- 10 Oct, 2019 1 commit
-
-
Xin Long authored
A helper sctp_ulpevent_nofity_peer_addr_change() will be extracted to make peer_addr_change event and enqueue it, and the helper will be called in sctp_assoc_add_peer() to send SCTP_ADDR_ADDED event. This event is described in rfc6458#section-6.1.2: SCTP_ADDR_ADDED: The address is now part of the association. Signed-off-by:
Xin Long <lucien.xin@gmail.com> Acked-by:
Neil Horman <nhorman@tuxdriver.com> Signed-off-by:
Jakub Kicinski <jakub.kicinski@netronome.com>
-
- 09 Oct, 2019 12 commits
-
-
Antonio Borneo authored
All the registers and the functionalities used in the callback dwmac5_flex_pps_config() are common between dwmac 4.10a [1] and 5.00a [2]. Reuse the same callback for dwmac 4.10a too. Tested on STM32MP15x, based on dwmac 4.10a. [1] DWC Ethernet QoS Databook 4.10a October 2014 [2] DWC Ethernet QoS Databook 5.00a September 2017 Signed-off-by:
Antonio Borneo <antonio.borneo@st.com> Signed-off-by:
-
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spiJakub Kicinski authored
Pull in a dependency for Vladimir's work on more precise packet time stamping. Mark Brown says: ==================== spi: Add a PTP API For detailed timestamping of operations. ==================== Signed-off-by: -
Eric Dumazet authored
This reverts commit 0ad646c8. As noticed by Jakub, this is no longer needed after commit 11fc7d5a ("tun: fix memory leak in error path") This no longer exports dev_get_valid_name() for the exclusive use of tun driver. Suggested-by:
Eric Dumazet <edumazet@google.com> Signed-off-by:
-
Jiri Pirko authored
__tipc_nl_compat_dumpit() calls tipc_nl_publ_dump() which expects the attrs to be available by genl_dumpit_info(cb)->attrs. Add info struct and attr parsing in compat dumpit function. Reported-by: syzbot+8d37c50ffb0f52941a5e@syzkaller.appspotmail.com Fixes: 057af707 ("net: tipc: have genetlink code to parse the attrs during dumpit") Signed-off-by:
Jiri Pirko <jiri@mellanox.com> Acked-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
-
Jiri Pirko authored
Individual dumpit ops (start, dumpit, done) are locked by genl_lock if !family->parallel_ops. However, multiple genl_family_rcv_msg_dumpit() calls may in in flight in parallel. Each has a separate struct genl_dumpit_info allocated but they share the same family->attrbuf. Fix this by allocating separate memory for attrs for dumpit ops, for non-parallel_ops (for parallel_ops it is done already). Reported-by: syzbot+495688b736534bb6c6ad@syzkaller.appspotmail.com Reported-by: syzbot+ff59dc711f2cff879a05@syzkaller.appspotmail.com Reported-by: syzbot+dbe02e13bcce52bcf182@syzkaller.appspotmail.com Reported-by: syzbot+9cb7edb2906ea1e83006@syzkaller.appspotmail.com Fixes: bf813b0a ("net: genetlink: parse attrs and store in contect info struct during dumpit") Signed-off-by:
-
Jakub Kicinski authored
Huazhong Tan says: ==================== This patch-set includes some new features for the HNS3 ethernet controller driver. [patch 01/06] adds support for configuring VF link status on the host. [patch 02/06] adds support for configuring VF spoof check. [patch 03/06] adds support for configuring VF trust. [patch 04/06] adds support for configuring VF bandwidth on the host. [patch 05/06] adds support for configuring VF MAC on the host. [patch 06/06] adds support for tx-scatter-gather-fraglist. ==================== Signed-off-by: -
Yunsheng Lin authored
The hardware supports up to 8 TX BD for non-tso skb and up to 63 TX BD for TSO skb. Currently, the hns3 driver supports RX skb with fraglist when HW GRO is enabled, when the stack forwards a RX skb with fraglist, the stack need to linearize the skb before sending to other interface without TX fraglist support. This patch adds support for TX fraglist. The performance increases from 1 GByte to 1.5 GByte for one iperf TCP stream during forwarding test after this patch. BTW, the minimum BD number of ring should be updated to 72 for supporting TX fraglist. This patch also changes the error handling of some function that called by hns3_fill_desc, which returns BD num when there is no error, change some macro to more meaningful name. Signed-off-by:
Yunsheng Lin <linyunsheng@huawei.com> Signed-off-by:
Huazhong Tan <tanhuazhong@huawei.com> Signed-off-by:
-
Huazhong Tan authored
This patch adds support of configuring VF MAC from the host for the HNS3 driver. BTW, the parameter init in the hns3_init_mac_addr is unnecessary now, since the MAC address will not read from NCL_CONFIG when doing reset, so it should be removed, otherwise it will affect VF's MAC address initialization. Signed-off-by:
-
Yonglong Liu authored
This patch adds support for configuring bandwidth of VF on the host for HNS3 drivers. Signed-off-by:
Yonglong Liu <liuyonglong@huawei.com> Signed-off-by:
-
Jian Shen authored
This patch adds supports for setting VF trust by host. If specified VF is trusted, then it can enable promisc(include allmulti mode). If a trusted VF enabled promisc, and being untrusted, host will disable promisc mode for this VF. For VF will update its promisc mode from set_rx_mode now, so it's unnecessary to set broadcst promisc mode when initialization or reset. Signed-off-by:
Jian Shen <shenjian15@huawei.com> Signed-off-by:
-
Jian Shen authored
This patch adds support for spoof check configuration for VFs. When it is enabled, "spoof checking" is done for both mac address and VLAN. For each VF, the HW ensures that the source MAC address (or VLAN) of every outgoing packet exists in the MAC-list (or VLAN-list) configured for RX filtering for that VF. If not, the packet is dropped. Signed-off-by:
-
Yufeng Mo authored
This patch adds support to configure VF link properties. The options are auto, enable, and disable. Even if the PF is down, the communication between VFs will be normal if the VFs are set to enable. The commands are as follows: 'ip link set <pf> vf <vf_id> state <auto|enable|disable>' change the VF status 'ip link show' show the setting status Signed-off-by:
Yufeng Mo <moyufeng@huawei.com> Signed-off-by:
-
- 08 Oct, 2019 5 commits
-
-
Eric Dumazet authored
syzbot reported a warning [1] that triggered after recent Jiri patch. This exposes a bug that we hit already in the past (see commit ff244c6b ("tun: handle register_netdevice() failures properly") for details) tun uses priv->destructor without an ndo_init() method. register_netdevice() can return an error, but will not call priv->destructor() in some cases. Jiri recent patch added one more. A long term fix would be to transfer the initialization of what we destroy in ->destructor() in the ndo_init() This looks a bit risky given the complexity of tun driver. A simpler fix is to detect after the failed register_netdevice() if the tun_free_netdev() function was called already. [1] ODEBUG: free active (active state 0) object type: timer_list hint: tun_flow_cleanup+0x0/0x280 drivers/net/tun.c:457 WARNING: CPU: 0 PID: 8653 at lib/debugobjects.c:481 debug_print_object+0x168/0x250 lib/debugobjects.c:481 Kernel panic - not syncing: panic_on_warn set ... CPU: 0 PID: 8653 Comm: syz-executor976 Not tainted 5.4.0-rc1-next-20191004 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x172/0x1f0 lib/dump_stack.c:113 panic+0x2dc/0x755 kernel/panic.c:220 __warn.cold+0x2f/0x3c kernel/panic.c:581 report_bug+0x289/0x300 lib/bug.c:195 fixup_bug arch/x86/kernel/traps.c:174 [inline] fixup_bug arch/x86/kernel/traps.c:169 [inline] do_error_trap+0x11b/0x200 arch/x86/kernel/traps.c:267 do_invalid_op+0x37/0x50 arch/x86/kernel/traps.c:286 invalid_op+0x23/0x30 arch/x86/entry/entry_64.S:1028 RIP: 0010:debug_print_object+0x168/0x250 lib/debugobjects.c:481 Code: dd 80 b9 e6 87 48 89 fa 48 c1 ea 03 80 3c 02 00 0f 85 b5 00 00 00 48 8b 14 dd 80 b9 e6 87 48 c7 c7 e0 ae e6 87 e8 80 84 ff fd <0f> 0b 83 05 e3 ee 80 06 01 48 83 c4 20 5b 41 5c 41 5d 41 5e 5d c3 RSP: 0018:ffff888095997a28 EFLAGS: 00010082 RAX: 0000000000000000 RBX: 0000000000000003 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff815cb526 RDI: ffffed1012b32f37 RBP: ffff888095997a68 R08: ffff8880a92ac580 R09: ffffed1015d04101 R10: ffffed1015d04100 R11: ffff8880ae820807 R12: 0000000000000001 R13: ffffffff88fb5340 R14: ffffffff81627110 R15: ffff8880aa41eab8 __debug_check_no_obj_freed lib/debugobjects.c:963 [inline] debug_check_no_obj_freed+0x2d4/0x43f lib/debugobjects.c:994 kfree+0xf8/0x2c0 mm/slab.c:3755 kvfree+0x61/0x70 mm/util.c:593 netdev_freemem net/core/dev.c:9384 [inline] free_netdev+0x39d/0x450 net/core/dev.c:9533 tun_set_iff drivers/net/tun.c:2871 [inline] __tun_chr_ioctl+0x317b/0x3f30 drivers/net/tun.c:3075 tun_chr_ioctl+0x2b/0x40 drivers/net/tun.c:3355 vfs_ioctl fs/ioctl.c:47 [inline] file_ioctl fs/ioctl.c:539 [inline] do_vfs_ioctl+0xdb6/0x13e0 fs/ioctl.c:726 ksys_ioctl+0xab/0xd0 fs/ioctl.c:743 __do_sys_ioctl fs/ioctl.c:750 [inline] __se_sys_ioctl fs/ioctl.c:748 [inline] __x64_sys_ioctl+0x73/0xb0 fs/ioctl.c:748 do_syscall_64+0xfa/0x760 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x441439 Code: e8 9c ae 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 3b 0a fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fff61c37438 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 0000000000441439 RDX: 0000000020000400 RSI: 00000000400454ca RDI: 0000000000000004 RBP: 00007fff61c37470 R08: 0000000000000001 R09: 0000000100000000 R10: 0000000000000000 R11: 0000000000000246 R12: ffffffffffffffff R13: 0000000000000005 R14: 0000000000000000 R15: 0000000000000000 Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: ff927412 ("net: introduce name_node struct to be used in hashlist") Signed-off-by:
syzbot <syzkaller@googlegroups.com> Signed-off-by:
-
Colin Ian King authored
There is a spelling mistake in a NL_SET_ERR_MSG_MOD message. Fix it. Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Colin Ian King authored
Don't populate const arrays on the stack but instead make them static. Makes the object code smaller by 1058 bytes. Before: text data bss dec hex filename 29879 6144 0 36023 8cb7 drivers/net/phy/mscc.o After: text data bss dec hex filename 28437 6528 0 34965 8895 drivers/net/phy/mscc.o (gcc version 9.2.1, amd64) Signed-off-by:
-
Colin Ian King authored
Don't populate the array exp_mask on the stack but instead make it static. Makes the object code smaller by 224 bytes. Before: text data bss dec hex filename 77832 2290 0 80122 138fa ethernet/netronome/nfp/bpf/jit.o After: text data bss dec hex filename 77544 2354 0 79898 1381a ethernet/netronome/nfp/bpf/jit.o (gcc version 9.2.1, amd64) Signed-off-by:
-
Vladimir Oltean authored
SPI is one of the interfaces used to access devices which have a POSIX clock driver (real time clocks, 1588 timers etc). The fact that the SPI bus is slow is not what the main problem is, but rather the fact that drivers don't take a constant amount of time in transferring data over SPI. When there is a high delay in the readout of time, there will be uncertainty in the value that has been read out of the peripheral. When that delay is constant, the uncertainty can at least be approximated with a certain accuracy which is fine more often than not. Timing jitter occurs all over in the kernel code, and is mainly caused by having to let go of the CPU for various reasons such as preemption, servicing interrupts, going to sleep, etc. Another major reason is CPU dynamic frequency scaling. It turns out that the problem of retrieving time from a SPI peripheral with high accuracy can be solved by the use of "PTP system timestamping" - a mechanism to correlate the time when the device has snapshotted its internal time counter with the Linux system time at that same moment. This is sufficient for having a precise time measurement - it is not necessary for the whole SPI transfer to be transmitted "as fast as possible", or "as low-jitter as possible". The system has to be low-jitter for a very short amount of time to be effective. This patch introduces a PTP system timestamping mechanism in struct spi_transfer. This is to be used by SPI device drivers when they need to know the exact time at which the underlying device's time was snapshotted. More often than not, SPI peripherals have a very exact timing for when their SPI-to-interconnect bridge issues a transaction for snapshotting and reading the time register, and that will be dependent on when the SPI-to-interconnect bridge figures out that this is what it should do, aka as soon as it sees byte N of the SPI transfer. Since spi_device drivers are the ones who'd know best how the peripheral behaves in this regard, expose a mechanism in spi_transfer which allows them to specify which word (or word range) from the transfer should be timestamped. Add a default implementation of the PTP system timestamping in the SPI core. This is not going to be satisfactory performance-wise, but should at least increase the likelihood that SPI device drivers will use PTP system timestamping in the future. There are 3 entry points from the core towards the SPI controller drivers: - transfer_one: The driver is passed individual spi_transfers to execute. This is the easiest to timestamp. - transfer_one_message: The core passes the driver an entire spi_message (a potential batch of spi_transfers). The core puts the same pre and post timestamp to all transfers within a message. This is not ideal, but nothing better can be done by default anyway, since the core has no insight into how the driver batches the transfers. - transfer: Like transfer_one_message, but for unqueued drivers (i.e. the driver implements its own queue scheduling). Signed-off-by:
Vladimir Oltean <olteanv@gmail.com> Link: https://lore.kernel.org/r/20190905010114.26718-3-olteanv@gmail.comSigned-off-by:
Mark Brown <broonie@kernel.org>
-
- 07 Oct, 2019 22 commits
-
-
David S. Miller authored
Ioana Ciornei says: ==================== dpaa2-eth: misc cleanup This patch set consists of some cleanup patches ranging from removing dead code to fixing a minor issue in ethtool stats. Also, unbounded while loops are removed from the driver by adding a maximum number of retries for DPIO portal commands. Changes in v2: - return -ETIMEDOUT where possible if the number of retries is hit ==================== Signed-off-by:David S. Miller <davem@davemloft.net>
-
Ioana Radulescu authored
Throughout the driver there are several places where we wait indefinitely for DPIO portal commands to be executed, while the portal returns a busy response code. Even though in theory we are guaranteed the portals become available eventually, in practice the QBMan hardware module may become unresponsive in various corner cases. Make sure we can never get stuck in an infinite while loop by adding a retry counter for all portal commands. Signed-off-by:
Ioana Radulescu <ruxandra.radulescu@nxp.com> Signed-off-by:
Ioana Ciornei <ioana.ciornei@nxp.com> Signed-off-by:
-
Ioana Radulescu authored
Don't print error message for a successful return value. Fixes: d84c3a4d ("dpaa2-eth: Add new DPNI statistics counters") Signed-off-by:
-
Ioana Radulescu authored
Remove one function call whose result was not used anywhere. Signed-off-by:
-
Colin Ian King authored
Don't populate the array tick_array on the stack but instead make it static. Makes the object code smaller by 29 bytes. Before: text data bss dec hex filename 19191 432 0 19623 4ca7 hisilicon/hns3/hns3pf/hclge_tm.o After: text data bss dec hex filename 19098 496 0 19594 4c8a hisilicon/hns3/hns3pf/hclge_tm.o (gcc version 9.2.1, amd64) Signed-off-by:
-
Colin Ian King authored
Don't populate the arrays port_map and sl_map on the stack but instead make them static. Makes the object code smaller by 64 bytes. Before: text data bss dec hex filename 49575 6872 64 56511 dcbf hisilicon/hns/hns_dsaf_main.o After: text data bss dec hex filename 49350 7032 64 56446 dc7e hisilicon/hns/hns_dsaf_main.o (gcc version 9.2.1, amd64) Signed-off-by:
-
David S. Miller authored
Jakub Kicinski says: ==================== net/tls: minor micro optimizations This set brings a number of minor code changes from my tree which don't have a noticeable impact on performance but seem reasonable nonetheless. First sk_msg_sg copy array is converted to a bitmap, zeroing that structure takes a lot of time, hence we should try to keep it small. Next two conditions are marked as unlikely, GCC seemed to had little trouble correctly reasoning about those. Patch 4 adds parameters to tls_device_decrypted() to avoid walking structures, as all callers already have the relevant pointers. Lastly two boolean members of TLS context structures are converted to a bitfield. ==================== Signed-off-by: -
Jakub Kicinski authored
Use a single bit instead of boolean to remember if packet was already decrypted. Signed-off-by:
Simon Horman <simon.horman@netronome.com> Signed-off-by:
-
Jakub Kicinski authored
Store async_capable on a single bit instead of a full integer to save space. Signed-off-by:
-
Jakub Kicinski authored
Avoid unnecessary pointer chasing and calculations, callers already have most of the state tls_device_decrypted() needs. Signed-off-by:
Dirk van der Merwe <dirk.vandermerwe@netronome.com> Signed-off-by:
-
Jakub Kicinski authored
Make sure GCC realizes it's unlikely that allocations will fail. Signed-off-by:
-
Jakub Kicinski authored
Tell GCC sk->err is not likely to be set. Signed-off-by:
-
Jakub Kicinski authored
Don't use bool array in struct sk_msg_sg, save 12 bytes. Signed-off-by:
-
Heiner Kallweit authored
Use helper skb_ensure_writable in two more places to simplify the code. Signed-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by:
-
David S. Miller authored
Consistent with how pskb_may_pull() also now does so. Signed-off-by: -
Heiner Kallweit authored
This function de-facto returns a bool, so let's change the return type accordingly. Signed-off-by:
-
David S. Miller authored
Sameeh Jubran says: ==================== ena: Support ethtool set_channels Difference from v2: * ethtool's set/get channels: Switched to using combined instead of separate rx/tx * Fixed error handling in set_channels * Fixed indentation and cosmetic issues as requested by Jakub Kicinski Difference from v1: * Dropped the print from patch 0002 - "net: ena: multiple queue creation related cleanups" as requested by David Miller ==================== Signed-off-by: -
Sameeh Jubran authored
Set channels callback enables the user to change the count of queues used by the driver using ethtool. We decided to currently support only equal number of rx and tx queues, this might change in the future. Also rename dev_up to dev_was_up in ena_update_queue_count() to make it clearer. Signed-off-by:
Sameeh Jubran <sameehj@amazon.com> Signed-off-by:
-
Sameeh Jubran authored
The number of queues can be derived using ethtool, no need to print it in ena_probe() Signed-off-by:
-
Sameeh Jubran authored
- Update ena_ethtool:ena_get_channels() to return adapter->max_io_queues so that ethtool -l returns the correct maximum queue number. - Change the name of ena_calc_io_queue_num() to ena_calc_max_io_queue_num() as it returns the maximum number of io queues and actual number of queues can be smaller if changed by ethtool -L which is implemented in a later commit. - Change variable name from io_queue_num to max_num_io_queues in ena_calc_max_io_queue_num() and ena_probe(). - Make all types of variables that convey the number and sizeof queues to be u32, for consistency with the API between the driver and the device. Signed-off-by:
Arthur Kiyanovski <akiyano@amazon.com> Signed-off-by:
-
Sameeh Jubran authored
Since we use the same IRQ and NAPI to service RX and TX then we need to use a combined channel instead of rx and tx channels. Signed-off-by:
-
Sameeh Jubran authored
- Rename ena_calc_queue_size() to ena_calc_io_queue_size() for clarity and consistency - Remove redundant number of io queues parameter in functions ena_enable_msix() and ena_enable_msix_and_set_admin_interrupts(), which already get adapter parameter, so use adapter->num_io_queues in the function instead. - Use the local variable ena_dev instead of ctx->ena_dev in ena_calc_io_queue_size - Fix multi row comment alignments Signed-off-by:
-
