- 18 Aug, 2017 11 commits
-
-
Julian Wiedmann authored
Taking a full copy via skb_realloc_headroom() on every xmit is overkill and wastes CPU time; all we actually need is to push on the qeth_hdr. So rework the L2 OSA TX path to avoid the copy. Minor complications arise because struct qeth_hdr must not cross a page boundary. So add a new helper qeth_push_hdr() that catches this, and falls back to the hdr cache that we already use for IQDs. This change uncovered that qeth's TX completion takes rather long. Now that we no longer free the original skb straight away and thus call skb->destructor later than before, throughput regresses significantly. For now, restore old behaviour by adding an explicit skb_orphan(), and a big TODO to improve the TX completion time. Tested-by:
Nils Hoppmann <niho@de.ibm.com> Signed-off-by:
Julian Wiedmann <jwi@linux.vnet.ibm.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Julian Wiedmann authored
After plenty of refactoring, use hd_len as single indication that the skb needs a dedicated header element. This preserves existing behaviour for TSO, as 'hdr' always points to skb->data. Signed-off-by:
-
Julian Wiedmann authored
This is a prerequisite for unifying the code to build header elements. The TSO header has a different size, so we can no longer rely on implicitly adding the size of a normal qeth_hdr. No functional change. Signed-off-by:
-
Julian Wiedmann authored
For TSO we need to skip the skb's qeth/IP/TCP headers when mapping it into buffer elements. Instead of (mis)using skb_pull(), pass a corresponding offset to fill_buffer() like we already do for IQDs. No actual change in the resulting TSO buffers. Signed-off-by:
-
Julian Wiedmann authored
The TSO code already calculates the length of its header element, no need to duplicate this in the low-level code again. Use this opportunity to make hd_len unsigned, and for TSO match its calculation to what tso_fill_header() does. No functional change. Signed-off-by:
-
Julian Wiedmann authored
For IQD we already need to fix up the qeth_hdr's length field, and future changes will require more flexibility for OSA as well. The device-specific path knows best what header length it requires, so just pass it from there. While at it, remove the unused qeth_card parameter. No functional change. Signed-off-by:
-
Julian Wiedmann authored
l2_hard_start_xmit() actually doesn't contain much shared code, and having device-specific paths makes isolated changes a lot easier. So split it into three routines for IQD, OSN and OSD/OSM/OSX. No functional change. Signed-off-by:
-
Dan Carpenter authored
"map" is a valid pointer. We wanted to return "err" instead. Also let's return a zero literal at the end. Fixes: 174a79ff ("bpf: sockmap with sk redirect support") Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Acked-by:
John Fastabend <john.fastabend@gmail.com> Acked-by:
Daniel Borkmann <daniel@iogearbox.net> Signed-off-by:
-
David S. Miller authored
Rick Farrington says: ==================== liquidio: initialization fixes for embedded firmware Fix problems when using an adapter w/embedded f/w (param "fw_type=none"). 1. Add support for PF FLR when exiting. 2. Skip some initialization (don't try to load f/w, activate consoles). 3. Issue credits BEFORE enabling DROQs. ==================== Signed-off-by: -
Rick Farrington authored
1. Issue credits BEFORE enabling DROQ's; this prevents PKTPF_ERR interrupt. Signed-off-by:
Rick Farrington <ricardo.farrington@cavium.com> Signed-off-by:
Felix Manlunas <felix.manlunas@cavium.com> Signed-off-by:
-
Rick Farrington authored
1. Add support for PF FLR when exiting (enables CORE_DRV_ACTIVE upon next driver init) 2. Skip some initialization (don't try to load f/w, activate consoles). Signed-off-by:
-
- 17 Aug, 2017 9 commits
-
-
David S. Miller authored
Daniel Borkmann says: ==================== Two BPF smap related followups Fixing preemption imbalance and consolidating prologue generation. Thanks! ==================== Signed-off-by: -
Daniel Borkmann authored
Given both program types are effecitvely doing the same in the prologue, just reuse the one that we had for tc and only adapt to the corresponding drop verdict value. That way, we don't need to have the duplicate from 8a31db56 ("bpf: add access to sock fields and pkt data from sk_skb programs") to maintain. Reported-by:
Alexei Starovoitov <ast@kernel.org> Signed-off-by:
-
Daniel Borkmann authored
In smap_do_verdict(), the fall-through branch leads to call preempt_enable() twice for the SK_REDIRECT, which creates an imbalance. Only enable it for all remaining cases again. Fixes: 174a79ff ("bpf: sockmap with sk redirect support") Reported-by:
-
Arvind Yadav authored
vio_device_id are not supposed to change at runtime. All functions working with vio_device_id provided by <asm/vio.h> work with const vio_device_id. So mark the non-const structs as const. Signed-off-by:
Arvind Yadav <arvind.yadav.cs@gmail.com> Signed-off-by:
-
Arvind Yadav authored
vio_device_id are not supposed to change at runtime. All functions working with vio_device_id provided by <asm/vio.h> work with const vio_device_id. So mark the non-const structs as const. Signed-off-by:
-
Daniel Borkmann authored
We are guaranteed to have a NULL ri->map in this branch since we test for it earlier, so we don't need to reset it here. Signed-off-by:
-
Daniel Borkmann authored
Using parent->regs[] when propagating REG_LIVE_READ for spilled regs doesn't work since parent->regs[] denote the set of normal registers but not spilled ones. Propagate to the correct regs. Fixes: dc503a8a ("bpf/verifier: track liveness for pruning") Reported-by:
Edward Cree <ecree@solarflare.com> Signed-off-by:
-
Colin Ian King authored
Media type is only set if h->ae_algo->ops->get_media_type is called so there is a possibility that media_type is uninitialized when it is used a switch statement. Fix this by initializing media_type to HNAE3_MEDIA_TYPE_UNKNOWN. Detected by CoverityScan, CID#1452624("Uninitialized scalar variable") Fixes: 496d03e9 ("net: hns3: Add Ethtool support to HNS3 driver") Signed-off-by:Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Colin Ian King authored
Trivial fix to spelling mistake in dev_info message Signed-off-by:
-
- 16 Aug, 2017 20 commits
-
-
David S. Miller authored
Stephen Hemminger says: ==================== vmbus sendpacket cleanups These patches remove and consolidate vmbus_sendpacket functions. They should go through the net-next tree since these API's were only used by the netvsc driver. ==================== Signed-off-by: -
stephen hemminger authored
The only usage of vmbus_sendpacket_ctl was by vmbus_sendpacket. Signed-off-by:
Stephen Hemminger <sthemmin@microsoft.com> Signed-off-by:
-
stephen hemminger authored
The function vmbus_sendpacket_pagebuffer_ctl was never used directly. Just have vmbus_send_pagebuffer Signed-off-by:
-
stephen hemminger authored
This function is not used anywhere in current code. Signed-off-by:
-
David S. Miller authored
Signed-off-by: -
David S. Miller authored
John Fastabend says: ==================== bpf: sockmap build fixes Two build fixes for sockmap, this should resolve the build errors and warnings that were reported. Thanks everyone. ==================== Signed-off-by: -
John Fastabend authored
Resolve issues with !CONFIG_BPF_SYSCALL and !STREAM_PARSER net/core/filter.c: In function ‘do_sk_redirect_map’: net/core/filter.c:1881:3: error: implicit declaration of function ‘__sock_map_lookup_elem’ [-Werror=implicit-function-declaration] sk = __sock_map_lookup_elem(ri->map, ri->ifindex); ^ net/core/filter.c:1881:6: warning: assignment makes pointer from integer without a cast [enabled by default] sk = __sock_map_lookup_elem(ri->map, ri->ifindex); Fixes: 174a79ff ("bpf: sockmap with sk redirect support") Reported-by:
Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by:
-
John Fastabend authored
psock will uninitialized in default case we need to do the same psock lookup and check as in other branch. Fixes compile warning below. kernel/bpf/sockmap.c: In function ‘smap_state_change’: kernel/bpf/sockmap.c:156:21: warning: ‘psock’ may be used uninitialized in this function [-Wmaybe-uninitialized] struct smap_psock *psock; Fixes: 174a79ff ("bpf: sockmap with sk redirect support") Reported-by:
-
Jiri Pirko authored
I made a stupid mistake using TC_CLSFLOWER_STATS instead of TC_SETUP_CLSFLOWER. Funny thing is that both are defined as "2" so it actually did not cause any harm. Anyway, fixing it now. Fixes: 2572ac53 ("net: sched: make type an argument for ndo_setup_tc") Signed-off-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
-
Eric Dumazet authored
tun_build_skb() is not thread safe since it uses per queue page frag, this will break things when multiple threads are sending through same queue. Switch to use per-thread generator (no lock involved). Fixes: 66ccbc9c ("tap: use build_skb() for small packet") Tested-by:
Jason Wang <jasowang@redhat.com> Signed-off-by:
Michael S. Tsirkin <mst@redhat.com> Signed-off-by:
-
Colin Ian King authored
Trivial fix to spelling mistakes in the mlx4 driver Signed-off-by:
Yuval Shaia <yuval.shaia@oracle.com> Reviewed-by:
Leon Romanovsky <leonro@mellanox.com> Signed-off-by:
-
Jesper Dangaard Brouer authored
The main purpose of this tracepoint is to monitor bulk dequeue in the network qdisc layer, as it cannot be deducted from the existing qdisc stats. The txq_state can be used for determining the reason for zero packet dequeues, see enum netdev_queue_state_t. Notice all packets doesn't necessary activate this tracepoint. As qdiscs with flag TCQ_F_CAN_BYPASS, can directly invoke sch_direct_xmit() when qdisc_qlen is zero. Remember that perf record supports filters like: perf record -e qdisc:qdisc_dequeue \ --filter 'ifindex == 4 && (packets > 1 || txq_state > 0)' Signed-off-by:
Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by:
-
David S. Miller authored
Simon Horman says: ==================== nfp: process MTU updates from firmware flower app The first patch of this series moves processing of control messages from a BH handler to a workqueue. That change makes it safe to process MTU updates from the firmware which is added by the second patch of this series. ==================== Signed-off-by: -
Simon Horman authored
Now that control message processing occurs in a workqueue rather than a BH handler MTU updates received from the firmware may be safely processed. Signed-off-by:
Simon Horman <simon.horman@netronome.com> Reviewed-by:
Jakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by:
-
Simon Horman authored
Processing of control messages is not time-critical and future processing of some messages will require taking the RTNL which is not possible in a BH handler. It seems simplest to move all control message processing to a workqueue. Signed-off-by:
-
John Fastabend authored
In the devmap alloc map logic we check to ensure that the sizeof the values are not greater than KMALLOC_MAX_SIZE. But, in the dev map case we ensure the value size is 4bytes earlier in the function because all values should be netdev ifindex values. The second check is harmless but is not needed so remove it. Signed-off-by:
-
David S. Miller authored
John Fastabend says: ==================== BPF: sockmap and sk redirect support This series implements a sockmap and socket redirect helper for BPF using a model similar to XDP netdev redirect. A sockmap is a BPF map type that holds references to sock structs. Then with a new sk redirect bpf helper BPF programs can use the map to redirect skbs between sockets, bpf_sk_redirect_map(map, key, flags) Finally, we need a call site to attach our BPF logic to do socket redirects. We added hooks to recv_sock using the existing strparser infrastructure to do this. The call site is added via the BPF attach map call. To enable users to use this infrastructure a new BPF program BPF_PROG_TYPE_SK_SKB is created that allows users to reference sock details, such as port and ip address fields, to build useful socket layer program. The sockmap datapath is as follows, recv -> strparser -> verdict/action where this series implements the drop and redirect actions. Additional, actions can be added as needed. A sample program is provided to illustrate how a sockmap can be integrated with cgroups and used to add/delete sockets in a sockmap. The program is simple but should show many of the key ideas. To test this work test_maps in selftests/bpf was leveraged. We added a set of tests to add sockets and do send/recv ops on the sockets to ensure correct behavior. Additionally, the selftests tests a series of negative test cases. We can expand on this in the future. I also have a basic test program I use with iperf/netperf clients that could be sent as an additional sample if folks want this. It needs a bit of cleanup to send to the list and wasn't included in this series. For people who prefer git over pulling patches out of their mail editor I've posted the code here, https://github.com/jrfastab/linux-kernel-xdp/tree/sockmap For some background information on the genesis of this work it might be helpful to review these slides from netconf 2017 by Thomas Graf, http://vger.kernel.org/netconf2017.html https://docs.google.com/a/covalent.io/presentation/d/1dwSKSBGpUHD3WO5xxzZWj8awV_-xL-oYhvqQMOBhhtk/edit?usp=sharing Thanks to Daniel Borkmann for reviewing and providing initial feedback. ==================== Acked-by: -
John Fastabend authored
This generates a set of sockets, attaches BPF programs, and sends some simple traffic using basic send/recv pattern. Additionally, we do a bunch of negative tests to ensure adding/removing socks out of the sockmap fail correctly. Signed-off-by:
-
John Fastabend authored
This adds tests to access new __sk_buff members from sk skb program type. Signed-off-by:
-
John Fastabend authored
This program binds a program to a cgroup and then matches hard coded IP addresses and adds these to a sockmap. This will receive messages from the backend and send them to the client. client:X <---> frontend:10000 client:X <---> backend:10001 To keep things simple this is only designed for 1:1 connections using hard coded values. A more complete example would allow many backends and clients. To run, # sockmap <cgroup2_dir> Signed-off-by:
-
