- 14 Feb, 2022 17 commits
-
-
David S. Miller authored
Sebastian Andrzej Siewior says: ==================== net: dev: PREEMPT_RT fixups. this series removes or replaces preempt_disable() and local_irq_save() sections which are problematic on PREEMPT_RT. Patch 2 makes netif_rx() work from any context after I found suggestions for it in an old thread. Should that work, then the context-specific variants could be removed. v2…v3: - #2 - Export __netif_rx() so it can be used by everyone. - Add a lockdep assert to check for interrupt context. - Update the kernel doc and mention that the skb is posted to backlog NAPI. - Use __netif_rx() also in drivers/net/*.c. - Added Toke''s review tag and kept Eric's desptite the changes made. v1…v2: - #1 and #2 - merge patch 1 und 2 from the series (as per Toke). - updated patch description and corrected the first commit number (as per Eric). - #2 - Provide netif_rx() as in v1 and additionally __netif_rx() without local_bh disable()+enable() for the loopback driver. __netif_rx() is not exported (loopback is built-in only) so it won't be used drivers. If this doesn't work then we can still export/ define a wrapper as Eric suggested. - Added a comment that netif_rx() considered legacy. - #3 - Moved ____napi_schedule() into rps_ipi_queued() and renamed it napi_schedule_rps(). https://lore.kernel.org/all/20220204201259.1095226-1-bigeasy@linutronix.de/ v1: https://lore.kernel.org/all/20220202122848.647635-1-bigeasy@linutronix.de ==================== Signed-off-by:David S. Miller <davem@davemloft.net>
-
Sebastian Andrzej Siewior authored
Disabling interrupts and in the RPS case locking input_pkt_queue is split into local_irq_disable() and optional spin_lock(). This breaks on PREEMPT_RT because the spinlock_t typed lock can not be acquired with disabled interrupts. The sections in which the lock is acquired is usually short in a sense that it is not causing long und unbounded latiencies. One exception is the skb_flow_limit() invocation which may invoke a BPF program (and may require sleeping locks). By moving local_irq_disable() + spin_lock() into rps_lock(), we can keep interrupts disabled on !PREEMPT_RT and enabled on PREEMPT_RT kernels. Without RPS on a PREEMPT_RT kernel, the needed synchronisation happens as part of local_bh_disable() on the local CPU. ____napi_schedule() is only invoked if sd is from the local CPU. Replace it with __napi_schedule_irqoff() which already disables interrupts on PREEMPT_RT as needed. Move this call to rps_ipi_queued() and rename the function to napi_schedule_rps as suggested by Jakub. Signed-off-by:
Sebastian Andrzej Siewior <bigeasy@linutronix.de> Reviewed-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Sebastian Andrzej Siewior authored
Dave suggested a while ago (eleven years by now) "Let's make netif_rx() work in all contexts and get rid of netif_rx_ni()". Eric agreed and pointed out that modern devices should use netif_receive_skb() to avoid the overhead. In the meantime someone added another variant, netif_rx_any_context(), which behaves as suggested. netif_rx() must be invoked with disabled bottom halves to ensure that pending softirqs, which were raised within the function, are handled. netif_rx_ni() can be invoked only from process context (bottom halves must be enabled) because the function handles pending softirqs without checking if bottom halves were disabled or not. netif_rx_any_context() invokes on the former functions by checking in_interrupts(). netif_rx() could be taught to handle both cases (disabled and enabled bottom halves) by simply disabling bottom halves while invoking netif_rx_internal(). The local_bh_enable() invocation will then invoke pending softirqs only if the BH-disable counter drops to zero. Eric is concerned about the overhead of BH-disable+enable especially in regard to the loopback driver. As critical as this driver is, it will receive a shortcut to avoid the additional overhead which is not needed. Add a local_bh_disable() section in netif_rx() to ensure softirqs are handled if needed. Provide __netif_rx() which does not disable BH and has a lockdep assert to ensure that interrupts are disabled. Use this shortcut in the loopback driver and in drivers/net/*.c. Make netif_rx_ni() and netif_rx_any_context() invoke netif_rx() so they can be removed once they are no more users left. Link: https://lkml.kernel.org/r/20100415.020246.218622820.davem@davemloft.netSigned-off-by:
Eric Dumazet <edumazet@google.com> Reviewed-by:
Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by:
-
Sebastian Andrzej Siewior authored
The preempt_disable() () section was introduced in commit cece1945 ("net: disable preemption before call smp_processor_id()") and adds it in case this function is invoked from preemtible context and because get_cpu() later on as been added. The get_cpu() usage was added in commit b0e28f1e ("net: netif_rx() must disable preemption") because ip_dev_loopback_xmit() invoked netif_rx() with enabled preemption causing a warning in smp_processor_id(). The function netif_rx() should only be invoked from an interrupt context which implies disabled preemption. The commit e30b38c2 ("ip: Fix ip_dev_loopback_xmit()") was addressing this and replaced netif_rx() with in netif_rx_ni() in ip_dev_loopback_xmit(). Based on the discussion on the list, the former patch (b0e28f1e) should not have been applied only the latter (e30b38c2). Remove get_cpu() and preempt_disable() since the function is supposed to be invoked from context with stable per-CPU pointers. Bottom halves have to be disabled at this point because the function may raise softirqs which need to be processed. Link: https://lkml.kernel.org/r/20100415.013347.98375530.davem@davemloft.netSigned-off-by: -
Dave Ertman authored
The status of support for RDMA is currently being tracked with two separate status flags. This is unnecessary with the current state of the driver. Simplify status tracking down to a single flag. Rename the helper function to denote the RDMA specific status and universally use the helper function to test the status bit. Signed-off-by:
Dave Ertman <david.m.ertman@intel.com> Tested-by:
Leszek Kaliszczuk <leszek.kaliszczuk@intel.com> Reviewed-by:
Leon Romanovsky <leonro@nvidia.com> Signed-off-by:
Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by:
-
David S. Miller authored
Colin Foster says: ==================== use bulk reads for ocelot statistics Ocelot loops over memory regions to gather stats on different ports. These regions are mostly continuous, and are ordered. This patch set uses that information to break the stats reads into regions that can get read in bulk. The motiviation is for general cleanup, but also for SPI. Performing two back-to-back reads on a SPI bus require toggling the CS line, holding, re-toggling the CS line, sending 3 address bytes, sending N padding bytes, then actually performing the read. Bulk reads could reduce almost all of that overhead, but require that the reads are performed via regmap_bulk_read. Verified with eth0 hooked up to the CPU port: NIC statistics: Good Rx Frames: 905 Rx Octets: 78848 Good Tx Frames: 691 Tx Octets: 52516 Rx + Tx 65-127 Octet Frames: 1574 Rx + Tx 128-255 Octet Frames: 22 Net Octets: 131364 Rx DMA chan 0: head_enqueue: 1 Rx DMA chan 0: tail_enqueue: 1032 Rx DMA chan 0: busy_dequeue: 628 Rx DMA chan 0: good_dequeue: 905 Tx DMA chan 0: head_enqueue: 346 Tx DMA chan 0: tail_enqueue: 345 Tx DMA chan 0: misqueued: 345 Tx DMA chan 0: empty_dequeue: 346 Tx DMA chan 0: good_dequeue: 691 p00_rx_octets: 52516 p00_rx_unicast: 691 p00_rx_frames_65_to_127_octets: 691 p00_tx_octets: 78848 p00_tx_unicast: 905 p00_tx_frames_65_to_127_octets: 883 p00_tx_frames_128_255_octets: 22 p00_tx_green_prio_0: 905 And with swp2 connected to swp3 with STP enabled: NIC statistics: tx_packets: 379 tx_bytes: 19708 rx_packets: 1 rx_bytes: 46 rx_octets: 64 rx_multicast: 1 rx_frames_below_65_octets: 1 rx_classified_drops: 1 tx_octets: 44630 tx_multicast: 387 tx_broadcast: 290 tx_frames_below_65_octets: 379 tx_frames_65_to_127_octets: 294 tx_frames_128_255_octets: 4 tx_green_prio_0: 298 tx_green_prio_7: 379 NIC statistics: tx_packets: 1 tx_bytes: 52 rx_packets: 713 rx_bytes: 34148 rx_octets: 46982 rx_multicast: 407 rx_broadcast: 306 rx_frames_below_65_octets: 399 rx_frames_65_to_127_octets: 310 rx_frames_128_to_255_octets: 4 rx_classified_drops: 399 rx_green_prio_0: 314 tx_octets: 64 tx_multicast: 1 tx_frames_below_65_octets: 1 tx_green_prio_7: 1 v1 > v2: reword commit messages v2 > v3: correctly mark this for net-next when sending v3 > v4: calloc array instead of zalloc per review v4 > v5: Apply CR suggestions for whitespace Fix calloc / zalloc mixup Properly destroy workqueues Add third commit to split long macros v5 > v6: Fix functionality - v5 was improperly tested Add bugfix for ethtool mutex lock Remove unnecessary ethtool stats reads v6 > v7: Remove mutex bug patch that was applied via net Rename function based on CR Add missed error check ==================== Signed-off-by: -
Colin Foster authored
Create and utilize bulk regmap reads instead of single access for gathering stats. The background reading of statistics happens frequently, and over a few contiguous memory regions. High speed PCIe buses and MMIO access will probably see negligible performance increase. Lower speed buses like SPI and I2C could see significant performance increase, since the bus configuration and register access times account for a large percentage of data transfer time. Signed-off-by:
Colin Foster <colin.foster@in-advantage.com> Reviewed-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Tested-by:
-
Colin Foster authored
Regmap supports bulk register reads. Ocelot does not. This patch adds support for Ocelot to invoke bulk regmap reads. That will allow any driver that performs consecutive reads over memory regions to optimize that access. Signed-off-by:
-
Colin Foster authored
In the ocelot.h file, several read / write macros were split across multiple lines, while others weren't. Split all macros that exceed the 80 character column width and match the style of the rest of the file. Signed-off-by:
-
Colin Foster authored
The ocelot_update_stats function only needs to read from one port, yet it was updating the stats for all ports. Update to only read the stats that are necessary. Signed-off-by:
-
David Ahern authored
Add reasons to __udp6_lib_rcv for skb drops. The only twist is that the NO_SOCKET takes precedence over the CSUM or other counters for that path (motivation behind this patch - csum counter was misleading). Signed-off-by:
David Ahern <dsahern@kernel.org> Signed-off-by:
-
David S. Miller authored
Joseph CHAMG says: ==================== ADD DM9051 ETHERNET DRIVER DM9051 is a spi interface chip, need cs/mosi/miso/clock with an interrupt gpio pin ==================== Signed-off-by: -
Joseph CHAMG authored
Add davicom dm9051 spi ethernet driver, The driver work for the device platform which has the spi master Signed-off-by:
Joseph CHAMG <josright123@gmail.com> Signed-off-by:
-
Joseph CHAMG authored
This is a new yaml base data file for configure davicom dm9051 with device tree Signed-off-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Tony Lu authored
The previous patch introduces a lock-free version of smc_tx_work() to solve unnecessary lock contention, which is expected to be held lock. So this adds comment to remind people to keep an eye out for locks. Suggested-by:
Stefan Raspl <raspl@linux.ibm.com> Signed-off-by:
Tony Lu <tonylu@linux.alibaba.com> Signed-off-by:
-
Kalash Nainwal authored
Generate RTM_NEWROUTE netlink notification when the route preference changes on an existing kernel generated default route in response to RA messages. Currently netlink notifications are generated only when this route is added or deleted but not when the route preference changes, which can cause userspace routing application state to go out of sync with kernel. Signed-off-by:
Kalash Nainwal <kalash@arista.com> Reviewed-by:
-
Davide Caratti authored
in current Linux, MTU policing does not take into account that packets at the TC ingress have the L2 header pulled. Thus, the same TC police action (with the same value of tcfp_mtu) behaves differently for ingress/egress. In addition, the full GSO size is compared to tcfp_mtu: as a consequence, the policer drops GSO packets even when individual segments have the L2 + L3 + L4 + payload length below the configured valued of tcfp_mtu. Improve the accuracy of MTU policing as follows: - account for mac_len for non-GSO packets at TC ingress. - compare MTU threshold with the segmented size for GSO packets. Also, add a kselftest that verifies the correct behavior. Signed-off-by:
Davide Caratti <dcaratti@redhat.com> Reviewed-by:
Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by:
-
- 13 Feb, 2022 10 commits
-
-
Kees Cook authored
With GCC 12, -Wstringop-overread was warning about an implicit cast from char[6] to char[8]. However, the extra 2 bytes are always thrown away, alignment doesn't matter, and the risk of hitting the edge of unallocated memory has been accepted, so this prototype can just be converted to a regular char *. Silences: net/core/dev.c: In function ‘bpf_prog_run_generic_xdp’: net/core/dev.c:4618:21: warning: ‘ether_addr_equal_64bits’ reading 8 bytes from a region of size 6 [-Wstringop-overread] 4618 | orig_host = ether_addr_equal_64bits(eth->h_dest, > skb->dev->dev_addr); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/dev.c:4618:21: note: referencing argument 1 of type ‘const u8[8]’ {aka ‘const unsigned char[8]’} net/core/dev.c:4618:21: note: referencing argument 2 of type ‘const u8[8]’ {aka ‘const unsigned char[8]’} In file included from net/core/dev.c:91: include/linux/etherdevice.h:375:20: note: in a call to function ‘ether_addr_equal_64bits’ 375 | static inline bool ether_addr_equal_64bits(const u8 addr1[6+2], | ^~~~~~~~~~~~~~~~~~~~~~~ Reported-by:Marc Kleine-Budde <mkl@pengutronix.de> Tested-by:
Kees Cook <keescook@chromium.org> Signed-off-by:
-
Horatiu Vultur authored
When CONFIG_IPV6 is not set, then the linking of the lan966x driver fails with the following error: drivers/net/ethernet/microchip/lan966x/lan966x_main.c:444: undefined reference to `ipv6_mc_check_mld' The fix consists in adding a check also for IS_ENABLED(CONFIG_IPV6) Fixes: 47aeea0d ("net: lan966x: Implement the callback SWITCHDEV_ATTR_ID_BRIDGE_MC_DISABLED") Reported-by:
kernel test robot <lkp@intel.com> Signed-off-by:
Horatiu Vultur <horatiu.vultur@microchip.com> Signed-off-by:
-
Horatiu Vultur authored
When CONFIG_PTP_1588_CLOCK is compiled as a module, then the linking of the lan966x fails because it can't find references to the following functions 'ptp_clock_index', 'ptp_clock_register' and 'ptp_clock_unregister' The fix consists in adding CONFIG_PTP_1588_CLOCK_OPTIONAL as a dependency for the driver. Fixes: d0964594 ("net: lan966x: Add support for ptp clocks") Reported-by:
-
David S. Miller authored
Raju Lakkaraju says: ==================== net: lan743x: PCI11010 / PCI11414 devices Enhancements This patch series adds support of the Ethernet function of the PCI11010 / PCI11414 devices to the LAN743x driver. The PCI1xxxx family of devices consists of a PCIe switch with a variety of embedded PCI endpoints on its downstream ports. The PCI11010 / PCI11414 devices include an Ethernet 10/100/1000/2500 function as one of those embedded endpoints. ==================== Signed-off-by: -
Raju Lakkaraju authored
Add support for Clause-45 MDIO PHY management Signed-off-by:
Raju Lakkaraju <Raju.Lakkaraju@microchip.com> Signed-off-by:
-
Raju Lakkaraju authored
This change facilitates the selection between SGMII and (R)GIII interfaces Signed-off-by:
-
Raju Lakkaraju authored
Increase MSI / MSI-X vectors supported from 8 to 16 and Interrupt De-assertion timers from 8 to 10 Signed-off-by:
-
Raju Lakkaraju authored
Add support for 4 Tx queues Signed-off-by:
-
Raju Lakkaraju authored
PCI11010/PCI11414 devices are enhancement of Ethernet LAN743x chip family. Signed-off-by:
-
M Chetan Kumar authored
This patch enables Intel M.2 7360 WWAN card support on IOSM Driver. Control path implementation is a reuse whereas data path implementation it uses a different protocol called as MUX Aggregation. The major portion of this patch covers the MUX Aggregation protocol implementation used for IP traffic communication. For M.2 7360 WWAN card, driver exposes 2 wwan AT ports for control communication. The user space application or the modem manager to use wwan AT port for data path establishment. During probe, driver reads the mux protocol device capability register to know the mux protocol version supported by device. Base on which the right mux protocol is initialized for data path communication. An overview of an Aggregation Protocol 1> An IP packet is encapsulated with 16 octet padding header to form a Datagram & the start offset of the Datagram is indexed into Datagram Header (DH). 2> Multiple such Datagrams are composed & the start offset of each DH is indexed into Datagram Table Header (DTH). 3> The Datagram Table (DT) is IP session specific & table_length item in DTH holds the number of composed datagram pertaining to that particular IP session. 4> And finally the offset of first DTH is indexed into DBH (Datagram Block Header). So in TX/RX flow Datagram Block (Datagram Block Header + Payload)is exchanged between driver & device. Signed-off-by:M Chetan Kumar <m.chetan.kumar@linux.intel.com> Signed-off-by:
-
- 12 Feb, 2022 1 commit
-
-
Jakub Kicinski authored
This reverts commit 038fcdaf. Christophe points out div64_u64() and do_div() have different calling conventions. One updates the param, the other returns the result. Reported-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Link: https://lore.kernel.org/all/056a7276-c6f0-cd7e-9e46-1d8507a0b6b1@wanadoo.fr/ Fixes: 038fcdaf ("net: ethernet: cavium: use div64_u64() instead of do_div()") Link: https://lore.kernel.org/r/20220211020544.3262694-1-kuba@kernel.orgSigned-off-by:
-
- 11 Feb, 2022 12 commits
-
-
Julia Lawall authored
Platform_driver probe functions aren't called with locks held and thus don't need GFP_ATOMIC. Use GFP_KERNEL instead. Problem found with Coccinelle. Signed-off-by:
Julia Lawall <Julia.Lawall@inria.fr> Link: https://lore.kernel.org/r/20220210204223.104181-1-Julia.Lawall@inria.frSigned-off-by:
-
Hariprasad Kelam authored
This patch fixes below error by using proper data type. drivers/net/ethernet/marvell/octeontx2/af/rpm.c: In function 'rpm_cfg_pfc_quanta_thresh': include/linux/find.h:40:23: error: array subscript 'long unsigned int[0]' is partly outside array bounds of 'u16[1]' {aka 'short unsigned int[1]'} [-Werror=array-bounds] 40 | val = *addr & GENMASK(size - 1, offset); Reported-by:Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
Hariprasad Kelam <hkelam@marvell.com> Link: https://lore.kernel.org/r/20220211155539.13931-1-hkelam@marvell.comSigned-off-by:
-
David S. Miller authored
Merge tag 'wireless-next-2022-02-11' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next wireless-next patches for v5.18 First set of patches for v5.18, with both wireless and stack patches. rtw89 now has AP mode support and wcn36xx has survey support. But otherwise pretty normal. Major changes: ath11k * add LDPC FEC type in 802.11 radiotap header * enable RX PPDU stats in monitor co-exist mode wcn36xx * implement survey reporting brcmfmac * add CYW43570 PCIE device rtw88 * rtw8821c: enable RFE 6 devices rtw89 * AP mode support mt76 * mt7916 support * background radar detection support
-
David S. Miller authored
Eric Dumazet says: ==================== ipv6: remove addrconf reliance on loopback Second patch in this series removes IPv6 requirement about the netns loopback device being the last device being dismantled. This was needed because rt6_uncached_list_flush_dev() and ip6_dst_ifdown() had to switch dst dev to a known device (loopback). Instead of loopback, we can use the (hidden) blackhole_netdev which is also always there. This will allow future simplfications of netdev_run_to() and other parts of the stack like default_device_exit_batch(). Last two patches are optimizations for both IP families. ==================== Signed-off-by: -
Eric Dumazet authored
This is an optimization to keep the per-cpu lists as short as possible: Whenever rt_flush_dev() changes one rtable dst.dev matching the disappearing device, it can can transfer the object to a quarantine list, waiting for a final rt_del_uncached_list(). Signed-off-by:
-
Eric Dumazet authored
This is an optimization to keep the per-cpu lists as short as possible: Whenever rt6_uncached_list_flush_dev() changes one rt6_info matching the disappearing device, it can can transfer the object to a quarantine list, waiting for a final rt6_uncached_list_del(). Signed-off-by:
-
Eric Dumazet authored
IPv6 addrconf notifiers wants the loopback device to be the last device being dismantled at netns deletion. This caused many limitations and work arounds. Back in linux-5.3, Mahesh added a per host blackhole_netdev that can be used whenever we need to make sure objects no longer refer to a disappearing device. If we attach to blackhole_netdev an ip6_ptr (allocate an idev), then we can use this special device (which is never freed) in place of the loopback_dev (which can be freed). This will permit improvements in netdev_run_todo() and other parts of the stack where had steps to make sure loopback_dev was the last device to disappear. Signed-off-by:
-
Eric Dumazet authored
This counter has never been visible, there is little point trying to maintain it. Signed-off-by:
-
Holger Brunck authored
The mv88e6352, mv88e6240 and mv88e6176 have a serdes interface. This patch allows to configure the output swing to a desired value in the phy-handle of the port. The value which is peak to peak has to be specified in microvolts. As the chips only supports eight dedicated values we return EINVAL if the value in the DTS does not match one of these values. Signed-off-by:
Holger Brunck <holger.brunck@hitachienergy.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Marek Behún <kabel@kernel.org> Signed-off-by:
-
Marek Behún authored
Common PHYs and network PCSes often have the possibility to specify peak-to-peak voltage on the differential pair - the default voltage sometimes needs to be changed for a particular board. Add properties `tx-p2p-microvolt` and `tx-p2p-microvolt-names` for this purpose. The second property is needed to specify the mode for the corresponding voltage in the `tx-p2p-microvolt` property, if the voltage is to be used only for speficic mode. More voltage-mode pairs can be specified. Example usage with only one voltage (it will be used for all supported PHY modes, the `tx-p2p-microvolt-names` property is not needed in this case): tx-p2p-microvolt = <915000>; Example usage with voltages for multiple modes: tx-p2p-microvolt = <915000>, <1100000>, <1200000>; tx-p2p-microvolt-names = "2500base-x", "usb", "pcie"; Add these properties into a separate file phy/transmit-amplitude.yaml, which should be referenced by any binding that uses it. Signed-off-by:
-
Guillaume Nault authored
The ->rtm_tos option is normally used to route packets based on both the destination address and the DS field. However it's ignored for IPv6 routes. Setting ->rtm_tos for IPv6 is thus invalid as the route is going to work only on the destination address anyway, so it won't behave as specified. Suggested-by:
Guillaume Nault <gnault@redhat.com> Reviewed-by:
Shuah Khan <skhan@linuxfoundation.org> Signed-off-by:
-
David S. Miller authored
Vladimir Oltean says: ==================== More aggressive DSA cleanup This series deletes some code which is apparently not needed. I've had these patches in my tree for a while, and testing on my boards didn't reveal any issues. Compared to the RFC v1 series, the only change is the addition of patch 3. https://patchwork.kernel.org/project/netdevbpf/cover/20220107184842.550334-1-vladimir.oltean@nxp.com/ ==================== Signed-off-by:
-
