summaryrefslogtreecommitdiff
path: root/drivers/net
AgeCommit message (Collapse)Author
2021-04-21bonding: 3ad: Fix the conflict between bond_update_slave_arr and the state ↵jinyiting
machine The bond works in mode 4, and performs down/up operations on the bond that is normally negotiated. The probability of bond-> slave_arr is NULL Test commands: ifconfig bond1 down ifconfig bond1 up The conflict occurs in the following process： __dev_open (CPU A) --bond_open --queue_delayed_work(bond->wq,&bond->ad_work,0); --bond_update_slave_arr --bond_3ad_get_active_agg_info ad_work(CPU B) --bond_3ad_state_machine_handler --ad_agg_selection_logic ad_work runs on cpu B. In the function ad_agg_selection_logic, all agg->is_active will be cleared. Before the new active aggregator is selected on CPU B, bond_3ad_get_active_agg_info failed on CPU A, bond->slave_arr will be set to NULL. The best aggregator in ad_agg_selection_logic has not changed, no need to update slave arr. The conflict occurred in that ad_agg_selection_logic clears agg->is_active under mode_lock, but bond_open -> bond_update_slave_arr is inspecting agg->is_active outside the lock. Also, bond_update_slave_arr is normal for potential sleep when allocating memory, so replace the WARN_ON with a call to might_sleep. Signed-off-by: jinyiting <jinyiting@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-04-21mt76: mt7921: reinit wpdma during drv_own if necessaryLorenzo Bianconi
Check dummy reg to reinitialized WPDMA during driver_own operation Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: move mt76_token_init in mt76_alloc_deviceLorenzo Bianconi
In order to remove duplicated code, move mt76_token_init in mt76_alloc_device routine Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: fix a precision vs width bug in printkDan Carpenter
Precision %.*s was intended instead of width %*s. The original code is potentially an information leak. Fixes: c7cc5ec57303 ("mt76: mt7921: rework mt7921_mcu_debug_msg_event routine") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7915: fix a precision vs width bug in printkDan Carpenter
Precision %.*s was intended instead of width %*s. The original code will still print unintended data from beyond the end of skb->data. Fixes: 665b2c780d63 ("mt76: mt7915: limit firmware log message printk to buffer length") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7615: fix a precision vs width bug in printkDan Carpenter
Precision "%.*s" was intended instead of width "%*s". The original code will print garbage from beyond the end of the skb->data. Fixes: d76d6c3ba2b0 ("mt76: mt7615: limit firmware log message printk to buffer length") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: debugfs: introduce napi_threaded nodeLorenzo Bianconi
Introduce napi_threaded debugfs knob in order to enable/disable NAPI threaded support Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: introduce mt7921_mcu_sta_add routineLorenzo Bianconi
mt7921_mcu_sta_add will be used to add and remove wtbl entries. Create broadcast wtbl entry after AP association Co-developed-by: Deren Wu <deren.wu@mediatek.com> Signed-off-by: Deren Wu <deren.wu@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: mt7921_stop should put device in fw_own stateSean Wang
mt7921_stop should put device in fw_own state to reduce power consumption. Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: unschedule mac_work before going to sleepLorenzo Bianconi
In order to wake the device less frequently and so reduce power consumpation, unschedule mac_work before going to sleep Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7663: add awake and doze time accountingLorenzo Bianconi
Similar to mt7921, introduce awake and doze time accounting for runtime pm. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: improve doze opportunityLorenzo Bianconi
Increase mt7921 mac work timeout in oder to have move sleep opportunities Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: get rid of mcu_reset function pointerLorenzo Bianconi
since mcu_reset it used only by mt7921, move the reset callback to mt7921_mcu_parse_response routine and get rid of the function pointer. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7915: do not read rf value from efuse in flash modeShayne Chen
Do not read rf value from efuse when driver is configured to use flash mode. Tested-by: Bruce Chuang <bruce-ss.chuang@mediatek.com> Signed-off-by: Shayne Chen <shayne.chen@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: move token utilities in mt76 common moduleLorenzo Bianconi
Move token management in mt76 common module since it is shared between mt7615, mt7915 and mt7921 drivers Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: move token_lock, token and token_count in mt76_devLorenzo Bianconi
Move token_lock, token and token_count data structures in mt76_dev. This is a preliminary patch to move token management in mt76 common module since it is shared by mt7615, mt7915 and mt7921 drivers. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: fix possible invalid register accessSean Wang
Disable the interrupt and synchronze for the pending irq handlers to ensure the irq tasklet is not being scheduled after the suspend to avoid the possible invalid register access acts when the host pcie controller is suspended. [17932.910534] mt7921e 0000:01:00.0: pci_pm_suspend+0x0/0x22c returned 0 after 21375 usecs [17932.910590] pcieport 0000:00:00.0: calling pci_pm_suspend+0x0/0x22c @ 18565, parent: pci0000:00 [17932.910602] pcieport 0000:00:00.0: pci_pm_suspend+0x0/0x22c returned 0 after 8 usecs [17932.910671] mtk-pcie 11230000.pcie: calling platform_pm_suspend+0x0/0x60 @ 22783, parent: soc [17932.910674] mtk-pcie 11230000.pcie: platform_pm_suspend+0x0/0x60 returned 0 after 0 usecs ... 17933.615352] x1 : 00000000000d4200 x0 : ffffff8269ca2300 [17933.620666] Call trace: [17933.623127] mt76_mmio_rr+0x28/0xf0 [mt76] [17933.627234] mt7921_rr+0x38/0x44 [mt7921e] [17933.631339] mt7921_irq_tasklet+0x54/0x1d8 [mt7921e] [17933.636309] tasklet_action_common+0x12c/0x16c [17933.640754] tasklet_action+0x24/0x2c [17933.644418] __do_softirq+0x16c/0x344 [17933.648082] irq_exit+0xa8/0xac [17933.651224] scheduler_ipi+0xd4/0x148 [17933.654890] handle_IPI+0x164/0x2d4 [17933.658379] gic_handle_irq+0x140/0x178 [17933.662216] el1_irq+0xb8/0x180 [17933.665361] cpuidle_enter_state+0xf8/0x204 [17933.669544] cpuidle_enter+0x38/0x4c [17933.673122] do_idle+0x1a4/0x2a8 [17933.676352] cpu_startup_entry+0x24/0x28 [17933.680276] rest_init+0xd4/0xe0 [17933.683508] arch_call_rest_init+0x10/0x18 [17933.687606] start_kernel+0x340/0x3b4 [17933.691279] Code: aa0003f5 d503201f f953eaa8 8b344108 (b9400113) [17933.697373] ---[ end trace a24b8e26ffbda3c5 ]--- [17933.767846] Kernel panic - not syncing: Fatal exception in interrupt Fixes: ffa1bf97425b ("mt76: mt7921: introduce PM support") Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: enable deep sleep when the device suspendsSean Wang
Enable the deep sleep mode in suspend handler to reduce the power consumption further. Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: introduce mt76_connac_mcu_set_deep_sleep utilitySean Wang
Introduce mt76_connac_mcu_set_deep_sleep to enable deep sleep mode and will be activated immediately when the host returns the ownership to the device. Co-developed-by: Leon Yen <leon.yen@mediatek.com> Signed-off-by: Leon Yen <leon.yen@mediatek.com> Co-developed-by: YN Chen <YN.Chen@mediatek.com> Signed-off-by: YN Chen <YN.Chen@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: introduce mt7921_wpdma_reinit_cond utility routineLorenzo Bianconi
Add mt7921_wpdma_reinit_cond to check dummy reg if driver needs to reinitialized WPDMA after driver_own operation Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Co-developed-by: Leon Yen <leon.yen@mediatek.com> Signed-off-by: Leon Yen <leon.yen@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: introduce mt7921_dma_{enable,disable} utilitiesLorenzo Bianconi
Introduce mt7921_dma_enable and mt7921_dma_disable utilities routine in order for code reusing between mt7921_dma_reset and mt7921_dma_init. This is a preliminary patch to reset dma during device driver_own request. Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: introduce mt7921_wpdma_reset utility routineLorenzo Bianconi
Introduce mt7921_wpdma_reset routine to reset wpdma during chip reset or driver_own request. Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: move mt7921_dma_reset in dma.cLorenzo Bianconi
Move mt7921_dma_reset routine in dma.c and make mt7921_dma_prefetch static. Moreover add force parameter to mt7921_dma_reset signature. This is a preliminary patch to reset dma mt7921_mcu_drv_pmctrl. Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7615: Fix a dereference of pointer sta before it is null checkedColin Ian King
Currently the assignment of idx dereferences pointer sta before sta is null checked, leading to a potential null pointer dereference. Fix this by assigning idx when it is required after the null check on pointer sta. Addresses-Coverity: ("Dereference before null check") Fixes: a4a5a430b076 ("mt76: mt7615: fix TSF configuration") Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: enable sw interruptsLorenzo Bianconi
Enable sw interrupts in order to wake the device from deep sleep receiving packets Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: add awake and doze time accountingLorenzo Bianconi
Introduce awake and doze time accounting for runtime pm. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: remove MT76_STATE_PM in tx pathLorenzo Bianconi
since tx/rx path is now relying pm ref counting, get rid of MT76_STATE_PM check in the tx path Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: use waitqueue for runtime-pmLorenzo Bianconi
Simplify the code using a wait_queue_head_t instead of a completion to wait the chip is fully awake in mt76_connac_pm_wake routine Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: do not run mt76_txq_schedule_all directlyLorenzo Bianconi
In order to not break runtime-pm, do run mt76_txq_schedule_all in mt7615_set_channel/mt7921_set_channel but rely on mt76_worker_schedule Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7615: rely on pm refcounting in mt7615_led_set_configLorenzo Bianconi
Rely on mt76_connac_pm_ref/mt76_connac_pm_unref utility routines in mt7615_led_set_config Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: alaways wake the device before scanningLorenzo Bianconi
move scanning check from mt76_connac_power_save_sched routine to mt7921_pm_power_save_work/mt7615_pm_power_save_work ones Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: get rid of useless MT76_STATE_PM in mt7921_mac_workLorenzo Bianconi
Remove useless MT76_STATE_PM check in mt7921_mac_work since mt7921_mutex_acquire will wake up the device if necessary Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: remove MT76_STATE_PM in mac_tx_freeLorenzo Bianconi
Get rid of MT76_STATE_PM chec in mt7615_mac_tx_free and mt7921_mac_tx_free since we already rely on mt76_connac_pm_unref in the NAPI callback. Remove mt76_connac_power_save_sched calls in mt7615_mac_tx_free and mt7921_mac_tx_free Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: check wake refcount in mcu_fw_pmctrlLorenzo Bianconi
In order to avoid synchronization races between tx and rx path, rely on mt76_connac_skip_fw_pmctrl putting the chip in sleep mode for mt7921 and mt7663 devices Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: unschedule ps_work in mt76_connac_pm_wakeLorenzo Bianconi
In order to avoid synchronization issues between wake and ps works, cancel ps_work in mt76_connac_pm_wake routine Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7663: rely on mt76_connac_pm_ref/mt76_connac_pm_unref in tx/rx napiLorenzo Bianconi
Introduce mt7615_poll_rx rx napi callback for mt7663. Do not access device registers in tx/rx napi if the device is not awake. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: rely on mt76_connac_pm_ref/mt76_connac_pm_unref in tx/rx napiLorenzo Bianconi
Introduce mt7921_poll_rx rx napi callback for mt7921. Do not access device registers in tx/rx napi if the device is not awake. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: dma: add the capability to define a custom rx napi poll routineLorenzo Bianconi
Add the capability to define a custom rx napi callback for each driver. This is a preliminary patch to properly support runtime-pm on rx side Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7663: rely on mt76_connac_pm_ref/mt76_connac_pm_unref in tx pathLorenzo Bianconi
Introduce mt7615_tx_worker routine as mt76 tx worker callback for mt7663. Rely on mt76_connac_pm_ref/mt76_connac_pm_unref to check PM state and increment/decrement wake counter Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: rely on mt76_connac_pm_ref/mt76_connac_pm_unref in tx pathLorenzo Bianconi
Introduce mt7921_tx_worker routine as mt76 tx worker callback for mt7921. Rely on mt76_connac_pm_ref/mt76_connac_pm_unref to check PM state and increment/decrement wake counter Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: introduce wake counter for fw_pmctrl synchronizationLorenzo Bianconi
Introduce wake counter and related spinlock in order to synchronize tx/rx path and fw_pmctrl request. Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7663: fix a race between mt7615_mcu_drv_pmctrl and mt7615_mcu_fw_pmctrlLorenzo Bianconi
Introduce a mutex in order to avoid a race between mt7615_mcu_lp_drv_pmctrl and mt7615_mcu_fw_pmctrl routines since they are run two independent works Fixes: 1f549009b5b2 ("mt76: mt7615: do not request {driver,fw}_own if already granted") Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: fix a race between mt7921_mcu_drv_pmctrl and mt7921_mcu_fw_pmctrlLorenzo Bianconi
Introduce a mutex in order to avoid a race between mt7921_mcu_drv_pmctrl and mt7921_mcu_fw_pmctrl routines since they are run two independent works Fixes: 1d8efc741df8 ("mt76: mt7921: introduce Runtime PM support") Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: mt7921: remove leftover function declarationLorenzo Bianconi
Get rid of leftover mt7921_mcu_add_bss_info routine declaration Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21mt76: connac: move mcu_update_arp_filter in mt76_connac moduleLorenzo Bianconi
Move mt76_connac_mcu_update_arp_filter in mt76_connac module since the code is shared between mt7615 and mt7921 Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
2021-04-21net: ethernet: ravb: Fix release of refclkAdam Ford
The call to clk_disable_unprepare() can happen before priv is initialized. This means moving clk_disable_unprepare out of out_release into a new label. Fixes: 8ef7adc6beb2 ("net: ethernet: ravb: Enable optional refclk") Signed-off-by: Adam Ford <aford173@gmail.com> Reviewed-by: Sergei Shtylyov <sergei.shtylyov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-04-21net: phy: intel-xway: enable integrated led functionsMartin Schiller
The Intel xway phys offer the possibility to deactivate the integrated LED function and to control the LEDs manually. If this was set by the bootloader, it must be ensured that the integrated LED function is enabled for all LEDs when loading the driver. Before commit 6e2d85ec0559 ("net: phy: Stop with excessive soft reset") the LEDs were enabled by a soft-reset of the PHY (using genphy_soft_reset). Initialize the XWAY_MDIO_LED with it's default value (which is applied during a soft reset) instead of adding back the soft reset. This brings back the default LED configuration while still preventing an excessive amount of soft resets. Fixes: 6e2d85ec0559 ("net: phy: Stop with excessive soft reset") Signed-off-by: Martin Schiller <ms@dev.tdt.de> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-04-21net: renesas: ravb: Fix a stuck issue when a lot of frames are receivedYoshihiro Shimoda
When a lot of frames were received in the short term, the driver caused a stuck of receiving until a new frame was received. For example, the following command from other device could cause this issue. $ sudo ping -f -l 1000 -c 1000 <this driver's ipaddress> The previous code always cleared the interrupt flag of RX but checks the interrupt flags in ravb_poll(). So, ravb_poll() could not call ravb_rx() in the next time until a new RX frame was received if ravb_rx() returned true. To fix the issue, always calls ravb_rx() regardless the interrupt flags condition. Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper") Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-04-21net: stmmac: fix TSO and TBS feature enabling during driver openOng Boon Leong
TSO and TBS cannot co-exist and current implementation requires two fixes: 1) stmmac_open() does not need to call stmmac_enable_tbs() because the MAC is reset in stmmac_init_dma_engine() anyway. 2) Inside stmmac_hw_setup(), we should call stmmac_enable_tso() for TX Q that is _not_ configured for TBS. Fixes: 579a25a854d4 ("net: stmmac: Initial support for TBS") Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-04-21stmmac: intel: set TSO/TBS TX Queues default settingsOng Boon Leong
TSO and TBS cannot coexist, for now we set Intel mGbE controller to use below TX Queue mapping: TxQ0 uses TSO and the rest of TXQs supports TBS. Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.3 (git 2.39.1) at 2025-07-31 08:12:40 +0000