kernel.git/drivers/net/ethernet/intel/igb, branch linux-6.2.y

igb: revert rtnl_lock() that causes deadlock

2023-03-30T10:51:35Z

commit 65f69851e44d71248b952a687e44759a7abb5016 upstream. The commit 6faee3d4ee8b ("igb: Add lock to avoid data race") adds rtnl_lock to eliminate a false data race shown below (FREE from device detaching) | (USE from netdev core) igb_remove | igb_ndo_get_vf_config igb_disable_sriov | vf >= adapter->vfs_allocated_count? kfree(adapter->vf_data) | adapter->vfs_allocated_count = 0 | | memcpy(... adapter->vf_data[vf] The above race will never happen and the extra rtnl_lock causes deadlock below [ 141.420169] [ 141.420672] __schedule+0x2dd/0x840 [ 141.421427] schedule+0x50/0xc0 [ 141.422041] schedule_preempt_disabled+0x11/0x20 [ 141.422678] __mutex_lock.isra.13+0x431/0x6b0 [ 141.423324] unregister_netdev+0xe/0x20 [ 141.423578] igbvf_remove+0x45/0xe0 [igbvf] [ 141.423791] pci_device_remove+0x36/0xb0 [ 141.423990] device_release_driver_internal+0xc1/0x160 [ 141.424270] pci_stop_bus_device+0x6d/0x90 [ 141.424507] pci_stop_and_remove_bus_device+0xe/0x20 [ 141.424789] pci_iov_remove_virtfn+0xba/0x120 [ 141.425452] sriov_disable+0x2f/0xf0 [ 141.425679] igb_disable_sriov+0x4e/0x100 [igb] [ 141.426353] igb_remove+0xa0/0x130 [igb] [ 141.426599] pci_device_remove+0x36/0xb0 [ 141.426796] device_release_driver_internal+0xc1/0x160 [ 141.427060] driver_detach+0x44/0x90 [ 141.427253] bus_remove_driver+0x55/0xe0 [ 141.427477] pci_unregister_driver+0x2a/0xa0 [ 141.428296] __x64_sys_delete_module+0x141/0x2b0 [ 141.429126] ? mntput_no_expire+0x4a/0x240 [ 141.429363] ? syscall_trace_enter.isra.19+0x126/0x1a0 [ 141.429653] do_syscall_64+0x5b/0x80 [ 141.429847] ? exit_to_user_mode_prepare+0x14d/0x1c0 [ 141.430109] ? syscall_exit_to_user_mode+0x12/0x30 [ 141.430849] ? do_syscall_64+0x67/0x80 [ 141.431083] ? syscall_exit_to_user_mode_prepare+0x183/0x1b0 [ 141.431770] ? syscall_exit_to_user_mode+0x12/0x30 [ 141.432482] ? do_syscall_64+0x67/0x80 [ 141.432714] ? exc_page_fault+0x64/0x140 [ 141.432911] entry_SYSCALL_64_after_hwframe+0x72/0xdc Since the igb_disable_sriov() will call pci_disable_sriov() before releasing any resources, the netdev core will synchronize the cleanup to avoid any races. This patch removes the useless rtnl_(un)lock to guarantee correctness. CC: stable@vger.kernel.org Fixes: 6faee3d4ee8b ("igb: Add lock to avoid data race") Reported-by: Corinna Vinschen Link: https://lore.kernel.org/intel-wired-lan/ZAcJvkEPqWeJHO2r@calimero.vinschen.de/ Signed-off-by: Lin Ma Tested-by: Corinna Vinschen Reviewed-by: Jacob Keller Reviewed-by: Simon Horman Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen Signed-off-by: Greg Kroah-Hartman

igb: conditionalize I2C bit banging on external thermal sensor support

2023-02-16T05:20:40Z

Commit a97f8783a937 ("igb: unbreak I2C bit-banging on i350") introduced code to change I2C settings to bit banging unconditionally. However, this patch introduced a regression: On an Intel S2600CWR Server Board with three NICs: - 1x dual-port copper Intel I350 Gigabit Network Connection [8086:1521] (rev 01) fw 1.63, 0x80000dda - 2x quad-port SFP+ with copper SFP Avago ABCU-5700RZ Intel I350 Gigabit Fiber Network Connection [8086:1522] (rev 01) fw 1.52.0 the SFP NICs no longer get link at all. Reverting commit a97f8783a937 or switching to the Intel out-of-tree driver both fix the problem. Per the igb out-of-tree driver, I2C bit banging on i350 depends on support for an external thermal sensor (ETS). However, commit a97f8783a937 added bit banging unconditionally. Additionally, the out-of-tree driver always calls init_thermal_sensor_thresh on probe, while our driver only calls init_thermal_sensor_thresh only in igb_reset(), and only if an ETS is present, ignoring the internal thermal sensor. The affected SFPs don't provide an ETS. Per Intel, the behaviour is a result of i350 firmware requirements. This patch fixes the problem by aligning the behaviour to the out-of-tree driver: - split igb_init_i2c() into two functions: - igb_init_i2c() only performs the basic I2C initialization. - igb_set_i2c_bb() makes sure that E1000_CTRL_I2C_ENA is set and enables bit-banging. - igb_probe() only calls igb_set_i2c_bb() if an ETS is present. - igb_probe() calls init_thermal_sensor_thresh() unconditionally. - igb_reset() aligns its behaviour to igb_probe(), i. e., call igb_set_i2c_bb() if an ETS is present and call init_thermal_sensor_thresh() unconditionally. Fixes: a97f8783a937 ("igb: unbreak I2C bit-banging on i350") Tested-by: Mateusz Palczewski Co-developed-by: Jamie Bainbridge Signed-off-by: Jamie Bainbridge Signed-off-by: Corinna Vinschen Signed-off-by: Tony Nguyen Link: https://lore.kernel.org/r/20230214185549.1306522-1-anthony.l.nguyen@intel.com Signed-off-by: Jakub Kicinski

igb: Fix PPS input and output using 3rd and 4th SDP

2023-02-15T04:44:18Z

Fix handling of the tsync interrupt to compare the pin number with IGB_N_SDP instead of IGB_N_EXTTS/IGB_N_PEROUT and fix the indexing to the perout array. Fixes: cf99c1dd7b77 ("igb: move PEROUT and EXTTS isr logic to separate functions") Reported-by: Matt Corallo Signed-off-by: Miroslav Lichvar Reviewed-by: Jacob Keller Tested-by: Gurucharan G (A Contingent worker at Intel) Signed-off-by: Tony Nguyen Link: https://lore.kernel.org/r/20230213185822.3960072-1-anthony.l.nguyen@intel.com Signed-off-by: Jakub Kicinski

igb: Initialize mailbox message for VF reset

2022-12-14T03:26:36Z

When a MAC address is not assigned to the VF, that portion of the message sent to the VF is not set. The memory, however, is allocated from the stack meaning that information may be leaked to the VM. Initialize the message buffer to 0 so that no information is passed to the VM in this case. Fixes: 6ddbc4cf1f4d ("igb: Indicate failure on vf reset for empty mac address") Reported-by: Akihiko Odaki Signed-off-by: Tony Nguyen Reviewed-by: Akihiko Odaki Reviewed-by: Leon Romanovsky Link: https://lore.kernel.org/r/20221212190031.3983342-1-anthony.l.nguyen@intel.com Signed-off-by: Jakub Kicinski

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

2022-12-09T02:19:59Z

No conflicts. Signed-off-by: Jakub Kicinski

igb: Allocate MSI-X vector when testing

2022-11-30T19:27:22Z

Without this change, the interrupt test fail with MSI-X environment: $ sudo ethtool -t enp0s2 offline [ 43.921783] igb 0000:00:02.0: offline testing starting [ 44.855824] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Down [ 44.961249] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX/TX [ 51.272202] igb 0000:00:02.0: testing shared interrupt [ 56.996975] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX/TX The test result is FAIL The test extra info: Register test (offline) 0 Eeprom test (offline) 0 Interrupt test (offline) 4 Loopback test (offline) 0 Link test (on/offline) 0 Here, "4" means an expected interrupt was not delivered. To fix this, route IRQs correctly to the first MSI-X vector by setting IVAR_MISC. Also, set bit 0 of EIMS so that the vector will not be masked. The interrupt test now runs properly with this change: $ sudo ethtool -t enp0s2 offline [ 42.762985] igb 0000:00:02.0: offline testing starting [ 50.141967] igb 0000:00:02.0: testing shared interrupt [ 56.163957] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX/TX The test result is PASS The test extra info: Register test (offline) 0 Eeprom test (offline) 0 Interrupt test (offline) 0 Loopback test (offline) 0 Link test (on/offline) 0 Fixes: 4eefa8f01314 ("igb: add single vector msi-x testing to interrupt test") Signed-off-by: Akihiko Odaki Reviewed-by: Maciej Fijalkowski Tested-by: Gurucharan G (A Contingent worker at Intel) Signed-off-by: Tony Nguyen

igb: Proactively round up to kmalloc bucket size

2022-11-04T18:09:57Z

In preparation for removing the "silently change allocation size" users of ksize(), explicitly round up all q_vector allocations so that allocations can be correctly compared to ksize(). Cc: Jesse Brandeburg Cc: Tony Nguyen Cc: "David S. Miller" Cc: Eric Dumazet Cc: Jakub Kicinski Cc: Paolo Abeni Cc: intel-wired-lan@lists.osuosl.org Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook Tested-by: Gurucharan (A Contingent worker at Intel) Reviewed-by: Michael J. Ruhl Signed-off-by: Tony Nguyen

igb: Do not free q_vector unless new one was allocated

2022-11-04T18:09:57Z

Avoid potential use-after-free condition under memory pressure. If the kzalloc() fails, q_vector will be freed but left in the original adapter->q_vector[v_idx] array position. Cc: Jesse Brandeburg Cc: Tony Nguyen Cc: "David S. Miller" Cc: Eric Dumazet Cc: Jakub Kicinski Cc: Paolo Abeni Cc: intel-wired-lan@lists.osuosl.org Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook Reviewed-by: Michael J. Ruhl Reviewed-by: Jacob Keller Tested-by: Gurucharan (A Contingent worker at Intel) Signed-off-by: Tony Nguyen

ptp: introduce helpers to adjust by scaled parts per million

2022-10-31T11:14:16Z

Many drivers implement the .adjfreq or .adjfine PTP op function with the same basic logic: 1. Determine a base frequency value 2. Multiply this by the abs() of the requested adjustment, then divide by the appropriate divisor (1 billion, or 65,536 billion). 3. Add or subtract this difference from the base frequency to calculate a new adjustment. A few drivers need the difference and direction rather than the combined new increment value. I recently converted the Intel drivers to .adjfine and the scaled parts per million (65.536 parts per billion) logic. To avoid overflow with minimal loss of precision, mul_u64_u64_div_u64 was used. The basic logic used by all of these drivers is very similar, and leads to a lot of duplicate code to perform the same task. Rather than keep this duplicate code, introduce diff_by_scaled_ppm and adjust_by_scaled_ppm. These helper functions calculate the difference or adjustment necessary based on the scaled parts per million input. The diff_by_scaled_ppm function returns true if the difference should be subtracted, and false otherwise. Update the Intel drivers to use the new helper functions. Other vendor drivers will be converted to .adjfine and this helper function in the following changes. Signed-off-by: Jacob Keller Acked-by: Richard Cochran Signed-off-by: David S. Miller

net: Remove the obsolte u64_stats_fetch_*_irq() users (drivers).

2022-10-29T03:13:54Z

Now that the 32bit UP oddity is gone and 32bit uses always a sequence count, there is no need for the fetch_irq() variants anymore. Convert to the regular interface. Signed-off-by: Thomas Gleixner Signed-off-by: Sebastian Andrzej Siewior Acked-by: Peter Zijlstra (Intel) Signed-off-by: Jakub Kicinski