- 15 Dec, 2020 29 commits
-
-
Pavan Chebbi authored
In bnxt_flash_package_from_fw_obj(), if firmware returns the NO_SPACE error, call __bnxt_flash_nvram() to create the UPDATE directory and then loop back and retry one more time. Since the first try may fail, we use the silent version to send the firmware commands. Reviewed-by:
Vasundhara Volam <vasundhara-v.volam@broadcom.com> Reviewed-by:
Edwin Peer <edwin.peer@broadcom.com> Signed-off-by:
Pavan Chebbi <pavan.chebbi@broadcom.com> Signed-off-by:
Michael Chan <michael.chan@broadcom.com> Signed-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Pavan Chebbi authored
On NICs with a smaller NVRAM, FW installation may fail after multiple updates due to fragmentation. The driver can retry when FW returns a special error code. To faciliate the retry, we restructure the logic that performs the flashing in a loop. The actual retry logic will be added in the next patch. Signed-off-by:
-
Michael Chan authored
This function will be modified in the next patch to retry flashing the firmware in a loop. To facilate that, we rearrange the code so that the steps that only need to be done once before the loop will be moved to the top of the function. Signed-off-by:
-
Pavan Chebbi authored
Refactor bnxt_flash_nvram() into __bnxt_flash_nvram() that takes an additional dir_item_len parameter. The new function will be used in subsequent patches with the dir_item_len parameter set to create the UPDATE directory during flashing. Signed-off-by:
-
Marcin Wojtas authored
Since its creation Marvell NIC driver for Armada 375/7k8k and CN913x SoC families mvpp2 has been lacking an entry in MAINTAINERS, which sometimes lead to unhandled bugs that persisted across several kernel releases. Signed-off-by:
Marcin Wojtas <mw@semihalf.com> Acked-by:
Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/20201211165114.26290-1-mw@semihalf.comSigned-off-by:
-
Vasily Averin authored
syzbot reproduces BUG_ON in skb_checksum_help(): tun creates (bogus) skb with huge partial-checksummed area and small ip packet inside. Then ip_rcv trims the skb based on size of internal ip packet, after that csum offset points beyond of trimmed skb. Then checksum_tg() called via netfilter hook triggers BUG_ON: offset = skb_checksum_start_offset(skb); BUG_ON(offset >= skb_headlen(skb)); To work around the problem this patch forces pskb_trim_rcsum_slow() to return -EINVAL in described scenario. It allows its callers to drop such kind of packets. Link: https://syzkaller.appspot.com/bug?id=b419a5ca95062664fe1a60b764621eb4526e2cd0 Reported-by: syzbot+7010af67ced6105e5ab6@syzkaller.appspotmail.com Signed-off-by:Vasily Averin <vvs@virtuozzo.com> Acked-by:
Willem de Bruijn <willemb@google.com> Link: https://lore.kernel.org/r/1b2494af-2c56-8ee2-7bc0-923fcad1cdf8@virtuozzo.comSigned-off-by:
-
Toke Høiland-Jørgensen authored
Jakub pointed out that the IP_ECN_set* helpers basically open-code csum16_add(), so let's switch them over to using the helper instead. v2: - Use __be16 for check_add stack variable in IP_ECN_set_ce() (kbot) v3: - Turns out we need __force casts to do arithmetic on __be16 types Reported-by:
Jonathan Morton <chromatix99@gmail.com> Tested-by:
Pete Heist <pete@heistp.net> Signed-off-by:
Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20201211142638.154780-1-toke@redhat.comSigned-off-by:
-
Wang Hai authored
I got a warining report: br_sysfs_addbr: can't create group bridge4/bridge ------------[ cut here ]------------ sysfs group 'bridge' not found for kobject 'bridge4' WARNING: CPU: 2 PID: 9004 at fs/sysfs/group.c:279 sysfs_remove_group fs/sysfs/group.c:279 [inline] WARNING: CPU: 2 PID: 9004 at fs/sysfs/group.c:279 sysfs_remove_group+0x153/0x1b0 fs/sysfs/group.c:270 Modules linked in: iptable_nat ... Call Trace: br_dev_delete+0x112/0x190 net/bridge/br_if.c:384 br_dev_newlink net/bridge/br_netlink.c:1381 [inline] br_dev_newlink+0xdb/0x100 net/bridge/br_netlink.c:1362 __rtnl_newlink+0xe11/0x13f0 net/core/rtnetlink.c:3441 rtnl_newlink+0x64/0xa0 net/core/rtnetlink.c:3500 rtnetlink_rcv_msg+0x385/0x980 net/core/rtnetlink.c:5562 netlink_rcv_skb+0x134/0x3d0 net/netlink/af_netlink.c:2494 netlink_unicast_kernel net/netlink/af_netlink.c:1304 [inline] netlink_unicast+0x4a0/0x6a0 net/netlink/af_netlink.c:1330 netlink_sendmsg+0x793/0xc80 net/netlink/af_netlink.c:1919 sock_sendmsg_nosec net/socket.c:651 [inline] sock_sendmsg+0x139/0x170 net/socket.c:671 ____sys_sendmsg+0x658/0x7d0 net/socket.c:2353 ___sys_sendmsg+0xf8/0x170 net/socket.c:2407 __sys_sendmsg+0xd3/0x190 net/socket.c:2440 do_syscall_64+0x33/0x40 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xa9 In br_device_event(), if the bridge sysfs fails to be added, br_device_event() should return error. This can prevent warining when removing bridge sysfs that do not exist. Fixes: bb900b27 ("bridge: allow creating bridge devices with netlink") Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
Wang Hai <wanghai38@huawei.com> Tested-by:
Nikolay Aleksandrov <nikolay@nvidia.com> Acked-by:
-
Björn Töpel authored
Instead doing the check for allocation in each loop, move it outside the while loop and do it every NAPI loop. This change boosts the xdpsock rxdrop scenario with 15% more packets-per-second. Reviewed-by:
Maciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by:
Björn Töpel <bjorn.topel@intel.com> Link: https://lore.kernel.org/r/20201211085410.59350-1-bjorn.topel@gmail.comSigned-off-by:
-
Zheng Yongjun authored
Simplify the return expression at mtk_eth_path.c file, simplify this all. Signed-off-by:
Zheng Yongjun <zhengyongjun3@huawei.com> Link: https://lore.kernel.org/r/20201211083801.1632-1-zhengyongjun3@huawei.comSigned-off-by:
-
Jakub Kicinski authored
George Cherian says: ==================== Add devlink and devlink health reporters to octeontx2 Add basic devlink and devlink health reporters. Devlink health reporters are added for NPA block. Address Jakub's comment to add devlink support for error reporting. https://www.spinics.net/lists/netdev/msg670712.html For now, I have dropped the NIX block health reporters. This series attempts to add health reporters only for the NPA block. As per Jakub's suggestion separate reporters per event is used and also got rid of the counters. ==================== Link: https://lore.kernel.org/r/20201211062526.2302643-1-george.cherian@marvell.comSigned-off-by:
-
George Cherian authored
Add Documentation for devlink health reporters for NPA block. Signed-off-by:
George Cherian <george.cherian@marvell.com> Signed-off-by:
-
George Cherian authored
Add health reporters for RVU NPA block. NPA Health reporters handle following HW event groups - GENERAL events - ERROR events - RAS events - RVU event Output: #devlink health pci/0002:01:00.0: reporter hw_npa_intr state healthy error 0 recover 0 grace_period 0 auto_recover true auto_dump true reporter hw_npa_gen state healthy error 0 recover 0 grace_period 0 auto_recover true auto_dump true reporter hw_npa_err state healthy error 0 recover 0 grace_period 0 auto_recover true auto_dump true reporter hw_npa_ras state healthy error 0 recover 0 grace_period 0 auto_recover true auto_dump true #devlink health dump show pci/0002:01:00.0 reporter hw_npa_err NPA_AF_ERR: NPA Error Interrupt Reg : 4096 AQ Doorbell Error #devlink health dump show pci/0002:01:00.0 reporter hw_npa_ras NPA_AF_RVU_RAS: NPA RAS Interrupt Reg : 0 Each reporter dump shows the Register value and the description of the cause. Signed-off-by:Sunil Kovvuri Goutham <sgoutham@marvell.com> Signed-off-by:
Jerin Jacob <jerinj@marvell.com> Signed-off-by:
-
George Cherian authored
Add devlink support to AF driver. Basic devlink support is added. Currently info_get is the only supported devlink ops. devlink ouptput looks like this # devlink dev pci/0002:01:00.0 # devlink dev info pci/0002:01:00.0: driver octeontx2-af # Signed-off-by:
-
Po-Hsu Lin authored
The cleanup function in this script that tries to delete hv-1 / hv-2 vm-1 / vm-2 netns will generate some uncessary error messages: Cannot remove namespace file "/run/netns/hv-2": No such file or directory Cannot remove namespace file "/run/netns/vm-1": No such file or directory Cannot remove namespace file "/run/netns/vm-2": No such file or directory Redirect it to /dev/null like other commands in the cleanup function to reduce confusion. Signed-off-by:
Po-Hsu Lin <po-hsu.lin@canonical.com> Link: https://lore.kernel.org/r/20201211042420.16411-1-po-hsu.lin@canonical.comSigned-off-by:
-
Cambda Zhu authored
For each TCP zero window probe, the icsk_backoff is increased by one and its max value is tcp_retries2. If tcp_retries2 is greater than 63, the probe0 timeout shift may exceed its max bits. On x86_64/ARMv8/MIPS, the shift count would be masked to range 0 to 63. And on ARMv7 the result is zero. If the shift count is masked, only several probes will be sent with timeout shorter than TCP_RTO_MAX. But if the timeout is zero, it needs tcp_retries2 times probes to end this false timeout. Besides, bitwise shift greater than or equal to the width is an undefined behavior. This patch adds a limit to the backoff. The max value of max_when is TCP_RTO_MAX and the min value of timeout base is TCP_RTO_MIN. The limit is the backoff from TCP_RTO_MIN to TCP_RTO_MAX. Signed-off-by:
Cambda Zhu <cambda@linux.alibaba.com> Link: https://lore.kernel.org/r/20201208091910.37618-1-cambda@linux.alibaba.comSigned-off-by:
-
Jakub Kicinski authored
Mat Martineau says: ==================== mptcp: Another set of miscellaneous MPTCP fixes This is another collection of MPTCP fixes and enhancements that we have tested in the MPTCP tree: Patch 1 cleans up cgroup attachment for in-kernel subflow sockets. Patches 2 and 3 make sure that deletion of advertised addresses by an MPTCP path manager when flushing all addresses behaves similarly to the remove-single-address operation, and adds related tests. Patches 4 and 8 do some minor cleanup. Patches 5-7 add MPTCP_FASTCLOSE functionality. Note that patch 6 adds MPTCP option parsing to tcp_reset(). Patch 9 optimizes skb size for outgoing MPTCP packets. ==================== Link: https://lore.kernel.org/r/20201210222506.222251-1-mathew.j.martineau@linux.intel.comSigned-off-by:
-
Paolo Abeni authored
Currently the xmit path of the MPTCP protocol creates smaller- than-max-size skbs, which is suboptimal for the performances. There are a few things to improve: - when coalescing to an existing skb, must clear the PUSH flag - tcp_build_frag() expect the available space as an argument. When coalescing is enable MPTCP already subtracted the to-be-coalesced skb len. We must increment said argument accordingly. Before: ./use_mptcp.sh netperf -H 127.0.0.1 -t TCP_STREAM [...] 131072 16384 16384 30.00 24414.86 After: ./use_mptcp.sh netperf -H 127.0.0.1 -t TCP_STREAM [...] 131072 16384 16384 30.05 28357.69 Signed-off-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by:
-
Paolo Abeni authored
There is no need to unconditionally acquire the join list lock, we can simply splice the join list into the subflow list and traverse only the latter. Signed-off-by:
-
Florian Westphal authored
parse the MPTCP FASTCLOSE subtype. If provided key matches the local one, schedule the work queue to close (with tcp reset) all subflows. The MPTCP socket moves to closed state immediately. Reviewed-by:
Matthieu Baerts <matthieu.baerts@tessares.net> Signed-off-by:
Florian Westphal <fw@strlen.de> Signed-off-by:
-
Florian Westphal authored
Because TCP-level resets only affect the subflow, there is a MPTCP option to indicate that the MPTCP-level connection should be closed immediately without a mptcp-level fin exchange. This is the 'MPTCP fast close option'. It can be carried on ack segments or TCP resets. In the latter case, its needed to parse mptcp options also for reset packets so that MPTCP can act accordingly. Next patch will add receive side fastclose support in MPTCP. Acked-by:
-
Florian Westphal authored
When processing options from tcp reset path its possible that tcp_done(ssk) drops the last reference on the mptcp socket which results in use-after-free. Reviewed-by:
-
Geliang Tang authored
Use the macro MPTCPOPT_HMAC_LEN instead of a constant in struct mptcp_options_received. Signed-off-by:
Geliang Tang <geliangtang@gmail.com> Signed-off-by:
-
Geliang Tang authored
This patch added the flush addrs testcase. In do_transfer, if the number of removing addresses is less than 8, use the del addr command to remove the addresses one by one. If the number is more than 8, use the flush addrs command to remove the addresses. Acked-by:
-
Geliang Tang authored
When the PM netlink flushes the addresses, invoke the remove address function mptcp_nl_remove_subflow_and_signal_addr to remove the addresses and the subflows. Since this function should not be invoked under lock, move __flush_addrs out of the pernet->lock. Acked-by:
-
Nicolas Rybowski authored
It has been observed that the kernel sockets created for the subflows (except the first one) are not in the same cgroup as their parents. That's because the additional subflows are created by kernel workers. This is a problem with eBPF programs attached to the parent's cgroup won't be executed for the children. But also with any other features of CGroup linked to a sk. This patch fixes this behaviour. As the subflow sockets are created by the kernel, we can't use 'mem_cgroup_sk_alloc' because of the current context being the one of the kworker. This is why we have to do low level memcg manipulation, if required. Suggested-by:
Nicolas Rybowski <nicolas.rybowski@tessares.net> Signed-off-by:
-
Loic Poulain authored
This patch checks that MHI queue is not full before waking up the net queue. This fix sporadic MHI queueing issues in xmit. Indeed xmit and its symmetric complete callback (ul_callback) can run concurently, it is then not safe to unconditionnaly waking the queue in the callback without checking queue fullness. Fixes: 3ffec6a1 ("net: Add mhi-net driver") Signed-off-by:
Loic Poulain <loic.poulain@linaro.org> Link: https://lore.kernel.org/r/1607599507-5879-1-git-send-email-loic.poulain@linaro.orgSigned-off-by:
-
Rasmus Villemoes authored
The 6220 and 6250 switches do not have a learn2all bit in global1, ATU control register; bit 3 is reserverd. On the switches that do have that bit, it is used to control whether learning frames are sent out the ports that have the message_port bit set. So rather than adding yet another chip method, use the existence of the ->port_setup_message_port method as a proxy for determining whether the learn2all bit exists (and should be set). Signed-off-by:
Rasmus Villemoes <rasmus.villemoes@prevas.dk> Link: https://lore.kernel.org/r/20201210110645.27765-1-rasmus.villemoes@prevas.dkSigned-off-by:
-
Eelco Chaudron authored
Currently, the exception actions are not processed correctly as the wrong dataset is passed. This change fixes this, including the misleading comment. In addition, a check was added to make sure we work on an IPv4 packet, and not just assume if it's not IPv6 it's IPv4. This was all tested using OVS with patch, https://patchwork.ozlabs.org/project/openvswitch/list/?series=21639, applied and sending packets with a TTL of 1 (and 0), both with IPv4 and IPv6. Fixes: 69929d4c ("net: openvswitch: fix TTL decrement action netlink message format") Signed-off-by:
Eelco Chaudron <echaudro@redhat.com> Link: https://lore.kernel.org/r/160733569860.3007.12938188180387116741.stgit@wsfd-netdev64.ntdv.lab.eng.bos.redhat.comSigned-off-by:
-
- 14 Dec, 2020 7 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-nextJakub Kicinski authored
Pablo Neira Ayuso says: ==================== Netfilter/IPVS updates for net-next 1) Missing dependencies in NFT_BRIDGE_REJECT, from Randy Dunlap. 2) Use atomic_inc_return() instead of atomic_add_return() in IPVS, from Yejune Deng. 3) Simplify check for overquota in xt_nfacct, from Kaixu Xia. 4) Move nfnl_acct_list away from struct net, from Miao Wang. 5) Pass actual sk in reject actions, from Jan Engelhardt. 6) Add timeout and protoinfo to ctnetlink destroy events, from Florian Westphal. 7) Four patches to generalize set infrastructure to support for multiple expressions per set element. * git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next: netfilter: nftables: netlink support for several set element expressions netfilter: nftables: generalize set extension to support for several expressions netfilter: nftables: move nft_expr before nft_set netfilter: nftables: generalize set expressions support netfilter: ctnetlink: add timeout and protoinfo to destroy events netfilter: use actual socket sk for REJECT action netfilter: nfnl_acct: remove data from struct net netfilter: Remove unnecessary conversion to bool ipvs: replace atomic_add_return() netfilter: nft_reject_bridge: fix build errors due to code movement ==================== Link: https://lore.kernel.org/r/20201212230513.3465-1-pablo@netfilter.orgSigned-off-by:
-
https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-nextJakub Kicinski authored
Daniel Borkmann says: ==================== pull-request: bpf-next 2020-12-14 1) Expose bpf_sk_storage_*() helpers to iterator programs, from Florent Revest. 2) Add AF_XDP selftests based on veth devs to BPF selftests, from Weqaar Janjua. 3) Support for finding BTF based kernel attach targets through libbpf's bpf_program__set_attach_target() API, from Andrii Nakryiko. 4) Permit pointers on stack for helper calls in the verifier, from Yonghong Song. 5) Fix overflows in hash map elem size after rlimit removal, from Eric Dumazet. 6) Get rid of direct invocation of llc in BPF selftests, from Andrew Delgadillo. 7) Fix xsk_recvmsg() to reorder socket state check before access, from Björn Töpel. 8) Add new libbpf API helper to retrieve ring buffer epoll fd, from Brendan Jackman. 9) Batch of minor BPF selftest improvements all over the place, from Florian Lehner, KP Singh, Jiri Olsa and various others. * https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (31 commits) selftests/bpf: Add a test for ptr_to_map_value on stack for helper access bpf: Permits pointers on stack for helper calls libbpf: Expose libbpf ring_buffer epoll_fd selftests/bpf: Add set_attach_target() API selftest for module target libbpf: Support modules in bpf_program__set_attach_target() API selftests/bpf: Silence ima_setup.sh when not running in verbose mode. selftests/bpf: Drop the need for LLVM's llc selftests/bpf: fix bpf_testmod.ko recompilation logic samples/bpf: Fix possible hang in xdpsock with multiple threads selftests/bpf: Make selftest compilation work on clang 11 selftests/bpf: Xsk selftests - adding xdpxceiver to .gitignore selftests/bpf: Drop tcp-{client,server}.py from Makefile selftests/bpf: Xsk selftests - Bi-directional Sockets - SKB, DRV selftests/bpf: Xsk selftests - Socket Teardown - SKB, DRV selftests/bpf: Xsk selftests - DRV POLL, NOPOLL selftests/bpf: Xsk selftests - SKB POLL, NOPOLL selftests/bpf: Xsk selftests framework bpf: Only provide bpf_sock_from_file with CONFIG_NET bpf: Return -ENOTSUPP when attaching to non-kernel BTF xsk: Validate socket state in xsk_recvmsg, prior touching socket members ... ==================== Link: https://lore.kernel.org/r/20201214214316.20642-1-daniel@iogearbox.netSigned-off-by:
-
Yonghong Song authored
Change bpf_iter_task.c such that pointer to map_value may appear on the stack for bpf_seq_printf() to access. Without previous verifier patch, the bpf_iter test will fail. Signed-off-by:
Yonghong Song <yhs@fb.com> Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Acked-by:
Song Liu <songliubraving@fb.com> Link: https://lore.kernel.org/bpf/20201210013350.943985-1-yhs@fb.com
-
Yonghong Song authored
Currently, when checking stack memory accessed by helper calls, for spills, only PTR_TO_BTF_ID and SCALAR_VALUE are allowed. Song discovered an issue where the below bpf program int dump_task(struct bpf_iter__task *ctx) { struct seq_file *seq = ctx->meta->seq; static char[] info = "abc"; BPF_SEQ_PRINTF(seq, "%s\n", info); return 0; } may cause a verifier failure. The verifier output looks like: ; struct seq_file *seq = ctx->meta->seq; 1: (79) r1 = *(u64 *)(r1 +0) ; BPF_SEQ_PRINTF(seq, "%s\n", info); 2: (18) r2 = 0xffff9054400f6000 4: (7b) *(u64 *)(r10 -8) = r2 5: (bf) r4 = r10 ; 6: (07) r4 += -8 ; BPF_SEQ_PRINTF(seq, "%s\n", info); 7: (18) r2 = 0xffff9054400fe000 9: (b4) w3 = 4 10: (b4) w5 = 8 11: (85) call bpf_seq_printf#126 R1_w=ptr_seq_file(id=0,off=0,imm=0) R2_w=map_value(id=0,off=0,ks=4,vs=4,imm=0) R3_w=inv4 R4_w=fp-8 R5_w=inv8 R10=fp0 fp-8_w=map_value last_idx 11 first_idx 0 regs=8 stack=0 before 10: (b4) w5 = 8 regs=8 stack=0 before 9: (b4) w3 = 4 invalid indirect read from stack off -8+0 size 8 Basically, the verifier complains the map_value pointer at "fp-8" location. To fix the issue, if env->allow_ptr_leaks is true, let us also permit pointers on the stack to be accessible by the helper. Reported-by:Alexei Starovoitov <ast@kernel.org> Signed-off-by:
-
Brendan Jackman authored
This provides a convenient perf ringbuf -> libbpf ringbuf migration path for users of external polling systems. It is analogous to perf_buffer__epoll_fd. Signed-off-by:
Brendan Jackman <jackmanb@google.com> Signed-off-by:
Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20201214113812.305274-1-jackmanb@google.com
-
Andrii Nakryiko authored
Add test for bpf_program__set_attach_target() API, validating it can find kernel module fentry target. Signed-off-by:
-
Andrii Nakryiko authored
Support finding kernel targets in kernel modules when using bpf_program__set_attach_target() API. This brings it up to par with what libbpf supports when doing declarative SEC()-based target determination. Some minor internal refactoring was needed to make sure vmlinux BTF can be loaded before bpf_object's load phase. Signed-off-by:
-
- 13 Dec, 2020 1 commit
-
-
Xie He authored
According to the X.25 documentation, there was a plan to implement X.25-over-802.2-LLC. It never finished but left various code stubs in the X.25 code. At this time it is unlikely that it would ever finish so it may be better to remove those code stubs. Also change the documentation to make it clear that this is not a ongoing plan anymore. Change words like "will" to "could", "would", etc. Cc: Martin Schiller <ms@dev.tdt.de> Signed-off-by:
Xie He <xie.he.0141@gmail.com> Link: https://lore.kernel.org/r/20201209033346.83742-1-xie.he.0141@gmail.comSigned-off-by:
-
- 12 Dec, 2020 3 commits
-
-
SeongJae Park authored
On a few of our systems, I found frequent 'unshare(CLONE_NEWNET)' calls make the number of active slab objects including 'sock_inode_cache' type rapidly and continuously increase. As a result, memory pressure occurs. In more detail, I made an artificial reproducer that resembles the workload that we found the problem and reproduce the problem faster. It merely repeats 'unshare(CLONE_NEWNET)' 50,000 times in a loop. It takes about 2 minutes. On 40 CPU cores / 70GB DRAM machine, the available memory continuously reduced in a fast speed (about 120MB per second, 15GB in total within the 2 minutes). Note that the issue don't reproduce on every machine. On my 6 CPU cores machine, the problem didn't reproduce. 'cleanup_net()' and 'fqdir_work_fn()' are functions that deallocate the relevant memory objects. They are asynchronously invoked by the work queues and internally use 'rcu_barrier()' to ensure safe destructions. 'cleanup_net()' works in a batched maneer in a single thread worker, while 'fqdir_work_fn()' works for each 'fqdir_exit()' call in the 'system_wq'. Therefore, 'fqdir_work_fn()' called frequently under the workload and made the contention for 'rcu_barrier()' high. In more detail, the global mutex, 'rcu_state.barrier_mutex' became the bottleneck. This commit avoids such contention by doing the 'rcu_barrier()' and subsequent lightweight works in a batched manner, as similar to that of 'cleanup_net()'. The fqdir hashtable destruction, which is done before the 'rcu_barrier()', is still allowed to run in parallel for fast processing, but this commit makes it to use a dedicated work queue instead of the 'system_wq', to make sure that the number of threads is bounded. Signed-off-by:
SeongJae Park <sjpark@amazon.de> Reviewed-by:
Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20201211112405.31158-1-sjpark@amazon.comSigned-off-by:
-
Krzysztof Kozlowski authored
If interrupt trigger is not set when requesting the interrupt, the core will take care of reading trigger type from Devicetree. There is no point to do it in the driver. Signed-off-by:
Krzysztof Kozlowski <krzk@kernel.org> Link: https://lore.kernel.org/r/20201210211824.214949-1-krzk@kernel.orgSigned-off-by:
-
DENG Qingfang authored
MT7530 has a global RX length register, so we are actually changing its MRU. Enable MTU normalization for this reason. Signed-off-by:
DENG Qingfang <dqfext@gmail.com> Acked-by:
Landen Chao <landen.chao@mediatek.com> Reviewed-by:
Vladimir Oltean <olteanv@gmail.com> Link: https://lore.kernel.org/r/20201210170322.3433-1-dqfext@gmail.comSigned-off-by:
-
