- 19 Jan, 2020 20 commits
-
-
Amit Cohen authored
Add a trap for NVE packets that the device decided to drop because their overlay source MAC is multicast. Signed-off-by:
Amit Cohen <amitc@mellanox.com> Acked-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
Ido Schimmel <idosch@mellanox.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Amit Cohen authored
Add packet trap that can report NVE packets that the device decided to drop because their overlay source MAC is multicast. Signed-off-by:
-
Amit Cohen authored
Test that the trap is triggered under the right conditions and that devlink counters increase. Signed-off-by:
-
Amit Cohen authored
Test that the trap is triggered under the right conditions and that devlink counters increase. Signed-off-by:
-
Amit Cohen authored
Add the trap IDs and trap group used to report tunnel drops. Register tunnel packet traps and associated tunnel trap group with devlink during driver initialization. Signed-off-by:
-
Amit Cohen authored
Add packet traps that can report packets that were dropped during tunnel decapsulation. Signed-off-by:
-
Amit Cohen authored
Move L3_DROPS case to appear after L2_DROPS case. Signed-off-by:
-
Amit Cohen authored
Initialize ECN mapping registers during router init according to INET_ECN_encapsulate() and INET_ECN_decapsulate(). For IP-in-IP encapsulation, this is required to ensure that ECN bits in the underlay are set in accordance with the kernel. For decapsulation, this is required to ensure that packets with invalid ECN combination in underlay and overlay are trapped to the kernel and not forwarded. Signed-off-by:
-
Amit Cohen authored
This register configures the actions that are done during IPinIP decapsulation based on the ECN bits. Signed-off-by:
-
Amit Cohen authored
This register performs mapping from overlay ECN to underlay ECN during IPinIP encapsulation. Signed-off-by:
-
Amit Cohen authored
Add a trap for packets that the device decided to drop because they are not supposed to be routed. For example, IGMP queries can be flooded by the device in layer 2 and reach the router. Such packets should not be routed and instead dropped. Signed-off-by:
-
Amit Cohen authored
Add packet trap that can report packets that reached the router, but are non-routable. For example, IGMP queries can be flooded by the device in layer 2 and reach the router. Such packets should not be routed and instead dropped. Signed-off-by:
-
Amit Cohen authored
Add test cases to check that packets routed through disabled RIFs and packets routed from disabled RIFs are dropped and devlink counters increase when the action is trap. Signed-off-by:
-
Amit Cohen authored
IRIF_DISABLED and ERIF_DISABLED are driver specific traps. Packets are dropped for these reasons when they need to be routed through/from existing router interfaces (RIF) which are disabled. Add devlink driver-specific traps and mlxsw trap IDs used to report these traps. Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linuxDavid S. Miller authored
Saeed Mahameed says: ==================== Mellanox, mlx5 E-Switch chains and prios This series has two parts, 1) A merge commit with mlx5-next branch that include updates for mlx5 HW layouts needed for this and upcoming submissions. 2) From Paul, Increase the number of chains and prios Currently the Mellanox driver supports offloading tc rules that are defined on the first 4 chains and the first 16 priorities. The restriction stems from the firmware flow level enforcement requiring a flow table of a certain level to point to a flow table of a higher level. This limitation may be ignored by setting the ignore_flow_level bit when creating flow table entries. Use unmanaged tables and ignore flow level to create more tables than declared by fs_core steering. Manually manage the connections between the tables themselves. HW table is instantiated for every tc <chain,prio> tuple. The miss rule of every table either jumps to the next <chain,prio> table, or continues to slow_fdb. This logic is realized by following this sequence: 1. Create an auto-grouped flow table for the specified priority with reserved entries Reserved entries are allocated at the end of the flow table. Flow groups are evaluated in sequence and therefore it is guaranteed that the flow group defined on the last FTEs will be the last to evaluate. Define a "match all" flow group on the reserved entries, providing the platform to add table miss actions. 2. Set the miss rule action to jump to the next <chain,prio> table or the slow_fdb. 3. Link the previous priority table to point to the new table by updating its miss rule. Please pull and let me know if there's any problem. ==================== Signed-off-by: -
Dejin Zheng authored
found a warning by the following command: ./scripts/checkpatch.pl -f drivers/net/phy/adin.c WARNING: msleep < 20ms can sleep for up to 20ms; see Documentation/timers/timers-howto.rst #628: FILE: drivers/net/phy/adin.c:628: + msleep(10); Signed-off-by:
Dejin Zheng <zhengdejin5@gmail.com> Signed-off-by:
-
David S. Miller authored
Vladimir Oltean says: ==================== Rate adaptation for Felix DSA switch When operating the MAC at 2.5Gbps (2500Base-X and USXGMII/QSXGMII) and in combination with certain PHYs, it is possible that the copper side may operate at lower link speeds. In this case, it is the PHY who has a MAC inside of it that emits pause frames towards the switch's MAC, telling it to slow down so that the transmission is lossless. These patches are the support needed for the switch side of things to work. ==================== Signed-off-by: -
Alex Marginean authored
If the serdes link is set to 2500 using interfce type 2500base-X, lower link speeds over on the line side should still be supported. Rate adaptation is done out of band, in our case using AQR PHYs this is done using flow control. Signed-off-by:
Alex Marginean <alexandru.marginean@nxp.com> Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by:
-
Alex Marginean authored
Flow control is used with 2500Base-X and AQR PHYs to do rate adaptation between line side 100/1000 links and MAC running at 2.5G. This is independent of the flow control configuration settled on line side though AN. In general, allowing the MAC to handle flow control even if not negotiated with the link partner should not be a problem, so the patch just enables it in all cases. Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-nextDavid S. Miller authored
Pablo Neira Ayuso says: ==================== Netfilter updates for net-next The following patchset contains Netfilter updates for net-next, they are: 1) Incorrect uapi header comment in bitwise, from Jeremy Sowden. 2) Fetch flow statistics if flow is still active. 3) Restrict flow matching on hardware based on input device. 4) Add nf_flow_offload_work_alloc() helper function. 5) Remove the last client of the FLOW_OFFLOAD_DYING flag, use teardown instead. 6) Use atomic bitwise operation to operate with flow flags. 7) Add nf_flowtable_hw_offload() helper function to check for the NF_FLOWTABLE_HW_OFFLOAD flag. 8) Add NF_FLOW_HW_REFRESH to retry hardware offload from the flowtable software datapath. 9) Remove indirect calls in xt_hashlimit, from Florian Westphal. 10) Add nf_flow_offload_tuple() helper to consolidate code. 11) Add nf_flow_table_offload_cmd() helper function. 12) A few whitespace cleanups in nf_tables in bitwise and the bitmap/hash set types, from Jeremy Sowden. 13) Cleanup netlink attribute checks in bitwise, from Jeremy Sowden. 14) Replace goto by return in error path of nft_bitwise_dump(), from Jeremy Sowden. 15) Add bitwise operation netlink attribute, also from Jeremy. 16) Add nft_bitwise_init_bool(), from Jeremy Sowden. 17) Add nft_bitwise_eval_bool(), also from Jeremy. 18) Add nft_bitwise_dump_bool(), from Jeremy Sowden. 19) Disallow hardware offload for other that NFT_BITWISE_BOOL, from Jeremy Sowden. 20) Add NFTA_BITWISE_DATA netlink attribute, again from Jeremy. 21) Add support for bitwise shift operation, from Jeremy Sowden. ==================== Signed-off-by:
-
- 18 Jan, 2020 1 commit
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queueDavid S. Miller authored
Jeff Kirsher says: ==================== Intel Wired LAN Driver Updates 2020-01-17 This series contains updates to igc, i40e, fm10k and ice drivers. Sasha fixes a typo in a code comment that referred to silicon that is not supported in the igc driver. Cleaned up a defined that was not being used. Added support for another i225 SKU which does not have an NVM. Added support for TCP segmentation offload (TSO) into igc. Added support for PHY power management control to provide a reliable and accurate indication of PHY reset completion. Jake adds support for the new txqueue parameter to the transmit timeout function in fm10k which reduces the code complexity when determining which transmit queue is stuck. Julio Faracco makes the similar changes that Jake did for fm10k, for i40e and ice drivers. Added support for the new txqueue parameter in the transmit timeout functions for i40e and ice. Colin Ian King cleans up a redundant initialization of a local variable. ==================== Signed-off-by:
-
- 17 Jan, 2020 19 commits
-
-
Colin Ian King authored
The variable xmit_done is being initialized with a value that is never read and it is being updated later with a new value. The initialization is redundant and can be removed. Addresses-Coverity: ("Unused value") Signed-off-by:Colin Ian King <colin.king@canonical.com> Tested-by:
Andrew Bowers <andrewx.bowers@intel.com> Signed-off-by:
Jeff Kirsher <jeffrey.t.kirsher@intel.com>
-
Julio Faracco authored
The scope of function .ndo_tx_timeout was changed to include the hang queue when a TX timeout event occurs. See commit 0290bd29 ("netdev: pass the stuck queue to the timeout handler") for more details. Now, drivers don't need to identify which queue is stopped. Drivers can simply use the queue index provided by dev_watchdog and execute all actions needed to restore network traffic. This commit do some cleanups into Intel ice driver to remove a redundant loop to find stopped queue. Signed-off-by:
Julio Faracco <jcfaracco@gmail.com> Tested-by:
-
Julio Faracco authored
The scope of function .ndo_tx_timeout was changed to include the hang queue when a TX timeout event occurs. See commit 0290bd29 ("netdev: pass the stuck queue to the timeout handler") for more details. Now, drivers don't need to identify which queue is stopped. Drivers can simply use the queue index provided by dev_watchdog and execute all actions needed to restore network traffic. This commit do some cleanups into Intel i40e driver to remove a redundant loop to find stopped queue. Signed-off-by:
-
Jacob Keller authored
Make use of the new txqueue parameter to the .ndo_tx_timeout function. In fm10k_tx_timeout, remove the now unnecessary loop to determine which Tx queue is stuck. Instead, just double check the specified queue This could be improved further to attempt resetting only the specific queue that got stuck. However, that is a much larger refactor and has been left as a future improvement. Signed-off-by:
Jacob Keller <jacob.e.keller@intel.com> Tested-by:
-
Sasha Neftin authored
PHY power management control should provide a reliable and accurate indication of PHY reset completion and decrease the delay time after a PHY reset Signed-off-by:
Sasha Neftin <sasha.neftin@intel.com> Tested-by:
Aaron Brown <aaron.f.brown@intel.com> Signed-off-by:
-
Sasha Neftin authored
TCP segmentation offload allows a device to segment a single frame into multiple frames with a data payload size specified in socket buffer. As a result we can now send data approximately up to seven percents fast than was previously possible on my system. Signed-off-by:
-
Sasha Neftin authored
Add support for blank NVM SKU Signed-off-by:
-
Sasha Neftin authored
Remove the unused IGC_FUNC_0 definition and make the code cleaner Signed-off-by:
-
Sasha Neftin authored
Fix typo in a context descriptor comment Signed-off-by:
-
Guillaume Nault authored
Mark function parameters as 'const' where possible. Signed-off-by:
Guillaume Nault <gnault@redhat.com> Acked-by:
Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by:
-
Vladimir Oltean authored
The felix_parse_ports_node function was tested only on device trees where all ports were enabled. Fix this check so that the driver continues to probe only with the ports where status is not "disabled", as expected. Fixes: bdeced75 ("net: dsa: felix: Add PCS operations for PHYLINK") Signed-off-by:
Florian Fainelli <f.fainelli@gmail.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
Alex Marginean authored
Some PHYs like VSC8234 don't like it when AN restarts on their system side and they restart line side AN too, going into an endless link up/down loop. Don't restart PCS AN if link is up already. Although in theory this feedback loop should be possible with the other in-band AN modes too, for some reason it was not seen with the VSC8514 QSGMII and AQR412 USXGMII PHYs. So keep this logic only for SGMII where the problem was found. Fixes: bdeced75 ("net: dsa: felix: Add PCS operations for PHYLINK") Suggested-by:
-
Alex Marginean authored
At least some PHYs (AQR412) don't advertise copper-side link status during system side AN. So remove this duplicate assignment to pcs->link and rely on the previous one for link state: the local indication from the MAC PCS. Fixes: bdeced75 ("net: dsa: felix: Add PCS operations for PHYLINK") Signed-off-by:
-
Ido Schimmel authored
The driver is named "mlxsw", not "mlx5". Fixes: d4255d75 ("devlink: document info versions for each driver") Signed-off-by:
-
Vladimir Oltean authored
It is not an error to unplug a cable from the ENETC port even with TSN offloads, so don't spam the log with link-related messages from the tc-taprio offload subsystem, a single notification is sufficient: [10972.351859] fsl_enetc 0000:00:00.0 eno0: Qbv PSPEED set speed link down. [10972.360241] fsl_enetc 0000:00:00.0 eno0: Link is Down Fixes: 2e47cb41 ("enetc: update TSN Qbv PSPEED set according to adjust link speed") Signed-off-by:
-
Alexandru Ardelean authored
Some bits of static data should have been made const from the start. This change adds the const qualifier where appropriate. Reviewed-by:
Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by:
-
Hongbo Yao authored
If CONFIG_INET is not set and CONFIG_NETDEVSIM=y. Building drivers/net/netdevsim/fib.o will get the following error: drivers/net/netdevsim/fib.o: In function `nsim_fib4_rt_hw_flags_set': fib.c:(.text+0x12b): undefined reference to `fib_alias_hw_flags_set' drivers/net/netdevsim/fib.o: In function `nsim_fib4_rt_destroy': fib.c:(.text+0xb11): undefined reference to `free_fib_info' Correct the Kconfig for netdevsim. Reported-by:
Hulk Robot <hulkci@huawei.com> Fixes: 48bb9eb4 ("netdevsim: fib: Add dummy implementation for FIB offload") Signed-off-by:
Hongbo Yao <yaohongbo@huawei.com> Acked-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Florian Fainelli authored
We maintain global statistics for an entire MDIO bus, as well as broken down, per MDIO bus address statistics. Given that it is possible for MDIO devices such as switches to access MDIO bus addresses for which there is not a mdio_device instance created (therefore not a a corresponding device directory in sysfs either), we also maintain per-address statistics under the statistics folder. The layout looks like this: /sys/class/mdio_bus/../statistics/ transfers errrors writes reads transfers_<addr> errors_<addr> writes_<addr> reads_<addr> When a mdio_device instance is registered, a statistics/ folder is created with the tranfers, errors, writes and reads attributes which point to the appropriate MDIO bus statistics structure. Statistics are 64-bit unsigned quantities and maintained through the u64_stats_sync.h helper functions. Signed-off-by:
-
Eric Dumazet authored
It seems nsim_fib6_rt_create() intent was to return either a valid pointer or an embedded error code. BUG: unable to handle page fault for address: fffffffffffffff4 PGD 9870067 P4D 9870067 PUD 9872067 PMD 0 Oops: 0000 [#1] PREEMPT SMP KASAN CPU: 0 PID: 22851 Comm: syz-executor.1 Not tainted 5.5.0-rc5-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:jhash2 include/linux/jhash.h:125 [inline] RIP: 0010:rhashtable_jhash2+0x76/0x2c0 lib/rhashtable.c:963 Code: b9 00 00 00 00 00 fc ff df 48 c1 e8 03 0f b6 14 08 4c 89 f0 83 e0 07 83 c0 03 38 d0 7c 08 84 d2 0f 85 30 02 00 00 49 8d 7e 04 <41> 8b 06 48 be 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 0f b6 RSP: 0018:ffffc90016127190 EFLAGS: 00010246 RAX: 0000000000000007 RBX: 00000000dfb3ab49 RCX: dffffc0000000000 RDX: 0000000000000000 RSI: ffffffff839ba7c8 RDI: fffffffffffffff8 RBP: ffffc900161271c0 R08: ffff8880951f8640 R09: ffffed1015d0703d R10: ffffed1015d0703c R11: ffff8880ae8381e3 R12: 00000000dfb3ab49 R13: 00000000dfb3ab49 R14: fffffffffffffff4 R15: 0000000000000007 FS: 00007f40bfbc6700(0000) GS:ffff8880ae800000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: fffffffffffffff4 CR3: 0000000093660000 CR4: 00000000001406f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rht_key_get_hash include/linux/rhashtable.h:133 [inline] rht_key_hashfn include/linux/rhashtable.h:159 [inline] rht_head_hashfn include/linux/rhashtable.h:174 [inline] __rhashtable_insert_fast.constprop.0+0xe15/0x1180 include/linux/rhashtable.h:723 rhashtable_insert_fast include/linux/rhashtable.h:832 [inline] nsim_fib6_rt_add drivers/net/netdevsim/fib.c:603 [inline] nsim_fib6_rt_insert drivers/net/netdevsim/fib.c:658 [inline] nsim_fib6_event drivers/net/netdevsim/fib.c:719 [inline] nsim_fib_event drivers/net/netdevsim/fib.c:744 [inline] nsim_fib_event_nb+0x1b16/0x2600 drivers/net/netdevsim/fib.c:772 notifier_call_chain+0xc2/0x230 kernel/notifier.c:83 __atomic_notifier_call_chain+0xa6/0x1a0 kernel/notifier.c:173 atomic_notifier_call_chain+0x2e/0x40 kernel/notifier.c:183 call_fib_notifiers+0x173/0x2a0 net/core/fib_notifier.c:35 call_fib6_notifiers+0x4b/0x60 net/ipv6/fib6_notifier.c:22 call_fib6_entry_notifiers+0xfb/0x150 net/ipv6/ip6_fib.c:399 fib6_add_rt2node net/ipv6/ip6_fib.c:1216 [inline] fib6_add+0x20cd/0x3ec0 net/ipv6/ip6_fib.c:1471 __ip6_ins_rt+0x54/0x80 net/ipv6/route.c:1315 ip6_ins_rt+0x96/0xd0 net/ipv6/route.c:1325 __ipv6_dev_ac_inc+0x76f/0xb20 net/ipv6/anycast.c:324 ipv6_sock_ac_join+0x4c1/0x790 net/ipv6/anycast.c:139 do_ipv6_setsockopt.isra.0+0x3908/0x4290 net/ipv6/ipv6_sockglue.c:670 ipv6_setsockopt+0xff/0x180 net/ipv6/ipv6_sockglue.c:944 udpv6_setsockopt+0x68/0xb0 net/ipv6/udp.c:1564 sock_common_setsockopt+0x94/0xd0 net/core/sock.c:3149 __sys_setsockopt+0x261/0x4c0 net/socket.c:2130 __do_sys_setsockopt net/socket.c:2146 [inline] __se_sys_setsockopt net/socket.c:2143 [inline] __x64_sys_setsockopt+0xbe/0x150 net/socket.c:2143 do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x45aff9 Fixes: 48bb9eb4 ("netdevsim: fib: Add dummy implementation for FIB offload") Signed-off-by:
Eric Dumazet <edumazet@google.com> Reported-by:
syzbot <syzkaller@googlegroups.com> Cc: Ido Schimmel <idosch@mellanox.com> Reviewed-by:
-
