series contains updates to iavf driver only. Dan Carpenter fixes some missing mutex unlocking. Stefan Assmann restores stopping watchdog from overriding to reset state. ==================== Signed-off-by: David S. Miller <[email protected]>
2021-12-14mmc: sdhci-tegra: Fix switch to HS400ES modePrathamesh Shete1-17/+26
When CMD13 is sent after switching to HS400ES mode, the bus is operating at either MMC_HIGH_26_MAX_DTR or MMC_HIGH_52_MAX_DTR. To meet Tegra SDHCI requirement at HS400ES mode, force SDHCI interface clock to MMC_HS200_MAX_DTR (200 MHz) so that host controller CAR clock and the interface clock are rate matched. Signed-off-by: Prathamesh Shete <[email protected]> Acked-by: Adrian Hunter <[email protected]> Fixes: dfc9700cef77 ("mmc: tegra: Implement HS400 enhanced strobe") Cc: [email protected] Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Ulf Hansson <[email protected]>
2021-12-14net: stmmac: fix tc flower deletion for VLAN priority Rx steeringOng Boon Leong2-13/+90
To replicate the issue:- 1) Add 1 flower filter for VLAN Priority based frame steering:- $ IFDEVNAME=eth0 $ tc qdisc add dev $IFDEVNAME ingress $ tc qdisc add dev $IFDEVNAME root mqprio num_tc 8 \ map 0 1 2 3 4 5 6 7 0 0 0 0 0 0 0 0 \ queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 hw 0 $ tc filter add dev $IFDEVNAME parent ffff: protocol 802.1Q \ flower vlan_prio 0 hw_tc 0 2) Get the 'pref' id $ tc filter show dev $IFDEVNAME ingress 3) Delete a specific tc flower record (say pref 49151) $ tc filter del dev $IFDEVNAME parent ffff: pref 49151 From dmesg, we will observe kernel NULL pointer ooops [ 197.170464] BUG: kernel NULL pointer dereference, address: 0000000000000000 [ 197.171367] #PF: supervisor read access in kernel mode [ 197.171367] #PF: error_code(0x0000) - not-present page [ 197.171367] PGD 0 P4D 0 [ 197.171367] Oops: 0000 [#1] PREEMPT SMP NOPTI <snip> [ 197.171367] RIP: 0010:tc_setup_cls+0x20b/0x4a0 [stmmac] <snip> [ 197.171367] Call Trace: [ 197.171367] <TASK> [ 197.171367] ? __stmmac_disable_all_queues+0xa8/0xe0 [stmmac] [ 197.171367] stmmac_setup_tc_block_cb+0x70/0x110 [stmmac] [ 197.171367] tc_setup_cb_destroy+0xb3/0x180 [ 197.171367] fl_hw_destroy_filter+0x94/0xc0 [cls_flower] The above issue is due to previous incorrect implementation of tc_del_vlan_flow(), shown below, that uses flow_cls_offload_flow_rule() to get struct flow_rule *rule which is no longer valid for tc filter delete operation. struct flow_rule *rule = flow_cls_offload_flow_rule(cls); struct flow_dissector *dissector = rule->match.dissector; So, to ensure tc_del_vlan_flow() deletes the right VLAN cls record for earlier configured RX queue (configured by hw_tc) in tc_add_vlan_flow(), this patch introduces stmmac_rfs_entry as driver-side flow_cls_offload record for 'RX frame steering' tc flower, currently used for VLAN priority. The implementation has taken consideration for future extension to include other type RX frame steering such as EtherType based. v2: - Clean up overly extensive backtrace and rewrite git message to better explain the kernel NULL pointer issue. Fixes: 0e039f5cf86c ("net: stmmac: add RX frame steering based on VLAN priority in tc flower") Tested-by: Kurt Kanzenbach <[email protected]> Signed-off-by: Ong Boon Leong <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-14PCI/MSI: Clear PCI_MSIX_FLAGS_MASKALL on errorThomas Gleixner1-1/+1
PCI_MSIX_FLAGS_MASKALL is set in the MSI-X control register at MSI-X interrupt setup time. It's cleared on success, but the error handling path only clears the PCI_MSIX_FLAGS_ENABLE bit. That's incorrect as the reset state of the PCI_MSIX_FLAGS_MASKALL bit is zero. That can be observed via lspci: Capabilities: [b0] MSI-X: Enable- Count=67 Masked+ Clear the bit in the error path to restore the reset state. Fixes: 438553958ba1 ("PCI/MSI: Enable and mask MSI-X early") Reported-by: Stefan Roese <[email protected]> Signed-off-by: Thomas Gleixner <[email protected]> Tested-by: Stefan Roese <[email protected]> Cc: [email protected] Cc: Bjorn Helgaas <[email protected]> Cc: Michal Simek <[email protected]> Cc: Marek Vasut <[email protected]> Cc: [email protected] Link: https://lore.kernel.org/r/87tufevoqx.ffs@tglx
2021-12-14PCI/MSI: Mask MSI-X vectors only on successStefan Roese1-3/+10
Masking all unused MSI-X entries is done to ensure that a crash kernel starts from a clean slate, which correponds to the reset state of the device as defined in the PCI-E specificion 3.0 and later: Vector Control for MSI-X Table Entries -------------------------------------- "00: Mask bit: When this bit is set, the function is prohibited from sending a message using this MSI-X Table entry. ... This bit’s state after reset is 1 (entry is masked)." A Marvell NVME device fails to deliver MSI interrupts after trying to enable MSI-X interrupts due to that masking. It seems to take the MSI-X mask bits into account even when MSI-X is disabled. While not specification compliant, this can be cured by moving the masking into the success path, so that the MSI-X table entries stay in device reset state when the MSI-X setup fails. [ tglx: Move it into the success path, add comment and amend changelog ] Fixes: aa8092c1d1f1 ("PCI/MSI: Mask all unused MSI-X entries") Signed-off-by: Stefan Roese <[email protected]> Signed-off-by: Thomas Gleixner <[email protected]> Cc: [email protected] Cc: Bjorn Helgaas <[email protected]> Cc: Michal Simek <[email protected]> Cc: Marek Vasut <[email protected]> Cc: [email protected] Link: https://lore.kernel.org/r/[email protected]
2021-12-14drm/i915/display: Fix an unsigned subtraction which can never be negative.Harshit Mogalapalli1-1/+1
smatch warning: drivers/gpu/drm/i915/display/intel_dmc.c:601 parse_dmc_fw() warn: unsigned 'fw->size - offset' is never less than zero Firmware size is size_t and offset is u32. So the subtraction is unsigned which can never be less than zero. Fixes: 3d5928a168a9 ("drm/i915/xelpd: Pipe A DMC plugging") Signed-off-by: Harshit Mogalapalli <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Signed-off-by: Lucas De Marchi <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] (cherry picked from commit 87bb2a410dcfb617b88e4695edf4beb6336dc314) Signed-off-by: Rodrigo Vivi <[email protected]>
2021-12-14USB: serial: option: add Telit FN990 compositionsDaniele Palmas1-0/+8
Add the following Telit FN990 compositions: 0x1070: tty, adb, rmnet, tty, tty, tty, tty 0x1071: tty, adb, mbim, tty, tty, tty, tty 0x1072: rndis, tty, adb, tty, tty, tty, tty 0x1073: tty, adb, ecm, tty, tty, tty, tty Signed-off-by: Daniele Palmas <[email protected]> Link: https://lore.kernel.org/r/[email protected] Cc: [email protected] Signed-off-by: Johan Hovold <[email protected]>
2021-12-14USB: serial: cp210x: fix CP2105 GPIO registrationJohan Hovold1-2/+4
When generalising GPIO support and adding support for CP2102N, the GPIO registration for some CP2105 devices accidentally broke. Specifically, when all the pins of a port are in "modem" mode, and thus unavailable for GPIO use, the GPIO chip would now be registered without having initialised the number of GPIO lines. This would in turn be rejected by gpiolib and some errors messages would be printed (but importantly probe would still succeed). Fix this by initialising the number of GPIO lines before registering the GPIO chip. Note that as for the other device types, and as when all CP2105 pins are muxed for LED function, the GPIO chip is registered also when no pins are available for GPIO use. Reported-by: Maarten Brock <[email protected]> Link: https://lore.kernel.org/r/[email protected] Fixes: c8acfe0aadbe ("USB: serial: cp210x: implement GPIO support for CP2102N") Cc: [email protected] # 4.19 Cc: Karoly Pados <[email protected]> Link: https://lore.kernel.org/r/[email protected] Reviewed-by: Greg Kroah-Hartman <[email protected]> Tested-by: Maarten Brock <[email protected]> Signed-off-by: Johan Hovold <[email protected]>
2021-12-14drm/ast: potential dereference of null pointerJiasheng Jiang1-1/+4
The return value of kzalloc() needs to be checked. To avoid use of null pointer '&ast_state->base' in case of the failure of alloc. Fixes: f0adbc382b8b ("drm/ast: Allocate initial CRTC state of the correct size") Signed-off-by: Jiasheng Jiang <[email protected]> Signed-off-by: Thomas Zimmermann <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-12-13scsi: pm8001: Fix phys_to_virt() usage on dma_addr_tJohn Garry1-17/+21
The driver supports a "direct" mode of operation, where the SMP req frame is directly copied into the command payload (and vice-versa for the SMP resp). To get at the SMP req frame data in the scatterlist the driver uses phys_to_virt() on the DMA mapped memory dma_addr_t . This is broken, and subsequently crashes as follows when an IOMMU is enabled: Unable to handle kernel paging request at virtual address ffff0000fcebfb00 ... pc : pm80xx_chip_smp_req+0x2d0/0x3d0 lr : pm80xx_chip_smp_req+0xac/0x3d0 pm80xx_chip_smp_req+0x2d0/0x3d0 pm8001_task_exec.constprop.0+0x368/0x520 pm8001_queue_command+0x1c/0x30 smp_execute_task_sg+0xdc/0x204 sas_discover_expander.part.0+0xac/0x6cc sas_discover_root_expander+0x8c/0x150 sas_discover_domain+0x3ac/0x6a0 process_one_work+0x1d0/0x354 worker_thread+0x13c/0x470 kthread+0x17c/0x190 ret_from_fork+0x10/0x20 Code: 371806e1 910006d6 6b16033f 54000249 (38766b05) ---[ end trace b91d59aaee98ea2d ]--- note: kworker/u192:0[7] exited with preempt_count 1 Instead use kmap_atomic(). -- Difference to v1: - use kmap_atomic() in both locations Difference to v2: - add whitespace around arithmetic (Damien) Link: https://lore.kernel.org/r/[email protected] Reviewed-by: Damien Le Moal <[email protected]> Signed-off-by: John Garry <[email protected]> Signed-off-by: Martin K. Petersen <[email protected]>
2021-12-13Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhostLinus Torvalds5-7/+9
Pull virtio fixes from Michael Tsirkin: "Misc virtio and vdpa bugfixes" * tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost: vdpa: Consider device id larger than 31 virtio/vsock: fix the transport to work with VMADDR_CID_ANY virtio_ring: Fix querying of maximum DMA mapping size for virtio device virtio: always enter drivers/virtio/ vduse: check that offset is within bounds in get_config() vdpa: check that offsets are within bounds vduse: fix memory corruption in vduse_dev_ioctl()
2021-12-13PCI: mt7621: Convert driver into 'bool'Sergio Paracuellos1-2/+2
The driver is not ready yet to be compiled as a module since it depends on some symbols not exported on MIPS. We have the following current problems: Building mips:allmodconfig ... failed -------------- Error log: ERROR: modpost: missing MODULE_LICENSE() in drivers/pci/controller/pcie-mt7621.o ERROR: modpost: "mips_cm_unlock_other" [drivers/pci/controller/pcie-mt7621.ko] undefined! ERROR: modpost: "mips_cpc_base" [drivers/pci/controller/pcie-mt7621.ko] undefined! ERROR: modpost: "mips_cm_lock_other" [drivers/pci/controller/pcie-mt7621.ko] undefined! ERROR: modpost: "mips_cm_is64" [drivers/pci/controller/pcie-mt7621.ko] undefined! ERROR: modpost: "mips_gcr_base" [drivers/pci/controller/pcie-mt7621.ko] undefined! Temporarily move from 'tristate' to 'bool' until a better solution is ready. Also RALINK is redundant because SOC_MT7621 already depends on it. Hence, simplify condition. Fixes: 2bdd5238e756 ("PCI: mt7621: Add MediaTek MT7621 PCIe host controller driver"). Signed-off-by: Sergio Paracuellos <[email protected]> Reviewed-and-Tested-by: Guenter Roeck <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2021-12-13iavf: do not override the adapter state in the watchdog task (again)Stefan Assmann1-1/+0
The watchdog task incorrectly changes the state to __IAVF_RESETTING, instead of letting the reset task take care of that. This was already resolved by commit 22c8fd71d3a5 ("iavf: do not override the adapter state in the watchdog task") but the problem was reintroduced by the recent code refactoring in commit 45eebd62999d ("iavf: Refactor iavf state machine tracking"). Fixes: 45eebd62999d ("iavf: Refactor iavf state machine tracking") Signed-off-by: Stefan Assmann <[email protected]> Tested-by: Konrad Jankowski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-12-13iavf: missing unlocks in iavf_watchdog_task()Dan Carpenter1-2/+2
This code was re-organized and there some unlocks missing now. Fixes: 898ef1cb1cb2 ("iavf: Combine init and watchdog state machines") Signed-off-by: Dan Carpenter <[email protected]> Tested-by: Konrad Jankowski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-12-13net: stmmac: Add GFP_DMA32 for rx buffers if no 64 capabilityDavid Wu1-4/+12
Use page_pool_alloc_pages instead of page_pool_dev_alloc_pages, which can give the gfp parameter, in the case of not supporting 64-bit width, using 32-bit address memory can reduce a copy from swiotlb. Signed-off-by: David Wu <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-13net: phy: add a note about refcountingRussell King (Oracle)1-0/+3
Recently, a patch has been submitted to "fix" the refcounting for a DT node in of_mdiobus_link_mdiodev(). This is not a leaked refcount. The refcount is passed to the new device. Sadly, coccicheck identifies this location as a leaked refcount, which means we're likely to keep getting patches to "fix" this. However, fixing this will cause breakage. Add a comment to state that the lack of of_node_put() here is intentional. Signed-off-by: Russell King (Oracle) <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-13net: ethernet: ti: add missing of_node_put before returnWang Qing1-9/+20
Fix following coccicheck warning: WARNING: Function "for_each_child_of_node" should have of_node_put() before return. Early exits from for_each_child_of_node should decrement the node reference counter. Signed-off-by: Wang Qing <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-13net: dsa: mv88e6xxx: Unforce speed & duplex in mac_link_down()Marek Behún1-0/+4
Commit 64d47d50be7a ("net: dsa: mv88e6xxx: configure interface settings in mac_config") removed forcing of speed and duplex from mv88e6xxx_mac_config(), where the link is forced down, and left it only in mv88e6xxx_mac_link_up(), by which time link is unforced. It seems that (at least on 88E6190) when changing cmode to 2500base-x, if the link is not forced down, but the speed or duplex are still forced, the forcing of new settings for speed & duplex doesn't take in mv88e6xxx_mac_link_up(). Fix this by unforcing speed & duplex in mv88e6xxx_mac_link_down(). Fixes: 64d47d50be7a ("net: dsa: mv88e6xxx: configure interface settings in mac_config") Signed-off-by: Marek Behún <[email protected]> Reviewed-by: Russell King (Oracle) <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-13net: bcmgenet: Fix NULL vs IS_ERR() checkingMiaoqian Lin1-2/+2
The phy_attach() function does not return NULL. It returns error pointers. Signed-off-by: Miaoqian Lin <[email protected]> Acked-by: Florian Fainelli <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-12-13usb: cdnsp: Fix incorrect status for control requestPawel Laszczak1-0/+2
Patch fixes incorrect status for control request. Without this fix all usb_request objects were returned to upper drivers with usb_reqest->status field set to -EINPROGRESS. Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver") cc: <[email protected]> Reported-by: Ken (Jian) He <[email protected]> Reviewed-by: Peter Chen <[email protected]> Signed-off-by: Pawel Laszczak <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-13usb: cdnsp: Fix issue in cdnsp_log_ep trace eventPawel Laszczak1-2/+2
Patch fixes incorrect order of __entry->stream_id and __entry->state parameters in TP_printk macro. Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver") cc: <[email protected]> Reviewed-by: Peter Chen <[email protected]> Signed-off-by: Pawel Laszczak <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-13usb: cdnsp: Fix incorrect calling of cdnsp_died functionPawel Laszczak1-1/+8
Patch restrict calling of cdnsp_died function during removing modules or software disconnect. This function was called because after transition controller to HALT state the driver starts handling the deferred interrupt. In this case such interrupt can be simple ignored. Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver") cc: <[email protected]> Reviewed-by: Peter Chen <[email protected]> Signed-off-by: Pawel Laszczak <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-13usb: xhci-mtk: fix list_del warning when enable list debugChunfeng Yun1-1/+1
There is warning of 'list_del corruption' when enable list debug (CONFIG_DEBUG_LIST=y), fix it by using list_del_init() Fixes: 4ce186665e7c ("usb: xhci-mtk: Do not use xhci's virt_dev in drop_endpoint") Cc: stable <[email protected]> Signed-off-by: Chunfeng Yun <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-13usb: gadget: u_ether: fix race in setting MAC address in setup phaseMarian Postevca1-10/+6
When listening for notifications through netlink of a new interface being registered, sporadically, it is possible for the MAC to be read as zero. The zero MAC address lasts a short period of time and then switches to a valid random MAC address. This causes problems for netd in Android, which assumes that the interface is malfunctioning and will not use it. In the good case we get this log: InterfaceController::getCfg() ifName usb0 hwAddr 92:a8:f0:73:79:5b ipv4Addr 0.0.0.0 flags 0x1002 In the error case we get these logs: InterfaceController::getCfg() ifName usb0 hwAddr 00:00:00:00:00:00 ipv4Addr 0.0.0.0 flags 0x1002 netd : interfaceGetCfg("usb0") netd : interfaceSetCfg() -> ServiceSpecificException (99, "[Cannot assign requested address] : ioctl() failed") The reason for the issue is the order in which the interface is setup, it is first registered through register_netdev() and after the MAC address is set. Fixed by first setting the MAC address of the net_device and after that calling register_netdev(). Fixes: bcd4a1c40bee885e ("usb: gadget: u_ether: construct with default values and add setters/getters") Cc: [email protected] Signed-off-by: Marian Postevca <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2021-12-13firmware: arm_scpi: Fix string overflow in SCPI genpd driverSudeep Holla1-3/+7
Without the bound checks for scpi_pd->name, it could result in the buffer overflow when copying the SCPI device name from the corresponding device tree node as the name string is set at maximum size of 30. Let us fix it by using devm_kasprintf so that the string buffer is allocated dynamically. Fixes: 8bec4337ad40 ("firmware: scpi: add device power domain support using genpd") Reported-by: Pedro Batista <[email protected]> Signed-off-by: Sudeep Holla <[email protected]> Cc: [email protected] Cc: Cristian Marussi <[email protected]> Link: https://lore.kernel.org/r/[email protected]' Signed-off-by: Arnd Bergmann <[email protected]>
2021-12-13Merge tag 'imx-fixes-5.16-2' of ↵Arnd Bergmann2-0/+23
git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into arm/fixes i.MX fixes for 5.16, round 2: - One fix on imx8m-blk-ctrl driver to get i.MX8MM MIPI reset work properly - Fix CSI_DATA07__ESAI_TX0 pad name in i.MX7ULL pin function header - Remove interconnect property from i.MX8MQ LCDIF device to fix the regression that LCDIF driver stops probe, because interconnect provider driver (imx-bus) hasn't been fully working. - Fix soc-imx driver to register SoC device only on i.MX platform. * tag 'imx-fixes-5.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux: soc: imx: Register SoC device only on i.MX boards soc: imx: imx8m-blk-ctrl: Fix imx8mm mipi reset ARM: dts: imx6ull-pinfunc: Fix CSI_DATA07__ESAI_TX0 pad name arm64: dts: imx8mq: remove interconnect property from lcdif Link: https://lore.kernel.org/r/20211211015625.GK4216@dragon Signed-off-by: Arnd Bergmann <[email protected]>
2021-12-13Merge tag 'tegra-for-5.16-firmware-fixes' of ↵Arnd Bergmann1-2/+3
git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into arm/fixes firmware: Fixes for v5.16-rc5 This contains a single fix for an incorrect usage of sizeof(). * tag 'tegra-for-5.16-firmware-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux: firmware: tegra: Fix error application of sizeof() to pointer Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Arnd Bergmann <[email protected]>
2021-12-13Merge tag 'amdtee-fix-for-v5.16' of ↵Arnd Bergmann1-3/+2
git://git.linaro.org/people/jens.wiklander/linux-tee into arm/fixes AMD-TEE fix IS_ERR() bug * tag 'amdtee-fix-for-v5.16' of git://git.linaro.org/people/jens.wiklander/linux-tee: tee: amdtee: fix an IS_ERR() vs NULL bug Link: https://lore.kernel.org/r/20211130070351.GA2308099@jade Signed-off-by: Arnd Bergmann <[email protected]>
2021-12-13Merge tag 'reset-fixes-for-v5.16' of git://git.pengutronix.de/pza/linux into ↵Arnd Bergmann1-8/+1
arm/fixes Reset controller fixes for v5.16 Well, just one: revert commit c045ceb5a145 ("reset: tegra-bpmp: Handle errors in BPMP response"), which exposed an issue with the Tegra194 HDA controller reset. BPMP response error handling will be reinstated once there's a fix for the HDA issue. * tag 'reset-fixes-for-v5.16' of git://git.pengutronix.de/pza/linux: reset: tegra-bpmp: Revert Handle errors in BPMP response Signed-off-by: Arnd Bergmann <[email protected]>
2021-12-13drm/mediatek: hdmi: Perform NULL pointer check for mtk_hdmi_confAngeloGioacchino Del Regno1-5/+7
In commit 41ca9caaae0b ("drm/mediatek: hdmi: Add check for CEA modes only") a check for CEA modes was added to function mtk_hdmi_bridge_mode_valid() in order to address possible issues on MT8167; moreover, with commit c91026a938c2 ("drm/mediatek: hdmi: Add optional limit on maximal HDMI mode clock") another similar check was introduced. Unfortunately though, at the time of writing, MT8173 does not provide any mtk_hdmi_conf structure and this is crashing the kernel with NULL pointer upon entering mtk_hdmi_bridge_mode_valid(), which happens as soon as a HDMI cable gets plugged in. To fix this regression, add a NULL pointer check for hdmi->conf in the said function, restoring HDMI functionality and avoiding NULL pointer kernel panics. Fixes: 41ca9caaae0b ("drm/mediatek: hdmi: Add check for CEA modes only") Fixes: c91026a938c2 ("drm/mediatek: hdmi: Add optional limit on maximal HDMI mode clock") Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Chun-Kuang Hu <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.