- 23 Nov, 2023 6 commits
-
-
Raju Rangoju authored
xgbe_get_link_ksettings() does not propagate correct speed and duplex information to ethtool during cable unplug. Due to which ethtool reports incorrect values for speed and duplex. Address this by propagating correct information. Fixes: 7c12aa08 ("amd-xgbe: Move the PHY support into amd-xgbe") Acked-by:
Shyam Sundar S K <Shyam-sundar.S-k@amd.com> Signed-off-by:
Raju Rangoju <Raju.Rangoju@amd.com> Reviewed-by:
Wojciech Drewek <wojciech.drewek@intel.com> Signed-off-by:
Paolo Abeni <pabeni@redhat.com>
-
Raju Rangoju authored
The existing implementation uses software logic to accumulate tx completions until the specified time (1ms) is met and then poll them. However, there exists a tiny gap which leads to a race between resetting and checking the tx_activate flag. Due to this the tx completions are not reported to upper layer and tx queue timeout kicks-in restarting the device. To address this, introduce a tx cleanup mechanism as part of the periodic maintenance process. Fixes: c5aa9e3b ("amd-xgbe: Initial AMD 10GbE platform driver") Acked-by:
-
Raju Rangoju authored
Force the mode change for SFI in Fixed PHY configurations. Fixed PHY configurations needs PLL to be enabled while doing mode set. When the SFP module isn't connected during boot, driver assumes AN is ON and attempts auto-negotiation. However, if the connected SFP comes up in Fixed PHY configuration the link will not come up as PLL isn't enabled while the initial mode set command is issued. So, force the mode change for SFI in Fixed PHY configuration to fix link issues. Fixes: e57f7a3f ("amd-xgbe: Prepare for working with more than one type of phy") Acked-by:
-
Lorenzo Bianconi authored
Fix a possible misalignment between page_pool stats and tx xdp_stats reported in veth_get_ethtool_stats routine. The issue can be reproduced configuring the veth pair with the following tx/rx queues: $ip link add v0 numtxqueues 2 numrxqueues 4 type veth peer name v1 \ numtxqueues 1 numrxqueues 1 and loading a simple XDP program on v0 that just returns XDP_PASS. In this case on v0 the page_pool stats overwrites tx xdp_stats for queue 1. Fix the issue incrementing pp_idx of dev->real_num_tx_queues * VETH_TQ_STATS_LEN since we always report xdp_stats for all tx queues in ethtool. Fixes: 4fc41805 ("net: veth: add page_pool stats") Signed-off-by:
Lorenzo Bianconi <lorenzo@kernel.org> Link: https://lore.kernel.org/r/c5b5d0485016836448453f12846c7c4ab75b094a.1700593593.git.lorenzo@kernel.orgSigned-off-by:
-
Suman Ghosh authored
It is possible to add a ntuple rule which would like to direct packet to a VF whose number of queues are greater/less than its PF's queue numbers. For example a PF can have 2 Rx queues but a VF created on that PF can have 8 Rx queues. As of today, ntuple rule will reject rule because it is checking the requested queue number against PF's number of Rx queues. As a part of this fix if the action of a ntuple rule is to move a packet to a VF's queue then the check is removed. Also, a debug information is printed to aware user that it is user's responsibility to cross check if the requested queue number on that VF is a valid one. Fixes: f0a1913f ("octeontx2-pf: Add support for ethtool ntuple filters") Signed-off-by:
Suman Ghosh <sumang@marvell.com> Reviewed-by:
Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20231121165624.3664182-1-sumang@marvell.comSigned-off-by:
-
Lech Perczak authored
Interface 4 is used by for QMI interface in stock firmware of MF28D, the router which uses MF290 modem. Rebind it to qmi_wwan after freeing it up from option driver. The proper configuration is: Interface mapping is: 0: QCDM, 1: (unknown), 2: AT (PCUI), 2: AT (Modem), 4: QMI T: Bus=01 Lev=02 Prnt=02 Port=00 Cnt=01 Dev#= 4 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=19d2 ProdID=0189 Rev= 0.00 S: Manufacturer=ZTE, Incorporated S: Product=ZTE LTE Technologies MSM C:* #Ifs= 5 Cfg#= 1 Atr=e0 MxPwr=500mA I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option E: Ad=82(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=option E: Ad=84(I) Atr=03(Int.) MxPS= 64 Ivl=2ms E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan E: Ad=86(I) Atr=03(Int.) MxPS= 64 Ivl=2ms E: Ad=87(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=05(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms Cc: Bjørn Mork <bjorn@mork.no> Signed-off-by:
Lech Perczak <lech.perczak@gmail.com> Link: https://lore.kernel.org/r/20231117231918.100278-3-lech.perczak@gmail.comSigned-off-by:
-
- 22 Nov, 2023 4 commits
-
-
Heiner Kallweit authored
This reverts commit efa5f131. I couldn't reproduce the reported issue. What I did, based on a pcap packet log provided by the reporter: - Used same chip version (RTL8168h) - Set MAC address to the one used on the reporters system - Replayed the EAPOL unicast packet that, according to the reporter, was filtered out by the mc filter. The packet was properly received. Therefore the root cause of the reported issue seems to be somewhere else. Disabling mc filtering completely for the most common chip version is a quite big hammer. Therefore revert the change and wait for further analysis results from the reporter. Cc: stable@vger.kernel.org Signed-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
D. Wythe authored
We found a data corruption issue during testing of SMC-R on Redis applications. The benchmark has a low probability of reporting a strange error as shown below. "Error: Protocol error, got "\xe2" as reply type byte" Finally, we found that the retrieved error data was as follows: 0xE2 0xD4 0xC3 0xD9 0x04 0x00 0x2C 0x20 0xA6 0x56 0x00 0x16 0x3E 0x0C 0xCB 0x04 0x02 0x01 0x00 0x00 0x20 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0xE2 It is quite obvious that this is a SMC DECLINE message, which means that the applications received SMC protocol message. We found that this was caused by the following situations: client server ¦ clc proposal -------------> ¦ clc accept <------------- ¦ clc confirm -------------> wait llc confirm send llc confirm ¦failed llc confirm ¦ x------ (after 2s)timeout wait llc confirm rsp wait decline (after 1s) timeout (after 2s) timeout ¦ decline --------------> ¦ decline <-------------- As a result, a decline message was sent in the implementation, and this message was read from TCP by the already-fallback connection. This patch double the client timeout as 2x of the server value, With this simple change, the Decline messages should never cross or collide (during Confirm link timeout). This issue requires an immediate solution, since the protocol updates involve a more long-term solution. Fixes: 0fb0b02b ("net/smc: adapt SMC client code to use the LLC flow") Signed-off-by:D. Wythe <alibuda@linux.alibaba.com> Reviewed-by:
Wen Gu <guwen@linux.alibaba.com> Reviewed-by:
Wenjia Zhang <wenjia@linux.ibm.com> Signed-off-by:
-
Nguyen Dinh Phi authored
syzbot reported an memory leak that happens when an skb is add to send_buff after virtual nci closed. This patch adds a variable to track if the ndev is running before handling new skb in send function. Signed-off-by:
Nguyen Dinh Phi <phind.uet@gmail.com> Reported-by: syzbot+6eb09d75211863f15e3e@syzkaller.appspotmail.com Closes: https://lore.kernel.org/lkml/00000000000075472b06007df4fb@google.com Reviewed-by: Bongsu Jeon Reviewed-by:
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by:
-
Hao Ge authored
We should clean the skb resource if genlmsg_put_reply failed. Fixes: 9d71b54b ("dpll: netlink: Add DPLL framework base functions") Signed-off-by:
Hao Ge <gehao@kylinos.cn> Reviewed-by:
Vadim Fedorenko <vadim.fedorenko@linux.dev> Link: https://lore.kernel.org/r/20231121013709.73323-1-gehao@kylinos.cnSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
- 21 Nov, 2023 21 commits
-
-
https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpfJakub Kicinski authored
Daniel Borkmann says: ==================== pull-request: bpf 2023-11-21 We've added 19 non-merge commits during the last 4 day(s) which contain a total of 18 files changed, 1043 insertions(+), 416 deletions(-). The main changes are: 1) Fix BPF verifier to validate callbacks as if they are called an unknown number of times in order to fix not detecting some unsafe programs, from Eduard Zingerman. 2) Fix bpf_redirect_peer() handling which missed proper stats accounting for veth and netkit and also generally fix missing stats for the latter, from Peilin Ye, Daniel Borkmann et al. * tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf: selftests/bpf: check if max number of bpf_loop iterations is tracked bpf: keep track of max number of bpf_loop callback iterations selftests/bpf: test widening for iterating callbacks bpf: widening for callback iterators selftests/bpf: tests for iterating callbacks bpf: verify callbacks as if they are called unknown number of times bpf: extract setup_func_entry() utility function bpf: extract __check_reg_arg() utility function selftests/bpf: fix bpf_loop_bench for new callback verification scheme selftests/bpf: track string payload offset as scalar in strobemeta selftests/bpf: track tcp payload offset as scalar in xdp_synproxy selftests/bpf: Add netkit to tc_redirect selftest selftests/bpf: De-veth-ize the tc_redirect test case bpf, netkit: Add indirect call wrapper for fetching peer dev bpf: Fix dev's rx stats for bpf_redirect_peer traffic veth: Use tstats per-CPU traffic counters netkit: Add tstats per-CPU traffic counters net: Move {l,t,d}stats allocation to core and convert veth & vrf net, vrf: Move dstats structure to core ==================== Link: https://lore.kernel.org/r/20231121193113.11796-1-daniel@iogearbox.netSigned-off-by:
-
Jakub Kicinski authored
There has been more than a few threads which went idle before the merge window and now people came back to them and started asking about next steps. We currently tell people to be patient and not to repost too often. Our "not too often", however, is still a few orders of magnitude faster than other subsystems. Or so I feel after hearing people talk about review rates at LPC. Clarify in the doc that if the discussion went idle for a week on netdev, 95% of the time there's no point waiting longer. Link: https://lore.kernel.org/r/20231120200109.620392-1-kuba@kernel.orgSigned-off-by:
-
Jose Ignacio Tornos Martinez authored
Using generic ASIX Electronics Corp. AX88179 Gigabit Ethernet device, the following test cycle has been implemented: - power on - check logs - shutdown - after detecting the system shutdown, disconnect power - after approximately 60 seconds of sleep, power is restored Running some cycles, sometimes error logs like this appear: kernel: ax88179_178a 2-9:1.0 (unnamed net_device) (uninitialized): Failed to write reg index 0x0001: -19 kernel: ax88179_178a 2-9:1.0 (unnamed net_device) (uninitialized): Failed to read reg index 0x0001: -19 ... These failed operation are happening during ax88179_reset execution, so the initialization could not be correct. In order to avoid this, we need to increase the delay after reset and clock initial operations. By using these larger values, many cycles have been run and no failed operations appear. It would be better to check some status register to verify when the operation has finished, but I do not have found any available information (neither in the public datasheets nor in the manufacturer's driver). The only available information for the necessary delays is the maufacturer's driver (original values) but the proposed values are not enough for the tested devices. Fixes: e2ca90c2 ("ax88179_178a: ASIX AX88179_178A USB 3.0/2.0 to gigabit ethernet adapter driver") Reported-by:Herb Wei <weihao.bj@ieisystem.com> Tested-by:
Jose Ignacio Tornos Martinez <jtornosm@redhat.com> Link: https://lore.kernel.org/r/20231120120642.54334-1-jtornosm@redhat.comSigned-off-by:
-
Simon Horman authored
indirect_call_wrapper.h is not, strictly speaking, networking specific. However, it's git history indicates that in practice changes go through netdev and thus the netdev maintainers have effectively been taking responsibility for it. Formalise this by adding it to the NETWORKING [GENERAL] section in the MAINTAINERS file. It is not clear how many other files under include/linux fall into this category and it would be interesting, as a follow-up, to audit that and propose further updates to the MAINTAINERS file as appropriate. Link: https://lore.kernel.org/netdev/20231116010310.4664dd38@kernel.org/Signed-off-by:
-
Paolo Abeni authored
Haiyang Zhang says: ==================== hv_netvsc: fix race of netvsc, VF register, and slave bit There are some races between netvsc probe, set notifier, VF register, and slave bit setting. This patch set fixes them. ==================== Link: https://lore.kernel.org/r/1700411023-14317-1-git-send-email-haiyangz@microsoft.comSigned-off-by:
-
Long Li authored
When a VF is being exposed form the kernel, it should be marked as "slave" before exposing to the user-mode. The VF is not usable without netvsc running as master. The user-mode should never see a VF without the "slave" flag. This commit moves the code of setting the slave flag to the time before VF is exposed to user-mode. Cc: stable@vger.kernel.org Fixes: 0c195567 ("netvsc: transparent VF management") Signed-off-by:
Long Li <longli@microsoft.com> Signed-off-by:
Haiyang Zhang <haiyangz@microsoft.com> Acked-by:
Stephen Hemminger <stephen@networkplumber.org> Acked-by:
Dexuan Cui <decui@microsoft.com> Signed-off-by:
-
Haiyang Zhang authored
If VF NIC is registered earlier, NETDEV_REGISTER event is replayed, but NETDEV_POST_INIT is not. Move register_netdevice_notifier() earlier, so the call back function is set before probing. Cc: stable@vger.kernel.org Fixes: e04e7a7b ("hv_netvsc: Fix a deadlock by getting rtnl lock earlier in netvsc_probe()") Reported-by:
-
Haiyang Zhang authored
The rtnl lock also needs to be held before rndis_filter_device_add() which advertises nvsp_2_vsc_capability / sriov bit, and triggers VF NIC offering and registering. If VF NIC finished register_netdev() earlier it may cause name based config failure. To fix this issue, move the call to rtnl_lock() before rndis_filter_device_add(), so VF will be registered later than netvsc / synthetic NIC, and gets a name numbered (ethX) after netvsc. Cc: stable@vger.kernel.org Fixes: e04e7a7b ("hv_netvsc: Fix a deadlock by getting rtnl lock earlier in netvsc_probe()") Reported-by:
-
Kunwu Chan authored
net/ipv4/route.c:783:46: warning: incorrect type in argument 2 (different base types) net/ipv4/route.c:783:46: expected unsigned int [usertype] key net/ipv4/route.c:783:46: got restricted __be32 [usertype] new_gw Fixes: 969447f2 ("ipv4: use new_gw for redirect neigh lookup") Suggested-by:
Eric Dumazet <edumazet@google.com> Signed-off-by:
Kunwu Chan <chentao@kylinos.cn> Link: https://lore.kernel.org/r/20231119141759.420477-1-chentao@kylinos.cnSigned-off-by:
-
Alexei Starovoitov authored
Eduard Zingerman says: ==================== verify callbacks as if they are called unknown number of times This series updates verifier logic for callback functions handling. Current master simulates callback body execution exactly once, which leads to verifier not detecting unsafe programs like below: static int unsafe_on_zero_iter_cb(__u32 idx, struct num_context *ctx) { ctx->i = 0; return 0; } SEC("?raw_tp") int unsafe_on_zero_iter(void *unused) { struct num_context loop_ctx = { .i = 32 }; __u8 choice_arr[2] = { 0, 1 }; bpf_loop(100, unsafe_on_zero_iter_cb, &loop_ctx, 0); return choice_arr[loop_ctx.i]; } This was reported previously in [0]. The basic idea of the fix is to schedule callback entry state for verification in env->head until some identical, previously visited state in current DFS state traversal is found. Same logic as with open coded iterators, and builds on top recent fixes [1] for those. The series is structured as follows: - patches #1,2,3 update strobemeta, xdp_synproxy selftests and bpf_loop_bench benchmark to allow convergence of the bpf_loop callback states; - patches #4,5 just shuffle the code a bit; - patch #6 is the main part of the series; - patch #7 adds test cases for #6; - patch #8 extend patch #6 with same speculative scalar widening logic, as used for open coded iterators; - patch #9 adds test cases for #8; - patch #10 extends patch #6 to track maximal number of callback executions specifically for bpf_loop(); - patch #11 adds test cases for #10. Veristat results comparing this series to master+patches #1,2,3 using selftests show the following difference: File Program States (A) States (B) States (DIFF) ------------------------- ------------- ---------- ---------- ------------- bpf_loop_bench.bpf.o benchmark 1 2 +1 (+100.00%) pyperf600_bpf_loop.bpf.o on_event 322 407 +85 (+26.40%) strobemeta_bpf_loop.bpf.o on_event 113 151 +38 (+33.63%) xdp_synproxy_kern.bpf.o syncookie_tc 341 291 -50 (-14.66%) xdp_synproxy_kern.bpf.o syncookie_xdp 344 301 -43 (-12.50%) Veristat results comparing this series to master using Tetragon BPF files [2] also show some differences. States diff varies from +2% to +15% on 23 programs out of 186, no new failures. Changelog: - V3 [5] -> V4, changes suggested by Andrii: - validate mark_chain_precision() result in patch #10; - renaming s/cumulative_callback_depth/callback_unroll_depth/. - V2 [4] -> V3: - fixes in expected log messages for test cases: - callback_result_precise; - parent_callee_saved_reg_precise_with_callback; - parent_stack_slot_precise_with_callback; - renamings (suggested by Alexei): - s/callback_iter_depth/cumulative_callback_depth/ - s/is_callback_iter_next/calls_callback/ - s/mark_callback_iter_next/mark_calls_callback/ - prepare_func_exit() updated to exit with -EFAULT when callee->in_callback_fn is true but calls_callback() is not true for callsite; - test case 'bpf_loop_iter_limit_nested' rewritten to use return value check instead of verifier log message checks (suggested by Alexei). - V1 [3] -> V2, changes suggested by Andrii: - small changes for error handling code in __check_func_call(); - callback body processing log is now matched in relevant verifier_subprog_precision.c tests; - R1 passed to bpf_loop() is now always marked as precise; - log level 2 message for bpf_loop() iteration termination instead of iteration depth messages; - __no_msg macro removed; - bpf_loop_iter_limit_nested updated to avoid using __no_msg; - commit message for patch #3 updated according to Alexei's request. [0] https://lore.kernel.org/bpf/CA+vRuzPChFNXmouzGG+wsy=6eMcfr1mFG0F3g7rbg-sedGKW3w@mail.gmail.com/ [1] https://lore.kernel.org/bpf/20231024000917.12153-1-eddyz87@gmail.com/ [2] git@github.com:cilium/tetragon.git [3] https://lore.kernel.org/bpf/20231116021803.9982-1-eddyz87@gmail.com/T/#t [4] https://lore.kernel.org/bpf/20231118013355.7943-1-eddyz87@gmail.com/T/#t [5] https://lore.kernel.org/bpf/20231120225945.11741-1-eddyz87@gmail.com/T/#t ==================== Link: https://lore.kernel.org/r/20231121020701.26440-1-eddyz87@gmail.comSigned-off-by:Alexei Starovoitov <ast@kernel.org>
-
Eduard Zingerman authored
Check that even if bpf_loop() callback simulation does not converge to a specific state, verification could proceed via "brute force" simulation of maximal number of callback calls. Signed-off-by:
Eduard Zingerman <eddyz87@gmail.com> Link: https://lore.kernel.org/r/20231121020701.26440-12-eddyz87@gmail.comSigned-off-by:
-
Eduard Zingerman authored
In some cases verifier can't infer convergence of the bpf_loop() iteration. E.g. for the following program: static int cb(__u32 idx, struct num_context* ctx) { ctx->i++; return 0; } SEC("?raw_tp") int prog(void *_) { struct num_context ctx = { .i = 0 }; __u8 choice_arr[2] = { 0, 1 }; bpf_loop(2, cb, &ctx, 0); return choice_arr[ctx.i]; } Each 'cb' simulation would eventually return to 'prog' and reach 'return choice_arr[ctx.i]' statement. At which point ctx.i would be marked precise, thus forcing verifier to track multitude of separate states with {.i=0}, {.i=1}, ... at bpf_loop() callback entry. This commit allows "brute force" handling for such cases by limiting number of callback body simulations using 'umax' value of the first bpf_loop() parameter. For this, extend bpf_func_state with 'callback_depth' field. Increment this field when callback visiting state is pushed to states traversal stack. For frame #N it's 'callback_depth' field counts how many times callback with frame depth N+1 had been executed. Use bpf_func_state specifically to allow independent tracking of callback depths when multiple nested bpf_loop() calls are present. Signed-off-by: -
Eduard Zingerman authored
A test case to verify that imprecise scalars widening is applied to callback entering state, when callback call is simulated repeatedly. Signed-off-by:
-
Eduard Zingerman authored
Callbacks are similar to open coded iterators, so add imprecise widening logic for callback body processing. This makes callback based loops behave identically to open coded iterators, e.g. allowing to verify programs like below: struct ctx { u32 i; }; int cb(u32 idx, struct ctx* ctx) { ++ctx->i; return 0; } ... struct ctx ctx = { .i = 0 }; bpf_loop(100, cb, &ctx, 0); ... Acked-by:Andrii Nakryiko <andrii@kernel.org> Signed-off-by:
-
Eduard Zingerman authored
A set of test cases to check behavior of callback handling logic, check if verifier catches the following situations: - program not safe on second callback iteration; - program not safe on zero callback iterations; - infinite loop inside a callback. Verify that callback logic works for bpf_loop, bpf_for_each_map_elem, bpf_user_ringbuf_drain, bpf_find_vma. Acked-by:
-
Eduard Zingerman authored
Prior to this patch callbacks were handled as regular function calls, execution of callback body was modeled exactly once. This patch updates callbacks handling logic as follows: - introduces a function push_callback_call() that schedules callback body verification in env->head stack; - updates prepare_func_exit() to reschedule callback body verification upon BPF_EXIT; - as calls to bpf_*_iter_next(), calls to callback invoking functions are marked as checkpoints; - is_state_visited() is updated to stop callback based iteration when some identical parent state is found. Paths with callback function invoked zero times are now verified first, which leads to necessity to modify some selftests: - the following negative tests required adding release/unlock/drop calls to avoid previously masked unrelated error reports: - cb_refs.c:underflow_prog - exceptions_fail.c:reject_rbtree_add_throw - exceptions_fail.c:reject_with_cp_reference - the following precision tracking selftests needed change in expected log trace: - verifier_subprog_precision.c:callback_result_precise (note: r0 precision is no longer propagated inside callback and I think this is a correct behavior) - verifier_subprog_precision.c:parent_callee_saved_reg_precise_with_callback - verifier_subprog_precision.c:parent_stack_slot_precise_with_callback Reported-by:Andrew Werner <awerner32@gmail.com> Closes: https://lore.kernel.org/bpf/CA+vRuzPChFNXmouzGG+wsy=6eMcfr1mFG0F3g7rbg-sedGKW3w@mail.gmail.com/Acked-by:
-
Eduard Zingerman authored
Move code for simulated stack frame creation to a separate utility function. This function would be used in the follow-up change for callbacks handling. Acked-by:
-
Eduard Zingerman authored
Split check_reg_arg() into two utility functions: - check_reg_arg() operating on registers from current verifier state; - __check_reg_arg() operating on a specific set of registers passed as a parameter; The __check_reg_arg() function would be used by a follow-up change for callbacks handling. Acked-by:
-
Eduard Zingerman authored
This is a preparatory change. A follow-up patch "bpf: verify callbacks as if they are called unknown number of times" changes logic for callbacks handling. While previously callbacks were verified as a single function call, new scheme takes into account that callbacks could be executed unknown number of times. This has dire implications for bpf_loop_bench: SEC("fentry/" SYS_PREFIX "sys_getpgid") int benchmark(void *ctx) { for (int i = 0; i < 1000; i++) { bpf_loop(nr_loops, empty_callback, NULL, 0); __sync_add_and_fetch(&hits, nr_loops); } return 0; } W/o callbacks change verifier sees it as a 1000 calls to empty_callback(). However, with callbacks change things become exponential: - i=0: state exploring empty_callback is scheduled with i=0 (a); - i=1: state exploring empty_callback is scheduled with i=1; ... - i=999: state exploring empty_callback is scheduled with i=999; - state (a) is popped from stack; - i=1: state exploring empty_callback is scheduled with i=1; ... Avoid this issue by rewriting outer loop as bpf_loop(). Unfortunately, this adds a function call to a loop at runtime, which negatively affects performance: throughput latency before: 149.919 ± 0.168 M ops/s, 6.670 ns/op after : 137.040 ± 0.187 M ops/s, 7.297 ns/op Acked-by: -
Eduard Zingerman authored
This change prepares strobemeta for update in callbacks verification logic. To allow bpf_loop() verification converge when multiple callback iterations are considered: - track offset inside strobemeta_payload->payload directly as scalar value; - at each iteration make sure that remaining strobemeta_payload->payload capacity is sufficient for execution of read_{map,str}_var functions; - make sure that offset is tracked as unbound scalar between iterations, otherwise verifier won't be able infer that bpf_loop callback reaches identical states. Acked-by: -
Eduard Zingerman authored
This change prepares syncookie_{tc,xdp} for update in callbakcs verification logic. To allow bpf_loop() verification converge when multiple callback itreations are considered: - track offset inside TCP payload explicitly, not as a part of the pointer; - make sure that offset does not exceed MAX_PACKET_OFF enforced by verifier; - make sure that offset is tracked as unbound scalar between iterations, otherwise verifier won't be able infer that bpf_loop callback reaches identical states. Acked-by:
-
- 20 Nov, 2023 9 commits
-
-
Martin KaFai Lau authored
Daniel Borkmann says: ==================== This fixes bpf_redirect_peer stats accounting for veth and netkit, and adds tstats in the first place for the latter. Utilise indirect call wrapper for bpf_redirect_peer, and improve test coverage of the latter also for netkit devices. Details in the patches, thanks! The series was targeted at bpf originally, and is done here as well, so it can trigger BPF CI. Jakub, if you think directly going via net is better since the majority of the diff touches net anyway, that is fine, too. Thanks! v2 -> v3: - Add kdoc for pcpu_stat_type (Simon) - Reject invalid type value in netdev_do_alloc_pcpu_stats (Simon) - Add Reviewed-by tags from list v1 -> v2: - Move stats allocation/freeing into net core (Jakub) - As prepwork for the above, move vrf's dstats over into the core - Add a check into stats alloc to enforce tstats upon implementing ndo_get_peer_dev - Add Acked-by tags from list Daniel Borkmann (6): net, vrf: Move dstats structure to core net: Move {l,t,d}stats allocation to core and convert veth & vrf netkit: Add tstats per-CPU traffic counters bpf, netkit: Add indirect call wrapper for fetching peer dev selftests/bpf: De-veth-ize the tc_redirect test case selftests/bpf: Add netkit to tc_redirect selftest ==================== Signed-off-by:Martin KaFai Lau <martin.lau@kernel.org>
-
Daniel Borkmann authored
Extend the existing tc_redirect selftest to also cover netkit devices for exercising the bpf_redirect_peer() code paths, so that we have both veth as well as netkit covered, all tests still pass after this change. Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Acked-by:
Stanislav Fomichev <sdf@google.com> Reviewed-by:
Nikolay Aleksandrov <razor@blackwall.org> Link: https://lore.kernel.org/r/20231114004220.6495-9-daniel@iogearbox.netSigned-off-by:
-
Daniel Borkmann authored
No functional changes to the test case, but just renaming various functions, variables, etc, to remove veth part of their name for making it more generic and reusable later on (e.g. for netkit). Signed-off-by:
-
Daniel Borkmann authored
ndo_get_peer_dev is used in tcx BPF fast path, therefore make use of indirect call wrapper and therefore optimize the bpf_redirect_peer() internal handling a bit. Add a small skb_get_peer_dev() wrapper which utilizes the INDIRECT_CALL_1() macro instead of open coding. Future work could potentially add a peer pointer directly into struct net_device in future and convert veth and netkit over to use it so that eventually ndo_get_peer_dev can be removed. Co-developed-by:
-
Peilin Ye authored
Traffic redirected by bpf_redirect_peer() (used by recent CNIs like Cilium) is not accounted for in the RX stats of supported devices (that is, veth and netkit), confusing user space metrics collectors such as cAdvisor [0], as reported by Youlun. Fix it by calling dev_sw_netstats_rx_add() in skb_do_redirect(), to update RX traffic counters. Devices that support ndo_get_peer_dev _must_ use the @tstats per-CPU counters (instead of @lstats, or @dstats). To make this more fool-proof, error out when ndo_get_peer_dev is set but @tstats are not selected. [0] Specifically, the "container_network_receive_{byte,packet}s_total" counters are affected. Fixes: 9aa1206e ("bpf: Add redirect_peer helper") Reported-by:Youlun Zhang <zhangyoulun@bytedance.com> Signed-off-by:
Peilin Ye <peilin.ye@bytedance.com> Co-developed-by:
-
Peilin Ye authored
Currently veth devices use the lstats per-CPU traffic counters, which only cover TX traffic. veth_get_stats64() actually populates RX stats of a veth device from its peer's TX counters, based on the assumption that a veth device can _only_ receive packets from its peer, which is no longer true: For example, recent CNIs (like Cilium) can use the bpf_redirect_peer() BPF helper to redirect traffic from NIC's tc ingress to veth's tc ingress (in a different netns), skipping veth's peer device. Unfortunately, this kind of traffic isn't currently accounted for in veth's RX stats. In preparation for the fix, use tstats (instead of lstats) to maintain both RX and TX counters for each veth device. We'll use RX counters for bpf_redirect_peer() traffic, and keep using TX counters for the usual "peer-to-peer" traffic. In veth_get_stats64(), calculate RX stats by _adding_ RX count to peer's TX count, in order to cover both kinds of traffic. veth_stats_rx() might need a name change (perhaps to "veth_stats_xdp()") for less confusion, but let's leave it to another patch to keep the fix minimal. Signed-off-by:
-
Daniel Borkmann authored
Add dev->tstats traffic accounting to netkit. The latter contains per-CPU RX and TX counters. The dev's TX counters are bumped upon pass/unspec as well as redirect verdicts, in other words, on everything except for drops. The dev's RX counters are bumped upon successful __netif_rx(), as well as from skb_do_redirect() (not part of this commit here). Using dev->lstats with having just a single packets/bytes counter and inferring one another's RX counters from the peer dev's lstats is not possible given skb_do_redirect() can also bump the device's stats. Signed-off-by:
-
Daniel Borkmann authored
Move {l,t,d}stats allocation to the core and let netdevs pick the stats type they need. That way the driver doesn't have to bother with error handling (allocation failure checking, making sure free happens in the right spot, etc) - all happening in the core. Co-developed-by: -
Daniel Borkmann authored
Just move struct pcpu_dstats out of the vrf into the core, and streamline the field names slightly, so they better align with the {t,l}stats ones. No functional change otherwise. A conversion of the u64s to u64_stats_t could be done at a separate point in future. This move is needed as we are moving the {t,l,d}stats allocation/freeing to the core. Signed-off-by:
-
