iavf: XDP Page Pool edition #17

alobakin · 2023-02-24T13:05:23Z

XDP and XSk parts are untested, has some TODOs to finish + lacks latest improvements and fixes from bpf-next's ice.

Not a secret there's a ton of code duplication between two and more Intel ethernet modules. Before introducing new changes, which would need to be copied over again, start decoupling the already existing duplicate functionality into a new module, which will be shared between several Intel Ethernet drivers. Add the lookup table which converts 8/10-bit hardware packet type into a parsed bitfield structure for easy checking packet format parameters, such as payload level, IP version, etc. This is currently used by i40e, ice and iavf and it's all the same in all three drivers. The only difference introduced in this implementation is that instead of defining a 256 (or 1024 in case of ice) element array, add unlikely() condition to limit the input to 154 (current maximum non-reserved packet type). There's no reason to waste 600 (or even 3600) bytes only to not hurt very unlikely exception packets. The hash computation function now takes payload level directly as a pkt_hash_type. There's a couple cases when non-IP ptypes are marked as L3 payload and in the previous versions their hash level would be 2, not 3. But skb_set_hash() only sees difference between L4 and non-L4, thus this won't change anything at all. The module is behind the hidden Kconfig symbol, which the drivers will select when needed. The exports are behind 'LIBIE' namespace to limit the scope of the functions. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Ever since build_skb() became stable, the old way with allocating an skb for storing the headers separately, which will be then copied manually, was slower, less flexible and thus obsolete. * it had higher pressure on MM since it actually allocates new pages, which then get split and refcount-biased (NAPI page cache); * it implies memcpy() of packet headers (40+ bytes per each frame); * the actual header length was calculated via eth_get_headlen(), which invokes Flow Dissector and thus wastes a bunch of CPU cycles; * XDP makes it even more weird since it requires headroom for long and also tailroom for some time (since mbuf landed). Take a look at the ice driver, which is built around work-arounds to make XDP work with it. Even on some quite low-end hardware (not a common case for 100G NICs) it was performing worse. The only advantage "legacy-rx" had is that it didn't require any reserved headroom and tailroom. But iavf didn't use this, as it always splits pages into two halves of 2k, while that save would only be useful when striding. And again, XDP effectively removes that sole pro. There's a train of features to land in IAVF soon: Page Pool, XDP, XSk, multi-buffer etc. Each new would require adding more and more Danse Macabre for absolutely no reason, besides making hotpath less and less effective. Remove the "feature" with all the related code. This includes at least one very hot branch (typically hit on each new frame), which was either always-true or always-false at least for a complete NAPI bulk of 64 frames, the whole private flags cruft and so on. Some stats: Function: add/remove: 0/2 grow/shrink: 0/7 up/down: 0/-774 (-774) RO Data: add/remove: 0/1 grow/shrink: 0/0 up/down: 0/-40 (-40) Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

@gfp

The Rx hotpath code of IAVF is not well-optimized TBH. Before doing any further buffer model changes, shake it up a bit. Notably: 1. Cache more variables on the stack. DMA device, Rx page size, NTC -- these are the most common things used all throughout the hotpath, often in loops on each iteration. Instead of fetching (or even calculating, as with the page size) them from the ring all the time, cache them on the stack at the beginning of the NAPI polling callback. NTC will be written back at the end, the rest are used read-only, so no sync needed. 2. Don't move the recycled buffers around the ring. The idea of passing the page of the right-now-recycled-buffer to a different buffer, in this case, the first one that needs to be allocated, moreover, on each new frame, is fundamentally wrong. It involves a few o' fetches, branches and then writes (and one Rx buffer struct is at least 32 bytes) where they're completely unneeded, but gives no good -- the result is the same as if we'd recycle it inplace, at the same position where it was used. So drop this and let the main refilling function take care of all the buffers, which were processed and now need to be recycled/refilled. 3. Don't allocate with %GPF_ATOMIC on ifup. This involved introducing the @gfp parameter to a couple functions. Doesn't change anything for Rx -> softirq. 4. 1 budget unit == 1 descriptor, not skb. There could be underflow when receiving a lot of fragmented frames. If each of them would consist of 2 frags, it means that we'd process 64 descriptors at the point where we pass the 32th skb to the stack. But the driver would count that only as a half, which could make NAPI re-enable interrupts prematurely and create unnecessary CPU load. 5. Shortcut !size case. It's super rare, but possible -- for example, if the last buffer of the fragmented frame contained only FCS, which was then stripped by the HW. Instead of checking for size several times when processing, quickly reuse the buffer and jump to the skb fields part. 6. Refill the ring after finishing the polling loop. Previously, the loop wasn't starting a new iteration after the 64th desc, meaning that we were always leaving 16 buffers non-refilled until the next NAPI poll. It's better to refill them while they're still hot, so do that right after exiting the loop as well. For a full cycle of 64 descs, there will be 4 refills of 16 descs from now on. Function: add/remove: 4/2 grow/shrink: 0/5 up/down: 473/-647 (-174) + up to 2% performance. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

As an intermediate step, remove all page splitting/recyclig code. Just always allocate a new page and don't touch its refcount, so that it gets freed by the core stack later. The change allows to greatly simplify certain parts of the code: Function: add/remove: 2/3 grow/shrink: 0/5 up/down: 543/-963 (-420) &iavf_rx_buf can even now retire in favor of just storing an array of pages used for Rx. Their DMA addresses can be stored in page::dma_addr -- use Page Pool's function for that. No surprise perf loses up to 30% here, but that regression will go away once PP lands. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

The current scheme with trying to pick the smallest buffer possible for the current MTU in order to flip/split pages is not very optimal. For example, on default MTU of 1500 it gives only 192 bytes of headroom, while XDP may require up to 258. But this also involves unnecessary code complication, which sometimes is even hard to follow. As page split is no more, always allocate order-0 pages. This optimizes performance a bit and drops some bytes off the object code. Next, always pick the maximum buffer length available for this %PAGE_SIZE to set it up in the hardware. This means it now becomes a constant value, which also has its positive impact. On x64 this means (without XDP): 4096 page 64 head, 320 tail 3712 HW buffer size 3686 max MTU w/o frags Previously, the maximum MTU w/o splitting a frame into several buffers was 3046. Increased buffer size allows us to reach the maximum frame size w/ frags supported by HW: 16382 bytes (MTU 16356). Reflect it in the netdev config as well. Relying on max single buffer size when calculating MTU was not correct. Move around a couple of fields in &iavf_ring after ::rx_buf_len removal to reduce holes and improve cache locality. Instead of providing the Rx definitions, which can and will be reused in rest of the drivers, exclusively for IAVF, do that in the libie header. Non-PP drivers could still use at least some of them and lose a couple copied lines. Function: add/remove: 0/0 grow/shrink: 3/9 up/down: 18/-265 (-247) + even reclaims a half percent of performance, nice. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Add a new flag, %PP_FLAG_DMA_MAP_WEAK, whill will tell PP to map pages with %DMA_ATTR_WEAK_ORDERING. To keep the code simple and optimized, map the following PP flags to DMA map attr flags: %PP_FLAG_DMA_MAP => %DMA_ATTR_SKIP_CPU_SYNC %PP_FLAG_DMA_MAP_WEAK => %DMA_ATTR_WEAK_ORDERING The first pair is done to be able to just pass it directly to dma_map_page_attrs(). When a driver wants Page Pool to maintain DMA mappings, it always sets this flag. Page Pool always skips CPU syncs when mapping to do that separately later, so having those two 1:1 avoids introducing ifs and/or bit-ors and keeps the code more compact. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Each driver is responsible for syncing buffers written by HW for CPU before accessing them. Almost each PP-enabled driver uses the same pattern, which could be shorthanded into a static inline to make driver code a little bit more compact. Introduce a pair of such functions. The first one takes the actual size of the data written by HW and is the main one to be used on Rx. The second picks max_len from the PP params and is designed for more extreme cases when the size is unknown, but the buffer still needs to be synced. Also constify pointer arguments of page_pool_get_dma_dir() and page_pool_get_dma_addr() to give a bit more room for optimization, as both of them are read-only. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Now that the IAVF driver simply uses dev_alloc_page() + free_page() with no custom recycling logics and one whole page per frame, it can easily be switched to using Page Pool API instead. Introduce libie_rx_page_pool_create(), a wrapper for creating a PP with the default libie settings applicable to all Intel hardware, and replace the alloc/free calls with the corresponding PP functions, including the newly added sync-for-CPU helpers. Use skb_mark_for_recycle() to bring back the recycling and restore the initial performance. From the important object code changes, worth mentioning that __iavf_alloc_rx_pages() is now inlined due to the greatly reduced size. The resulting driver is on par with the pre-series code and 1-2% slower than the "optimized" version right before the recycling removal. But the number of locs and object code bytes slaughtered is much more important here after all, not speaking of that there's still a vast space for optimization and improvements. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Next stop, per-queue private stats. They have only subtle differences from driver to driver and can easily be resolved. Define common structures, inline helpers and Ethtool helpers to collect, update and export the statistics. Use u64_stats_t right from the start, as well as the corresponding helpers to ensure tear-free operations. For the NAPI parts of both Rx and Tx, also define small onstack containers to update them in polling loops and then sync the actual containers once a loop ends. The drivers will be switched to use this API later on a per-driver basis, along with conversion to PP. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Expand the libie generic per-queue stats with the generic Page Pool stats provided by the API itself, when CONFIG_PAGE_POOL is enable. When it's not, there'll be no such fields in the stats structure, so no space wasted. They are also a bit special in terms of how they are obtained. One &page_pool accumulates statistics until it's destroyed obviously, which happens on ifdown. So, in order to not lose any statistics, get the stats and store in the queue container before destroying a pool. This container survives ifups/downs, so it basically stores the statistics accumulated since the very first pool was allocated on this queue. When it's needed to export the stats, first get the numbers from this container and then add the "live" numbers -- the ones that the current active pool returns. The result values will always represent the actual device-lifetime* stats. There's a cast from &page_pool_stats to `u64 *` in a couple functions, but they are guarded with stats asserts to make sure it's safe to do. FWIW it saves a lot of object code. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

iavf is pretty much ready for using the generic libie stats, so drop all the custom code and just use generic definitions. The only thing is that it previously lacked the counter of Tx queue stops. It's present in the other drivers, so add it here as well. The rest is straightforward. There were two fields in the Tx stats struct, which didn't belong there. The first one has never been used, wipe it; and move the other to the queue structure. Plus move around a couple fields in &iavf_ring to account stats structs' alignment. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Currently, the test relies on that only dropped ("xmitted") frames will be recycled and if a frame became an skb, it will be freed later by the stack and never come back to its page_pool. So, it easily gets broken by trying to recycle skbs: test_xdp_do_redirect:PASS:pkt_count_xdp 0 nsec test_xdp_do_redirect:FAIL:pkt_count_zero unexpected pkt_count_zero: actual 9936 != expected 2 test_xdp_do_redirect:PASS:pkt_count_tc 0 nsec That huge mismatch happened because after the TC ingress hook zeroes the magic, the page gets recycled when skb is freed, not returned to the MM layer. "Live frames" mode initializes only new pages and keeps the recycled ones as is by design, so they appear with zeroed magic on the Rx path again. Expand the possible magic values from two: 0 (was "xmitted"/dropped or did hit the TC hook) and 0x42 (hit the input XDP prog) to three: the new one will mark frames hit the TC hook, so that they will elide both @pkt_count_zero and @pkt_count_xdp. They can then be recycled to their page_pool or returned to the page allocator, this won't affect the counters anyhow. Just make sure to mark them as "input" (0x42) when they appear on the Rx path again. Also make an enum from those magics, so that they will be always visible and can be changed in just one place anytime. This also eases adding any new marks later on. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

skb_mark_for_recycle() is guarded with CONFIG_PAGE_POOL, this creates unneeded complication when using it in the generic code. For now, it's only used in the drivers always selecting Page Pool, so this works. Move the guards so that preprocessor will cut out only the operation itself and the function will still be a noop on !PAGE_POOL systems, but available there as well. No functional changes. Reported-by: kernel test robot <lkp@intel.com> Link: https://lore.kernel.org/oe-kbuild-all/202303020342.Wi2PRFFH-lkp@intel.com Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

__xdp_build_skb_from_frame() state(d): /* Until page_pool get SKB return path, release DMA here */ Page Pool got skb pages recycling in April 2021, but missed this function. xdp_release_frame() is relevant only for Page Pool backed frames and it detaches the page from the corresponding page_pool in order to make it freeable via page_frag_free(). It can instead just mark the output skb as eligible for recycling if the frame is backed by a pp. No change for other memory model types (the same condition check as before). cpumap redirect and veth on Page Pool drivers now become zero-alloc (or almost). Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

__xdp_build_skb_from_frame() was the last user of {__,}xdp_release_frame(), which detaches pages from the page_pool. All the consumers now recycle Page Pool skbs and page, except mlx5, stmmac and tsnep drivers, which use page_pool_release_page() directly (might change one day). It's safe to assume this functionality is not needed anymore and can be removed (in favor of recycling). Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

The current version of Intel 'ice' driver allows for using zero for the ring lenghth in 'configure queue' VIRTCHNL message. Such a value indicates the ring should not be configured. Implement the same handling in i40e driver. Instead of returning an 'invalid parameter' error for zero-sized rings, just skip that ring during queue pair configuration. That unified handling is needed for AF_XDP implementation for 'iavf' driver. In that use case we sometimes need to configure Tx ring only for a given queue pair. Signed-off-by: Michal Kubiak <michal.kubiak@intel.com>

The XDP and AF_XDP feature is initialized using .ndo functions. Those functions are always synchronous and may require some serious queues reconfiguration including changing the number of queues. Performing such a reconfiguration implies sending a bunch of VIRTCHNL messages to the PF in order to disable queues, re-enable and re-configure them, or update the RSS LUT. By definition, those VIRTCHNL messages are sent asynchronously, so the result of each VIRTCHNL operation can be received from the PF via admin queue after some time. Moreover, the previous implementation of some VIRTCHNL functions (e.g. 'iavf_disable_queues()' or 'iavf_enable_queues()' does not allow to call them selectively for specific queues only. In order to addres those problems and cover all scenarios of XDP and AF_XDP initialization, implement a polling mechanism with a timeout for blocking the execution of XDP .ndo functions until the result of VIRTCHNL operation on PF is known to the driver. Also, refactor the existing VIRTCHNL API by adding functions for selective queue enabling, disabling and configuration. Signed-off-by: Michal Kubiak <michal.kubiak@intel.com>

Port of commit 22bf877 ("ice: introduce XDP_TX fallback path"). The patch handles the case, when queue number is not sufficient for the current number of CPUs. To avoid dropping some packets redirected from other interfaces, XDP TxQs are allowed to be shared between CPUs, which imposes the locking requirement. Static key approach has little to none performance penalties when sharing is not needed. This mechanism is much more applicable when dealing with VFs. In fact, maximum number of queue pairs that ice PF can give to an iavf VF is 16, which allows up to 8 XDP TxQs, so without XDP TxQ sharing, some redirected packets can be dropped even on a 10 CPU system. Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>

Add necessary functions and data structures to support AF_XDP feature. Implement handling of 'XDP_SETUP_XSK_POOL' in .ndo_bpf(). Also, implement functions for selectively stopping only those queues which take part in XDP socket creation. Signed-off-by: Michal Kubiak <michal.kubiak@intel.com>

Implement Tx handling for AF_XDP feature in zero-copy mode. Add '.ndo_xdp_xmit()' and '.ndo_xsk_wakeup()' implementations to support AF_XDP Tx path. Also, add Tx interrupt handling function for zero-copy mode. Signed-off-by: Michal Kubiak <michal.kubiak@intel.com>

Implement RX packet processing specific to AF_XDP ZC. All actions except XDP_PASS are supported, the skb path will be implemented in later patches. Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>

For now, filling the skb fields on Rx is a bit scattered across RQ polling function. This makes it harder to reuse the code on XSk Rx path and also sometimes costs some CPU (e.g. doing a lookup for the decoded packet type two times). Make it consistent and do everything in iavf_process_skb_fields(). First of all, get the packet type and decode it. Then, move to hash, csum and VLAN, which is moved here too. iavf_receive_skb() becomes then the classic eth_type_trans() + napi_gro_receive() pair. Finally, make the fields processing function global and the skb receive function static inline in order to call them from a different file later on. Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Construct skb and fill in its fields, when AF_XDP is enabled on the ring, if XDP program returns XDP_PASS. (will be fixed up). Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>

Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

Upon physical link change, firmware reports to the kernel about the change along with the details like speed, lmac_type_id, etc. Kernel derives lmac_type based on lmac_type_id received from firmware. In a few scenarios, firmware returns an invalid lmac_type_id, which is resulting in below kernel panic. This patch adds the missing validation of the lmac_type_id field. Internal error: Oops: 96000005 [#1] PREEMPT SMP [ 35.321595] Modules linked in: [ 35.328982] CPU: 0 PID: 31 Comm: kworker/0:1 Not tainted 5.4.210-g2e3169d8e1bc-dirty #17 [ 35.337014] Hardware name: Marvell CN103XX board (DT) [ 35.344297] Workqueue: events work_for_cpu_fn [ 35.352730] pstate: 40400089 (nZcv daIf +PAN -UAO) [ 35.360267] pc : strncpy+0x10/0x30 [ 35.366595] lr : cgx_link_change_handler+0x90/0x180 Fixes: 61071a8 ("octeontx2-af: Forward CGX link notifications to PFs") Signed-off-by: Hariprasad Kelam <hkelam@marvell.com> Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com> Signed-off-by: Sai Krishna <saikrishnag@marvell.com> Reviewed-by: Simon Horman <simon.horman@corigine.com> Signed-off-by: David S. Miller <davem@davemloft.net>

The cited commit adds a compeletion to remove dependency on rtnl lock. But it causes a deadlock for multiple encapsulations: crash> bt ffff8aece8a64000 PID: 1514557 TASK: ffff8aece8a64000 CPU: 3 COMMAND: "tc" #0 [ffffa6d14183f368] __schedule at ffffffffb8ba7f45 #1 [ffffa6d14183f3f8] schedule at ffffffffb8ba8418 #2 [ffffa6d14183f418] schedule_preempt_disabled at ffffffffb8ba8898 #3 [ffffa6d14183f428] __mutex_lock at ffffffffb8baa7f8 #4 [ffffa6d14183f4d0] mutex_lock_nested at ffffffffb8baabeb #5 [ffffa6d14183f4e0] mlx5e_attach_encap at ffffffffc0f48c17 [mlx5_core] #6 [ffffa6d14183f628] mlx5e_tc_add_fdb_flow at ffffffffc0f39680 [mlx5_core] #7 [ffffa6d14183f688] __mlx5e_add_fdb_flow at ffffffffc0f3b636 [mlx5_core] #8 [ffffa6d14183f6f0] mlx5e_tc_add_flow at ffffffffc0f3bcdf [mlx5_core] #9 [ffffa6d14183f728] mlx5e_configure_flower at ffffffffc0f3c1d1 [mlx5_core] #10 [ffffa6d14183f790] mlx5e_rep_setup_tc_cls_flower at ffffffffc0f3d529 [mlx5_core] #11 [ffffa6d14183f7a0] mlx5e_rep_setup_tc_cb at ffffffffc0f3d714 [mlx5_core] #12 [ffffa6d14183f7b0] tc_setup_cb_add at ffffffffb8931bb8 #13 [ffffa6d14183f810] fl_hw_replace_filter at ffffffffc0dae901 [cls_flower] #14 [ffffa6d14183f8d8] fl_change at ffffffffc0db5c57 [cls_flower] #15 [ffffa6d14183f970] tc_new_tfilter at ffffffffb8936047 #16 [ffffa6d14183fac8] rtnetlink_rcv_msg at ffffffffb88c7c31 #17 [ffffa6d14183fb50] netlink_rcv_skb at ffffffffb8942853 #18 [ffffa6d14183fbc0] rtnetlink_rcv at ffffffffb88c1835 #19 [ffffa6d14183fbd0] netlink_unicast at ffffffffb8941f27 #20 [ffffa6d14183fc18] netlink_sendmsg at ffffffffb8942245 #21 [ffffa6d14183fc98] sock_sendmsg at ffffffffb887d482 #22 [ffffa6d14183fcb8] ____sys_sendmsg at ffffffffb887d81a #23 [ffffa6d14183fd38] ___sys_sendmsg at ffffffffb88806e2 #24 [ffffa6d14183fe90] __sys_sendmsg at ffffffffb88807a2 #25 [ffffa6d14183ff28] __x64_sys_sendmsg at ffffffffb888080f #26 [ffffa6d14183ff38] do_syscall_64 at ffffffffb8b9b6a8 #27 [ffffa6d14183ff50] entry_SYSCALL_64_after_hwframe at ffffffffb8c0007c crash> bt 0xffff8aeb07544000 PID: 1110766 TASK: ffff8aeb07544000 CPU: 0 COMMAND: "kworker/u20:9" #0 [ffffa6d14e6b7bd8] __schedule at ffffffffb8ba7f45 #1 [ffffa6d14e6b7c68] schedule at ffffffffb8ba8418 #2 [ffffa6d14e6b7c88] schedule_timeout at ffffffffb8baef88 #3 [ffffa6d14e6b7d10] wait_for_completion at ffffffffb8ba968b #4 [ffffa6d14e6b7d60] mlx5e_take_all_encap_flows at ffffffffc0f47ec4 [mlx5_core] #5 [ffffa6d14e6b7da0] mlx5e_rep_update_flows at ffffffffc0f3e734 [mlx5_core] #6 [ffffa6d14e6b7df8] mlx5e_rep_neigh_update at ffffffffc0f400bb [mlx5_core] #7 [ffffa6d14e6b7e50] process_one_work at ffffffffb80acc9c #8 [ffffa6d14e6b7ed0] worker_thread at ffffffffb80ad012 #9 [ffffa6d14e6b7f10] kthread at ffffffffb80b615d #10 [ffffa6d14e6b7f50] ret_from_fork at ffffffffb8001b2f After the first encap is attached, flow will be added to encap entry's flows list. If neigh update is running at this time, the following encaps of the flow can't hold the encap_tbl_lock and sleep. If neigh update thread is waiting for that flow's init_done, deadlock happens. Fix it by holding lock outside of the for loop. If neigh update is running, prevent encap flows from offloading. Since the lock is held outside of the for loop, concurrent creation of encap entries is not allowed. So remove unnecessary wait_for_completion call for res_ready. Fixes: 95435ad ("net/mlx5e: Only access fully initialized flows in neigh update") Signed-off-by: Chris Mi <cmi@nvidia.com> Reviewed-by: Roi Dayan <roid@nvidia.com> Reviewed-by: Vlad Buslov <vladbu@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

Currently, the per cpu upcall counters are allocated after the vport is created and inserted into the system. This could lead to the datapath accessing the counters before they are allocated resulting in a kernel Oops. Here is an example: PID: 59693 TASK: ffff0005f4f51500 CPU: 0 COMMAND: "ovs-vswitchd" #0 [ffff80000a39b5b0] __switch_to at ffffb70f0629f2f4 #1 [ffff80000a39b5d0] __schedule at ffffb70f0629f5cc #2 [ffff80000a39b650] preempt_schedule_common at ffffb70f0629fa60 #3 [ffff80000a39b670] dynamic_might_resched at ffffb70f0629fb58 #4 [ffff80000a39b680] mutex_lock_killable at ffffb70f062a1388 #5 [ffff80000a39b6a0] pcpu_alloc at ffffb70f0594460c #6 [ffff80000a39b750] __alloc_percpu_gfp at ffffb70f05944e68 #7 [ffff80000a39b760] ovs_vport_cmd_new at ffffb70ee6961b90 [openvswitch] ... PID: 58682 TASK: ffff0005b2f0bf00 CPU: 0 COMMAND: "kworker/0:3" #0 [ffff80000a5d2f40] machine_kexec at ffffb70f056a0758 #1 [ffff80000a5d2f70] __crash_kexec at ffffb70f057e2994 #2 [ffff80000a5d3100] crash_kexec at ffffb70f057e2ad8 #3 [ffff80000a5d3120] die at ffffb70f0628234c #4 [ffff80000a5d31e0] die_kernel_fault at ffffb70f062828a8 #5 [ffff80000a5d3210] __do_kernel_fault at ffffb70f056a31f4 #6 [ffff80000a5d3240] do_bad_area at ffffb70f056a32a4 #7 [ffff80000a5d3260] do_translation_fault at ffffb70f062a9710 #8 [ffff80000a5d3270] do_mem_abort at ffffb70f056a2f74 #9 [ffff80000a5d32a0] el1_abort at ffffb70f06297dac #10 [ffff80000a5d32d0] el1h_64_sync_handler at ffffb70f06299b24 #11 [ffff80000a5d3410] el1h_64_sync at ffffb70f056812dc #12 [ffff80000a5d3430] ovs_dp_upcall at ffffb70ee6963c84 [openvswitch] #13 [ffff80000a5d3470] ovs_dp_process_packet at ffffb70ee6963fdc [openvswitch] #14 [ffff80000a5d34f0] ovs_vport_receive at ffffb70ee6972c78 [openvswitch] #15 [ffff80000a5d36f0] netdev_port_receive at ffffb70ee6973948 [openvswitch] #16 [ffff80000a5d3720] netdev_frame_hook at ffffb70ee6973a28 [openvswitch] #17 [ffff80000a5d3730] __netif_receive_skb_core.constprop.0 at ffffb70f06079f90 We moved the per cpu upcall counter allocation to the existing vport alloc and free functions to solve this. Fixes: 95637d9 ("net: openvswitch: release vport resources on failure") Fixes: 1933ea3 ("net: openvswitch: Add support to count upcall packets") Signed-off-by: Eelco Chaudron <echaudro@redhat.com> Reviewed-by: Simon Horman <simon.horman@corigine.com> Acked-by: Aaron Conole <aconole@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>

The buffer is used to save register mapping in a sample. Normally perf samples don't have any register so the string should be empty. But it missed to initialize the buffer when the size is 0. And it's passed to PyUnicode_FromString() with a garbage data. So it returns NULL due to invalid input (instead of an empty unicode string object) which causes a segfault like below: Thread 2.1 "perf" received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7ffff7c83780 (LWP 193775)] 0x00007ffff6dbca2e in PyDict_SetItem () from /lib/x86_64-linux-gnu/libpython3.11.so.1.0 (gdb) bt #0 0x00007ffff6dbca2e in PyDict_SetItem () from /lib/x86_64-linux-gnu/libpython3.11.so.1.0 #1 0x00007ffff6dbf848 in PyDict_SetItemString () from /lib/x86_64-linux-gnu/libpython3.11.so.1.0 #2 0x000055555575824d in pydict_set_item_string_decref (val=0x0, key=0x5555557f96e3 "iregs", dict=0x7ffff5f7f780) at util/scripting-engines/trace-event-python.c:145 #3 set_regs_in_dict (evsel=0x555555efc370, sample=0x7fffffffb870, dict=0x7ffff5f7f780) at util/scripting-engines/trace-event-python.c:776 #4 get_perf_sample_dict (sample=sample@entry=0x7fffffffb870, evsel=evsel@entry=0x555555efc370, al=al@entry=0x7fffffffb2e0, addr_al=addr_al@entry=0x0, callchain=callchain@entry=0x7ffff63ef440) at util/scripting-engines/trace-event-python.c:923 #5 0x0000555555758ec1 in python_process_tracepoint (sample=0x7fffffffb870, evsel=0x555555efc370, al=0x7fffffffb2e0, addr_al=0x0) at util/scripting-engines/trace-event-python.c:1044 #6 0x00005555555c5db8 in process_sample_event (tool=<optimized out>, event=<optimized out>, sample=<optimized out>, evsel=0x555555efc370, machine=0x555555ef4d68) at builtin-script.c:2421 #7 0x00005555556b7793 in perf_session__deliver_event (session=0x555555ef4b60, event=0x7ffff62ff7d0, tool=0x7fffffffc150, file_offset=30672, file_path=0x555555efb8a0 "perf.data") at util/session.c:1639 #8 0x00005555556bc864 in do_flush (show_progress=true, oe=0x555555efb700) at util/ordered-events.c:245 #9 __ordered_events__flush (oe=oe@entry=0x555555efb700, how=how@entry=OE_FLUSH__FINAL, timestamp=timestamp@entry=0) at util/ordered-events.c:324 #10 0x00005555556bd06e in ordered_events__flush (oe=oe@entry=0x555555efb700, how=how@entry=OE_FLUSH__FINAL) at util/ordered-events.c:342 #11 0x00005555556b9d63 in __perf_session__process_events (session=0x555555ef4b60) at util/session.c:2465 #12 perf_session__process_events (session=0x555555ef4b60) at util/session.c:2627 #13 0x00005555555cb1d0 in __cmd_script (script=0x7fffffffc150) at builtin-script.c:2839 #14 cmd_script (argc=<optimized out>, argv=<optimized out>) at builtin-script.c:4365 #15 0x0000555555650811 in run_builtin (p=p@entry=0x555555ed8948 <commands+456>, argc=argc@entry=4, argv=argv@entry=0x7fffffffe240) at perf.c:323 #16 0x0000555555597eb3 in handle_internal_command (argv=0x7fffffffe240, argc=4) at perf.c:377 #17 run_argv (argv=<synthetic pointer>, argcp=<synthetic pointer>) at perf.c:421 #18 main (argc=4, argv=0x7fffffffe240) at perf.c:537 Fixes: 51cfe7a ("perf python: Avoid 2 leak sanitizer issues") Cc: Peter Zijlstra <peterz@infradead.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Ingo Molnar <mingo@kernel.org> Acked-by: Ian Rogers <irogers@google.com> Signed-off-by: Namhyung Kim <namhyung@kernel.org>

Alexei reported: After fast forwarding bpf-next today bpf_nf test started to fail when run twice: $ ./test_progs -t bpf_nf #17 bpf_nf:OK Summary: 1/10 PASSED, 0 SKIPPED, 0 FAILED $ ./test_progs -t bpf_nf All error logs: test_bpf_nf_ct:PASS:test_bpf_nf__open_and_load 0 nsec test_bpf_nf_ct:PASS:iptables-legacy -t raw -A PREROUTING -j CONNMARK --set-mark 42/0 0 nsec (network_helpers.c:102: errno: Address already in use) Failed to bind socket test_bpf_nf_ct:FAIL:start_server unexpected start_server: actual -1 < expected 0 #17/1 bpf_nf/xdp-ct:FAIL test_bpf_nf_ct:PASS:test_bpf_nf__open_and_load 0 nsec test_bpf_nf_ct:PASS:iptables-legacy -t raw -A PREROUTING -j CONNMARK --set-mark 42/0 0 nsec (network_helpers.c:102: errno: Address already in use) Failed to bind socket test_bpf_nf_ct:FAIL:start_server unexpected start_server: actual -1 < expected 0 #17/2 bpf_nf/tc-bpf-ct:FAIL #17 bpf_nf:FAIL Summary: 0/8 PASSED, 0 SKIPPED, 1 FAILED I was able to locally reproduce as well. Rearrange the connection teardown so that the client closes its connection first so that we don't need to linger in TCP time-wait. Fixes: e81fbd4 ("selftests/bpf: Add existing connection bpf_*_ct_lookup() test") Reported-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/CAADnVQ+0dnDq_v_vH1EfkacbfGnHANaon7zsw10pMb-D9FS0Pw@mail.gmail.com Link: https://lore.kernel.org/bpf/20230626131942.5100-1-daniel@iogearbox.net

Petr Machata says: ==================== mlxsw: Permit enslavement to netdevices with uppers The mlxsw driver currently makes the assumption that the user applies configuration in a bottom-up manner. Thus netdevices need to be added to the bridge before IP addresses are configured on that bridge or SVI added on top of it. Enslaving a netdevice to another netdevice that already has uppers is in fact forbidden by mlxsw for this reason. Despite this safety, it is rather easy to get into situations where the offloaded configuration is just plain wrong. As an example, take a front panel port, configure an IP address: it gets a RIF. Now enslave the port to the bridge, and the RIF is gone. Remove the port from the bridge again, but the RIF never comes back. There is a number of similar situations, where changing the configuration there and back utterly breaks the offload. Similarly, detaching a front panel port from a configured topology means unoffloading of this whole topology -- VLAN uppers, next hops, etc. Attaching the port back is then not permitted at all. If it were, it would not result in a working configuration, because much of mlxsw is written to react to changes in immediate configuration. There is nothing that would go visit netdevices in the attached-to topology and offload existing routes and VLAN memberships, for example. In this patchset, introduce a number of replays to be invoked so that this sort of post-hoc offload is supported. Then remove the vetoes that disallowed enslavement of front panel ports to other netdevices with uppers. The patchset progresses as follows: - In patch #1, fix an issue in the bridge driver. To my knowledge, the issue could not have resulted in a buggy behavior previously, and thus is packaged with this patchset instead of being sent separately to net. - In patch #2, add a new helper to the switchdev code. - In patch #3, drop mlxsw selftests that will not be relevant after this patchset anymore. - Patches #4, #5, #6, #7 and #8 prepare the codebase for smoother introduction of the rest of the code. - Patches #9, #10, #11, #12, #13 and #14 replay various aspects of upper configuration when a front panel port is introduced into a topology. Individual patches take care of bridge and LAG RIF memberships, switchdev replay, nexthop and neighbors replay, and MACVLAN offload. - Patches #15 and #16 introduce RIFs for newly-relevant netdevices when a front panel port is enslaved (in which case all uppers are newly relevant), or, respectively, deslaved (in which case the newly-relevant netdevice is the one being deslaved). - Up until this point, the introduced scaffolding was not really used, because mlxsw still forbids enslavement of mlxsw netdevices to uppers with uppers. In patch #17, this condition is finally relaxed. A sizable selftest suite is available to test all this new code. That will be sent in a separate patchset. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>

The following processes run into a deadlock. CPU 41 was waiting for CPU 29 to handle a CSD request while holding spinlock "crashdump_lock", but CPU 29 was hung by that spinlock with IRQs disabled. PID: 17360 TASK: ffff95c1090c5c40 CPU: 41 COMMAND: "mrdiagd" !# 0 [ffffb80edbf37b58] __read_once_size at ffffffff9b871a40 include/linux/compiler.h:185:0 !# 1 [ffffb80edbf37b58] atomic_read at ffffffff9b871a40 arch/x86/include/asm/atomic.h:27:0 !# 2 [ffffb80edbf37b58] dump_stack at ffffffff9b871a40 lib/dump_stack.c:54:0 # 3 [ffffb80edbf37b78] csd_lock_wait_toolong at ffffffff9b131ad5 kernel/smp.c:364:0 # 4 [ffffb80edbf37b78] __csd_lock_wait at ffffffff9b131ad5 kernel/smp.c:384:0 # 5 [ffffb80edbf37bf8] csd_lock_wait at ffffffff9b13267a kernel/smp.c:394:0 # 6 [ffffb80edbf37bf8] smp_call_function_many at ffffffff9b13267a kernel/smp.c:843:0 # 7 [ffffb80edbf37c50] smp_call_function at ffffffff9b13279d kernel/smp.c:867:0 # 8 [ffffb80edbf37c50] on_each_cpu at ffffffff9b13279d kernel/smp.c:976:0 # 9 [ffffb80edbf37c78] flush_tlb_kernel_range at ffffffff9b085c4b arch/x86/mm/tlb.c:742:0 #10 [ffffb80edbf37cb8] __purge_vmap_area_lazy at ffffffff9b23a1e0 mm/vmalloc.c:701:0 #11 [ffffb80edbf37ce0] try_purge_vmap_area_lazy at ffffffff9b23a2cc mm/vmalloc.c:722:0 #12 [ffffb80edbf37ce0] free_vmap_area_noflush at ffffffff9b23a2cc mm/vmalloc.c:754:0 #13 [ffffb80edbf37cf8] free_unmap_vmap_area at ffffffff9b23bb3b mm/vmalloc.c:764:0 #14 [ffffb80edbf37cf8] remove_vm_area at ffffffff9b23bb3b mm/vmalloc.c:1509:0 #15 [ffffb80edbf37d18] __vunmap at ffffffff9b23bb8a mm/vmalloc.c:1537:0 #16 [ffffb80edbf37d40] vfree at ffffffff9b23bc85 mm/vmalloc.c:1612:0 #17 [ffffb80edbf37d58] megasas_free_host_crash_buffer [megaraid_sas] at ffffffffc020b7f2 drivers/scsi/megaraid/megaraid_sas_fusion.c:3932:0 #18 [ffffb80edbf37d80] fw_crash_state_store [megaraid_sas] at ffffffffc01f804d drivers/scsi/megaraid/megaraid_sas_base.c:3291:0 #19 [ffffb80edbf37dc0] dev_attr_store at ffffffff9b56dd7b drivers/base/core.c:758:0 #20 [ffffb80edbf37dd0] sysfs_kf_write at ffffffff9b326acf fs/sysfs/file.c:144:0 #21 [ffffb80edbf37de0] kernfs_fop_write at ffffffff9b325fd4 fs/kernfs/file.c:316:0 #22 [ffffb80edbf37e20] __vfs_write at ffffffff9b29418a fs/read_write.c:480:0 #23 [ffffb80edbf37ea8] vfs_write at ffffffff9b294462 fs/read_write.c:544:0 #24 [ffffb80edbf37ee8] SYSC_write at ffffffff9b2946ec fs/read_write.c:590:0 #25 [ffffb80edbf37ee8] SyS_write at ffffffff9b2946ec fs/read_write.c:582:0 #26 [ffffb80edbf37f30] do_syscall_64 at ffffffff9b003ca9 arch/x86/entry/common.c:298:0 #27 [ffffb80edbf37f58] entry_SYSCALL_64 at ffffffff9ba001b1 arch/x86/entry/entry_64.S:238:0 PID: 17355 TASK: ffff95c1090c3d80 CPU: 29 COMMAND: "mrdiagd" !# 0 [ffffb80f2d3c7d30] __read_once_size at ffffffff9b0f2ab0 include/linux/compiler.h:185:0 !# 1 [ffffb80f2d3c7d30] native_queued_spin_lock_slowpath at ffffffff9b0f2ab0 kernel/locking/qspinlock.c:368:0 # 2 [ffffb80f2d3c7d58] pv_queued_spin_lock_slowpath at ffffffff9b0f244b arch/x86/include/asm/paravirt.h:674:0 # 3 [ffffb80f2d3c7d58] queued_spin_lock_slowpath at ffffffff9b0f244b arch/x86/include/asm/qspinlock.h:53:0 # 4 [ffffb80f2d3c7d68] queued_spin_lock at ffffffff9b8961a6 include/asm-generic/qspinlock.h:90:0 # 5 [ffffb80f2d3c7d68] do_raw_spin_lock_flags at ffffffff9b8961a6 include/linux/spinlock.h:173:0 # 6 [ffffb80f2d3c7d68] __raw_spin_lock_irqsave at ffffffff9b8961a6 include/linux/spinlock_api_smp.h:122:0 # 7 [ffffb80f2d3c7d68] _raw_spin_lock_irqsave at ffffffff9b8961a6 kernel/locking/spinlock.c:160:0 # 8 [ffffb80f2d3c7d88] fw_crash_buffer_store [megaraid_sas] at ffffffffc01f8129 drivers/scsi/megaraid/megaraid_sas_base.c:3205:0 # 9 [ffffb80f2d3c7dc0] dev_attr_store at ffffffff9b56dd7b drivers/base/core.c:758:0 #10 [ffffb80f2d3c7dd0] sysfs_kf_write at ffffffff9b326acf fs/sysfs/file.c:144:0 #11 [ffffb80f2d3c7de0] kernfs_fop_write at ffffffff9b325fd4 fs/kernfs/file.c:316:0 #12 [ffffb80f2d3c7e20] __vfs_write at ffffffff9b29418a fs/read_write.c:480:0 #13 [ffffb80f2d3c7ea8] vfs_write at ffffffff9b294462 fs/read_write.c:544:0 #14 [ffffb80f2d3c7ee8] SYSC_write at ffffffff9b2946ec fs/read_write.c:590:0 #15 [ffffb80f2d3c7ee8] SyS_write at ffffffff9b2946ec fs/read_write.c:582:0 #16 [ffffb80f2d3c7f30] do_syscall_64 at ffffffff9b003ca9 arch/x86/entry/common.c:298:0 #17 [ffffb80f2d3c7f58] entry_SYSCALL_64 at ffffffff9ba001b1 arch/x86/entry/entry_64.S:238:0 The lock is used to synchronize different sysfs operations, it doesn't protect any resource that will be touched by an interrupt. Consequently it's not required to disable IRQs. Replace the spinlock with a mutex to fix the deadlock. Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com> Link: https://lore.kernel.org/r/20230828221018.19471-1-junxiao.bi@oracle.com Reviewed-by: Mike Christie <michael.christie@oracle.com> Cc: stable@vger.kernel.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

Fix an error detected by memory sanitizer: ``` ==4033==WARNING: MemorySanitizer: use-of-uninitialized-value #0 0x55fb0fbedfc7 in read_alias_info tools/perf/util/pmu.c:457:6 #1 0x55fb0fbea339 in check_info_data tools/perf/util/pmu.c:1434:2 #2 0x55fb0fbea339 in perf_pmu__check_alias tools/perf/util/pmu.c:1504:9 #3 0x55fb0fbdca85 in parse_events_add_pmu tools/perf/util/parse-events.c:1429:32 #4 0x55fb0f965230 in parse_events_parse tools/perf/util/parse-events.y:299:6 #5 0x55fb0fbdf6b2 in parse_events__scanner tools/perf/util/parse-events.c:1822:8 #6 0x55fb0fbdf8c1 in __parse_events tools/perf/util/parse-events.c:2094:8 #7 0x55fb0fa8ffa9 in parse_events tools/perf/util/parse-events.h:41:9 #8 0x55fb0fa8ffa9 in test_event tools/perf/tests/parse-events.c:2393:8 #9 0x55fb0fa8f458 in test__pmu_events tools/perf/tests/parse-events.c:2551:15 #10 0x55fb0fa6d93f in run_test tools/perf/tests/builtin-test.c:242:9 #11 0x55fb0fa6d93f in test_and_print tools/perf/tests/builtin-test.c:271:8 #12 0x55fb0fa6d082 in __cmd_test tools/perf/tests/builtin-test.c:442:5 #13 0x55fb0fa6d082 in cmd_test tools/perf/tests/builtin-test.c:564:9 #14 0x55fb0f942720 in run_builtin tools/perf/perf.c:322:11 #15 0x55fb0f942486 in handle_internal_command tools/perf/perf.c:375:8 #16 0x55fb0f941dab in run_argv tools/perf/perf.c:419:2 #17 0x55fb0f941dab in main tools/perf/perf.c:535:3 ``` Fixes: 7b723db ("perf pmu: Be lazy about loading event info files from sysfs") Signed-off-by: Ian Rogers <irogers@google.com> Cc: James Clark <james.clark@arm.com> Cc: Kan Liang <kan.liang@linux.intel.com> Link: https://lore.kernel.org/r/20230914022425.1489035-1-irogers@google.com Signed-off-by: Namhyung Kim <namhyung@kernel.org>

The following call trace shows a deadlock issue due to recursive locking of mutex "device_mutex". First lock acquire is in target_for_each_device() and second in target_free_device(). PID: 148266 TASK: ffff8be21ffb5d00 CPU: 10 COMMAND: "iscsi_ttx" #0 [ffffa2bfc9ec3b18] __schedule at ffffffffa8060e7f #1 [ffffa2bfc9ec3ba0] schedule at ffffffffa8061224 #2 [ffffa2bfc9ec3bb8] schedule_preempt_disabled at ffffffffa80615ee #3 [ffffa2bfc9ec3bc8] __mutex_lock at ffffffffa8062fd7 #4 [ffffa2bfc9ec3c40] __mutex_lock_slowpath at ffffffffa80631d3 #5 [ffffa2bfc9ec3c50] mutex_lock at ffffffffa806320c #6 [ffffa2bfc9ec3c68] target_free_device at ffffffffc0935998 [target_core_mod] #7 [ffffa2bfc9ec3c90] target_core_dev_release at ffffffffc092f975 [target_core_mod] #8 [ffffa2bfc9ec3ca0] config_item_put at ffffffffa79d250f #9 [ffffa2bfc9ec3cd0] config_item_put at ffffffffa79d2583 #10 [ffffa2bfc9ec3ce0] target_devices_idr_iter at ffffffffc0933f3a [target_core_mod] #11 [ffffa2bfc9ec3d00] idr_for_each at ffffffffa803f6fc #12 [ffffa2bfc9ec3d60] target_for_each_device at ffffffffc0935670 [target_core_mod] #13 [ffffa2bfc9ec3d98] transport_deregister_session at ffffffffc0946408 [target_core_mod] #14 [ffffa2bfc9ec3dc8] iscsit_close_session at ffffffffc09a44a6 [iscsi_target_mod] #15 [ffffa2bfc9ec3df0] iscsit_close_connection at ffffffffc09a4a88 [iscsi_target_mod] #16 [ffffa2bfc9ec3df8] finish_task_switch at ffffffffa76e5d07 #17 [ffffa2bfc9ec3e78] iscsit_take_action_for_connection_exit at ffffffffc0991c23 [iscsi_target_mod] #18 [ffffa2bfc9ec3ea0] iscsi_target_tx_thread at ffffffffc09a403b [iscsi_target_mod] #19 [ffffa2bfc9ec3f08] kthread at ffffffffa76d8080 #20 [ffffa2bfc9ec3f50] ret_from_fork at ffffffffa8200364 Fixes: 36d4cb4 ("scsi: target: Avoid that EXTENDED COPY commands trigger lock inversion") Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com> Link: https://lore.kernel.org/r/20230918225848.66463-1-junxiao.bi@oracle.com Reviewed-by: Mike Christie <michael.christie@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

@0

Add test to validate BPF verifier's register range bounds tracking logic. The main bulk is a lot of auto-generated tests based on a small set of seed values for lower and upper 32 bits of full 64-bit values. Currently we validate only range vs const comparisons, but the idea is to start validating range over range comparisons in subsequent patch set. When setting up initial register ranges we treat registers as one of u64/s64/u32/s32 numeric types, and then independently perform conditional comparisons based on a potentially different u64/s64/u32/s32 types. This tests lots of tricky cases of deriving bounds information across different numeric domains. Given there are lots of auto-generated cases, we guard them behind SLOW_TESTS=1 envvar requirement, and skip them altogether otherwise. With current full set of upper/lower seed value, all supported comparison operators and all the combinations of u64/s64/u32/s32 number domains, we get about 7.7 million tests, which run in about 35 minutes on my local qemu instance without parallelization. But we also split those tests by init/cond numeric types, which allows to rely on test_progs's parallelization of tests with `-j` option, getting run time down to about 5 minutes on 8 cores. It's still something that shouldn't be run during normal test_progs run. But we can run it a reasonable time, and so perhaps a nightly CI test run (once we have it) would be a good option for this. We also add a small set of tricky conditions that came up during development and triggered various bugs or corner cases in either selftest's reimplementation of range bounds logic or in verifier's logic itself. These are fast enough to be run as part of normal test_progs test run and are great for a quick sanity checking. Let's take a look at test output to understand what's going on: $ sudo ./test_progs -t reg_bounds_crafted torvalds#191/1 reg_bounds_crafted/(u64)[0; 0xffffffff] (u64)< 0:OK ... torvalds#191/115 reg_bounds_crafted/(u64)[0; 0x17fffffff] (s32)< 0:OK ... torvalds#191/137 reg_bounds_crafted/(u64)[0xffffffff; 0x100000000] (u64)== 0:OK Each test case is uniquely and fully described by this generated string. E.g.: "(u64)[0; 0x17fffffff] (s32)< 0". This means that we initialize a register (R6) in such a way that verifier knows that it can have a value in [(u64)0; (u64)0x17fffffff] range. Another register (R7) is also set up as u64, but this time a constant (zero in this case). They then are compared using 32-bit signed < operation. Resulting TRUE/FALSE branches are evaluated (including cases where it's known that one of the branches will never be taken, in which case we validate that verifier also determines this as a dead code). Test validates that verifier's final register state matches expected state based on selftest's own reg_state logic, implemented from scratch for cross-checking purposes. These test names can be conveniently used for further debugging, and if -vv verboseness is requested we can get a corresponding verifier log (with mark_precise logs filtered out as irrelevant and distracting). Example below is slightly redacted for brevity, omitting irrelevant register output in some places, marked with [...]. $ sudo ./test_progs -a 'reg_bounds_crafted/(u32)[0; U32_MAX] (s32)< -1' -vv ... VERIFIER LOG: ======================== func#0 @0 0: R1=ctx(off=0,imm=0) R10=fp0 0: (05) goto pc+2 3: (85) call bpf_get_current_pid_tgid#14 ; R0_w=scalar() 4: (bc) w6 = w0 ; R0_w=scalar() R6_w=scalar(smin=0,smax=umax=4294967295,var_off=(0x0; 0xffffffff)) 5: (85) call bpf_get_current_pid_tgid#14 ; R0_w=scalar() 6: (bc) w7 = w0 ; R0_w=scalar() R7_w=scalar(smin=0,smax=umax=4294967295,var_off=(0x0; 0xffffffff)) 7: (b4) w1 = 0 ; R1_w=0 8: (b4) w2 = -1 ; R2=4294967295 9: (ae) if w6 < w1 goto pc-9 9: R1=0 R6=scalar(smin=0,smax=umax=4294967295,var_off=(0x0; 0xffffffff)) 10: (2e) if w6 > w2 goto pc-10 10: R2=4294967295 R6=scalar(smin=0,smax=umax=4294967295,var_off=(0x0; 0xffffffff)) 11: (b4) w1 = -1 ; R1_w=4294967295 12: (b4) w2 = -1 ; R2_w=4294967295 13: (ae) if w7 < w1 goto pc-13 ; R1_w=4294967295 R7=4294967295 14: (2e) if w7 > w2 goto pc-14 14: R2_w=4294967295 R7=4294967295 15: (bc) w0 = w6 ; [...] R6=scalar(id=1,smin=0,smax=umax=4294967295,var_off=(0x0; 0xffffffff)) 16: (bc) w0 = w7 ; [...] R7=4294967295 17: (ce) if w6 s< w7 goto pc+3 ; R6=scalar(id=1,smin=0,smax=umax=4294967295,smin32=-1,var_off=(0x0; 0xffffffff)) R7=4294967295 18: (bc) w0 = w6 ; [...] R6=scalar(id=1,smin=0,smax=umax=4294967295,smin32=-1,var_off=(0x0; 0xffffffff)) 19: (bc) w0 = w7 ; [...] R7=4294967295 20: (95) exit from 17 to 21: [...] 21: (bc) w0 = w6 ; [...] R6=scalar(id=1,smin=umin=umin32=2147483648,smax=umax=umax32=4294967294,smax32=-2,var_off=(0x80000000; 0x7fffffff)) 22: (bc) w0 = w7 ; [...] R7=4294967295 23: (95) exit from 13 to 1: [...] 1: [...] 1: (b7) r0 = 0 ; R0_w=0 2: (95) exit processed 24 insns (limit 1000000) max_states_per_insn 0 total_states 2 peak_states 2 mark_read 1 ===================== Verifier log above is for `(u32)[0; U32_MAX] (s32)< -1` use cases, where u32 range is used for initialization, followed by signed < operator. Note how we use w6/w7 in this case for register initialization (it would be R6/R7 for 64-bit types) and then `if w6 s< w7` for comparison at instruction #17. It will be `if R6 < R7` for 64-bit unsigned comparison. Above example gives a good impression of the overall structure of a BPF programs generated for reg_bounds tests. In the future, this "framework" can be extended to test not just conditional jumps, but also arithmetic operations. Adding randomized testing is another possibility. Some implementation notes. We basically have our own generics-like operations on numbers, where all the numbers are stored in u64, but how they are interpreted is passed as runtime argument enum num_t. Further, `struct range` represents a bounds range, and those are collected together into a minimal `struct reg_state`, which collects range bounds across all four numberical domains: u64, s64, u32, s64. Based on these primitives and `enum op` representing possible conditional operation (<, <=, >, >=, ==, !=), there is a set of generic helpers to perform "range arithmetics", which is used to maintain struct reg_state. We simulate what verifier will do for reg bounds of R6 and R7 registers using these range and reg_state primitives. Simulated information is used to determine branch taken conclusion and expected exact register state across all four number domains. Implementation of "range arithmetics" is more generic than what verifier is currently performing: it allows range over range comparisons and adjustments. This is the intended end goal of this patch set overall and verifier logic is enhanced in subsequent patches in this series to handle range vs range operations, at which point selftests are extended to validate these conditions as well. For now it's range vs const cases only. Note that tests are split into multiple groups by their numeric types for initialization of ranges and for comparison operation. This allows to use test_progs's -j parallelization to speed up tests, as we now have 16 groups of parallel running tests. Overall reduction of running time that allows is pretty good, we go down from more than 30 minutes to slightly less than 5 minutes running time. Acked-by: Eduard Zingerman <eddyz87@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com> Link: https://lore.kernel.org/r/20231112010609.848406-8-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org>

Petr Machata says: ==================== mlxsw: Support CFF flood mode The registers to configure to initialize a flood table differ between the controlled and CFF flood modes. In therefore needs to be an op. Add it, hook up the current init to the existing families, and invoke the op. PGT is an in-HW table that maps addresses to sets of ports. Then when some HW process needs a set of ports as an argument, instead of embedding the actual set in the dynamic configuration, what gets configured is the address referencing the set. The HW then works with the appropriate PGT entry. Among other allocations, the PGT currently contains two large blocks for bridge flooding: one for 802.1q and one for 802.1d. Within each of these blocks are three tables, for unknown-unicast, multicast and broadcast flooding: . . . | 802.1q | 802.1d | . . . | UC | MC | BC | UC | MC | BC | \______ _____/ \_____ ______/ v v FID flood vectors Thus each FID (which corresponds to an 802.1d bridge or one VLAN in an 802.1q bridge) uses three flood vectors spread across a fairly large region of PGT. This way of organizing the flood table (called "controlled") is not very flexible. E.g. to decrease a bridge scale and store more IP MC vectors, one would need to completely rewrite the bridge PGT blocks, or resort to hacks such as storing individual MC flood vectors into unused part of the bridge table. In order to address these shortcomings, Spectrum-2 and above support what is called CFF flood mode, for Compressed FID Flooding. In CFF flood mode, each FID has a little table of its own, with three entries adjacent to each other, one for unknown-UC, one for MC, one for BC. This allows for a much more fine-grained approach to PGT management, where bits of it are allocated on demand. . . . | FID | FID | FID | FID | FID | . . . |U|M|B|U|M|B|U|M|B|U|M|B|U|M|B| \_____________ _____________/ v FID flood vectors Besides the FID table organization, the CFF flood mode also impacts Router Subport (RSP) table. This table contains flood vectors for rFIDs, which are FIDs that reference front panel ports or LAGs. The RSP table contains two entries per front panel port and LAG, one for unknown-UC traffic, and one for everything else. Currently, the FW allocates and manages the table in its own part of PGT. rFIDs are marked with flood_rsp bit and managed specially. In CFF mode, rFIDs are managed as all other FIDs. The driver therefore has to allocate and maintain the flood vectors. Like with bridge FIDs, this is more work, but increases flexibility of the system. The FW currently supports both the controlled and CFF flood modes. To shed complexity, in the future it should only support CFF flood mode. Hence this patchset, which adds CFF flood mode support to mlxsw. Since mlxsw needs to maintain both the controlled mode as well as CFF mode support, we will keep the layout as compatible as possible. The bridge tables will stay in the same overall shape, just their inner organization will change from flood mode -> FID to FID -> flood mode. Likewise will RSP be kept as a contiguous block of PGT memory, as was the case when the FW maintained it. - The way FIDs get configured under the CFF flood mode differs from the currently used controlled mode. The simple approach of having several globally visible arrays for spectrum.c to statically choose from no longer works. Patch #1 thus privatizes all FID initialization and finalization logic, and exposes it as ops instead. - Patch #2 renames the ops that are specific to the controlled mode, to make room in the namespace for the CFF variants. Patch #3 extracts a helper to compute flood table base out of mlxsw_sp_fid_flood_table_mid(). - The op fid_setup configured fid_offset, i.e. the number of this FID within its family. For rFIDs in CFF mode, to determine this number, the driver will need to do fallible queries. Thus in patch #4, make the FID setup operation fallible as well. - Flood mode initialization routine differs between the controlled and CFF flood modes. The controlled mode needs to configure flood table layout, which the CFF mode does not need to do. In patch #5, move mlxsw_sp_fid_flood_table_init() up so that the following patch can make use of it. In patch #6, add an op to be invoked per table (if defined). - The current way of determining PGT allocation size depends on the number of FIDs and number of flood tables. RFIDs however have PGT footprint depending not on number of FIDs, but on number of ports and LAGs, because which ports an rFID should flood to does not depend on the FID itself, but on the port or LAG that it references. Therefore in patch #7, add FID family ops for determining PGT allocation size. - As elaborated above, layout of PGT will differ between controlled and CFF flood modes. In CFF mode, it will further differ between rFIDs and other FIDs (as described at previous patch). The way to pack the SFMR register to configure a FID will likewise differ from controlled to CFF. Thus in patches #8 and #9 add FID family ops to determine PGT base address for a FID and to pack SFMR. - Patches #10 and #11 add more bits for RSP support. In patch #10, add a new traffic type enumerator, for non-UC traffic. This is a combination of BC and MC traffic, but the way that mlxsw maps these mnemonic names to actual traffic type configurations requires that we have a new name to describe this class of traffic. Patch #11 then adds hooks necessary for RSP table maintenance. As ports come and go, and join and leave LAGs, it is necessary to update flood vectors that the rFIDs use. These new hooks will make that possible. - Patches #12, #13 and #14 introduce flood profiles. These have been implicit so far, but the way that CFF flood mode works with profile IDs requires that we make them explicit. Thus in patch #12, introduce flood profile objects as a set of flood tables that FID families then refer to. The FID code currently only uses a single flood profile. In patch #13, add a flood profile ID to flood profile objects. In patch #14, when in CFF mode, configure SFFP according to the existing flood profiles (or the one that exists as of that point). - Patches #15 and #16 add code to implement, respectively, bridge FIDs and RSP FIDs in CFF mode. - In patch #17, toggle flood_mode_prefer_cff on Spectrum-2 and above, which makes the newly-added code live. ==================== Link: https://lore.kernel.org/r/cover.1701183891.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>

When creating ceq_0 during probing irdma, cqp.sc_cqp will be sent as a cqp_request to cqp->sc_cqp.sq_ring. If the request is pending when removing the irdma driver or unplugging its aux device, cqp.sc_cqp will be dereferenced as wrong struct in irdma_free_pending_cqp_request(). PID: 3669 TASK: ffff88aef892c000 CPU: 28 COMMAND: "kworker/28:0" #0 [fffffe0000549e38] crash_nmi_callback at ffffffff810e3a34 #1 [fffffe0000549e40] nmi_handle at ffffffff810788b2 #2 [fffffe0000549ea0] default_do_nmi at ffffffff8107938f #3 [fffffe0000549eb8] do_nmi at ffffffff81079582 #4 [fffffe0000549ef0] end_repeat_nmi at ffffffff82e016b4 [exception RIP: native_queued_spin_lock_slowpath+1291] RIP: ffffffff8127e72b RSP: ffff88aa841ef778 RFLAGS: 00000046 RAX: 0000000000000000 RBX: ffff88b01f849700 RCX: ffffffff8127e47e RDX: 0000000000000000 RSI: 0000000000000004 RDI: ffffffff83857ec0 RBP: ffff88afe3e4efc8 R8: ffffed15fc7c9dfa R9: ffffed15fc7c9dfa R10: 0000000000000001 R11: ffffed15fc7c9df9 R12: 0000000000740000 R13: ffff88b01f849708 R14: 0000000000000003 R15: ffffed1603f092e1 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0000 -- <NMI exception stack> -- #5 [ffff88aa841ef778] native_queued_spin_lock_slowpath at ffffffff8127e72b #6 [ffff88aa841ef7b0] _raw_spin_lock_irqsave at ffffffff82c22aa4 #7 [ffff88aa841ef7c8] __wake_up_common_lock at ffffffff81257363 #8 [ffff88aa841ef888] irdma_free_pending_cqp_request at ffffffffa0ba12cc [irdma] #9 [ffff88aa841ef958] irdma_cleanup_pending_cqp_op at ffffffffa0ba1469 [irdma] #10 [ffff88aa841ef9c0] irdma_ctrl_deinit_hw at ffffffffa0b2989f [irdma] #11 [ffff88aa841efa28] irdma_remove at ffffffffa0b252df [irdma] #12 [ffff88aa841efae8] auxiliary_bus_remove at ffffffff8219afdb #13 [ffff88aa841efb00] device_release_driver_internal at ffffffff821882e6 #14 [ffff88aa841efb38] bus_remove_device at ffffffff82184278 #15 [ffff88aa841efb88] device_del at ffffffff82179d23 #16 [ffff88aa841efc48] ice_unplug_aux_dev at ffffffffa0eb1c14 [ice] #17 [ffff88aa841efc68] ice_service_task at ffffffffa0d88201 [ice] #18 [ffff88aa841efde8] process_one_work at ffffffff811c589a #19 [ffff88aa841efe60] worker_thread at ffffffff811c71ff #20 [ffff88aa841eff10] kthread at ffffffff811d87a0 #21 [ffff88aa841eff50] ret_from_fork at ffffffff82e0022f Fixes: 44d9e52 ("RDMA/irdma: Implement device initialization definitions") Link: https://lore.kernel.org/r/20231130081415.891006-1-lishifeng@sangfor.com.cn Suggested-by: "Ismail, Mustafa" <mustafa.ismail@intel.com> Signed-off-by: Shifeng Li <lishifeng@sangfor.com.cn> Reviewed-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

Test runners on debug kernels occasionally fail with: # # RUN tls_err.13_aes_gcm.poll_partial_rec_async ... # # tls.c:1883:poll_partial_rec_async:Expected poll(&pfd, 1, 5) (0) == 1 (1) # # tls.c:1870:poll_partial_rec_async:Expected status (256) == 0 (0) # # poll_partial_rec_async: Test failed at step #17 # # FAIL tls_err.13_aes_gcm.poll_partial_rec_async # not ok 699 tls_err.13_aes_gcm.poll_partial_rec_async # # FAILED: 698 / 699 tests passed. This points to the second poll() in the test which is expected to wait for the sender to send the rest of the data. Apparently under some conditions that doesn't happen within 5ms, bump the timeout to 20ms. Fixes: 23fcb62 ("selftests: tls: add tests for poll behavior") Link: https://lore.kernel.org/r/20240213142055.395564-1-kuba@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>

An errant disk backup on my desktop got into debugfs and triggered the following deadlock scenario in the amdgpu debugfs files. The machine also hard-resets immediately after those lines are printed (although I wasn't able to reproduce that part when reading by hand): [ 1318.016074][ T1082] ====================================================== [ 1318.016607][ T1082] WARNING: possible circular locking dependency detected [ 1318.017107][ T1082] 6.8.0-rc7-00015-ge0c8221b72c0 #17 Not tainted [ 1318.017598][ T1082] ------------------------------------------------------ [ 1318.018096][ T1082] tar/1082 is trying to acquire lock: [ 1318.018585][ T1082] ffff98c44175d6a0 (&mm->mmap_lock){++++}-{3:3}, at: __might_fault+0x40/0x80 [ 1318.019084][ T1082] [ 1318.019084][ T1082] but task is already holding lock: [ 1318.020052][ T1082] ffff98c4c13f55f8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: amdgpu_debugfs_mqd_read+0x6a/0x250 [amdgpu] [ 1318.020607][ T1082] [ 1318.020607][ T1082] which lock already depends on the new lock. [ 1318.020607][ T1082] [ 1318.022081][ T1082] [ 1318.022081][ T1082] the existing dependency chain (in reverse order) is: [ 1318.023083][ T1082] [ 1318.023083][ T1082] -> #2 (reservation_ww_class_mutex){+.+.}-{3:3}: [ 1318.024114][ T1082] __ww_mutex_lock.constprop.0+0xe0/0x12f0 [ 1318.024639][ T1082] ww_mutex_lock+0x32/0x90 [ 1318.025161][ T1082] dma_resv_lockdep+0x18a/0x330 [ 1318.025683][ T1082] do_one_initcall+0x6a/0x350 [ 1318.026210][ T1082] kernel_init_freeable+0x1a3/0x310 [ 1318.026728][ T1082] kernel_init+0x15/0x1a0 [ 1318.027242][ T1082] ret_from_fork+0x2c/0x40 [ 1318.027759][ T1082] ret_from_fork_asm+0x11/0x20 [ 1318.028281][ T1082] [ 1318.028281][ T1082] -> #1 (reservation_ww_class_acquire){+.+.}-{0:0}: [ 1318.029297][ T1082] dma_resv_lockdep+0x16c/0x330 [ 1318.029790][ T1082] do_one_initcall+0x6a/0x350 [ 1318.030263][ T1082] kernel_init_freeable+0x1a3/0x310 [ 1318.030722][ T1082] kernel_init+0x15/0x1a0 [ 1318.031168][ T1082] ret_from_fork+0x2c/0x40 [ 1318.031598][ T1082] ret_from_fork_asm+0x11/0x20 [ 1318.032011][ T1082] [ 1318.032011][ T1082] -> #0 (&mm->mmap_lock){++++}-{3:3}: [ 1318.032778][ T1082] __lock_acquire+0x14bf/0x2680 [ 1318.033141][ T1082] lock_acquire+0xcd/0x2c0 [ 1318.033487][ T1082] __might_fault+0x58/0x80 [ 1318.033814][ T1082] amdgpu_debugfs_mqd_read+0x103/0x250 [amdgpu] [ 1318.034181][ T1082] full_proxy_read+0x55/0x80 [ 1318.034487][ T1082] vfs_read+0xa7/0x360 [ 1318.034788][ T1082] ksys_read+0x70/0xf0 [ 1318.035085][ T1082] do_syscall_64+0x94/0x180 [ 1318.035375][ T1082] entry_SYSCALL_64_after_hwframe+0x46/0x4e [ 1318.035664][ T1082] [ 1318.035664][ T1082] other info that might help us debug this: [ 1318.035664][ T1082] [ 1318.036487][ T1082] Chain exists of: [ 1318.036487][ T1082] &mm->mmap_lock --> reservation_ww_class_acquire --> reservation_ww_class_mutex [ 1318.036487][ T1082] [ 1318.037310][ T1082] Possible unsafe locking scenario: [ 1318.037310][ T1082] [ 1318.037838][ T1082] CPU0 CPU1 [ 1318.038101][ T1082] ---- ---- [ 1318.038350][ T1082] lock(reservation_ww_class_mutex); [ 1318.038590][ T1082] lock(reservation_ww_class_acquire); [ 1318.038839][ T1082] lock(reservation_ww_class_mutex); [ 1318.039083][ T1082] rlock(&mm->mmap_lock); [ 1318.039328][ T1082] [ 1318.039328][ T1082] *** DEADLOCK *** [ 1318.039328][ T1082] [ 1318.040029][ T1082] 1 lock held by tar/1082: [ 1318.040259][ T1082] #0: ffff98c4c13f55f8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: amdgpu_debugfs_mqd_read+0x6a/0x250 [amdgpu] [ 1318.040560][ T1082] [ 1318.040560][ T1082] stack backtrace: [ 1318.041053][ T1082] CPU: 22 PID: 1082 Comm: tar Not tainted 6.8.0-rc7-00015-ge0c8221b72c0 #17 3316c85d50e282c5643b075d1f01a4f6365e39c2 [ 1318.041329][ T1082] Hardware name: Gigabyte Technology Co., Ltd. B650 AORUS PRO AX/B650 AORUS PRO AX, BIOS F20 12/14/2023 [ 1318.041614][ T1082] Call Trace: [ 1318.041895][ T1082] <TASK> [ 1318.042175][ T1082] dump_stack_lvl+0x4a/0x80 [ 1318.042460][ T1082] check_noncircular+0x145/0x160 [ 1318.042743][ T1082] __lock_acquire+0x14bf/0x2680 [ 1318.043022][ T1082] lock_acquire+0xcd/0x2c0 [ 1318.043301][ T1082] ? __might_fault+0x40/0x80 [ 1318.043580][ T1082] ? __might_fault+0x40/0x80 [ 1318.043856][ T1082] __might_fault+0x58/0x80 [ 1318.044131][ T1082] ? __might_fault+0x40/0x80 [ 1318.044408][ T1082] amdgpu_debugfs_mqd_read+0x103/0x250 [amdgpu 8fe2afaa910cbd7654c8cab23563a94d6caebaab] [ 1318.044749][ T1082] full_proxy_read+0x55/0x80 [ 1318.045042][ T1082] vfs_read+0xa7/0x360 [ 1318.045333][ T1082] ksys_read+0x70/0xf0 [ 1318.045623][ T1082] do_syscall_64+0x94/0x180 [ 1318.045913][ T1082] ? do_syscall_64+0xa0/0x180 [ 1318.046201][ T1082] ? lockdep_hardirqs_on+0x7d/0x100 [ 1318.046487][ T1082] ? do_syscall_64+0xa0/0x180 [ 1318.046773][ T1082] ? do_syscall_64+0xa0/0x180 [ 1318.047057][ T1082] ? do_syscall_64+0xa0/0x180 [ 1318.047337][ T1082] ? do_syscall_64+0xa0/0x180 [ 1318.047611][ T1082] entry_SYSCALL_64_after_hwframe+0x46/0x4e [ 1318.047887][ T1082] RIP: 0033:0x7f480b70a39d [ 1318.048162][ T1082] Code: 91 ba 0d 00 f7 d8 64 89 02 b8 ff ff ff ff eb b2 e8 18 a3 01 00 0f 1f 84 00 00 00 00 00 80 3d a9 3c 0e 00 00 74 17 31 c0 0f 05 <48> 3d 00 f0 ff ff 77 5b c3 66 2e 0f 1f 84 00 00 00 00 00 53 48 83 [ 1318.048769][ T1082] RSP: 002b:00007ffde77f5c68 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 1318.049083][ T1082] RAX: ffffffffffffffda RBX: 0000000000000800 RCX: 00007f480b70a39d [ 1318.049392][ T1082] RDX: 0000000000000800 RSI: 000055c9f2120c00 RDI: 0000000000000008 [ 1318.049703][ T1082] RBP: 0000000000000800 R08: 000055c9f2120a94 R09: 0000000000000007 [ 1318.050011][ T1082] R10: 0000000000000000 R11: 0000000000000246 R12: 000055c9f2120c00 [ 1318.050324][ T1082] R13: 0000000000000008 R14: 0000000000000008 R15: 0000000000000800 [ 1318.050638][ T1082] </TASK> amdgpu_debugfs_mqd_read() holds a reservation when it calls put_user(), which may fault and acquire the mmap_sem. This violates the established locking order. Bounce the mqd data through a kernel buffer to get put_user() out of the illegal section. Fixes: 445d85e ("drm/amdgpu: add debugfs interface for reading MQDs") Cc: stable@vger.kernel.org # v6.5+ Reviewed-by: Shashank Sharma <shashank.sharma@amd.com> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

vhost_worker will call tun call backs to receive packets. If too many illegal packets arrives, tun_do_read will keep dumping packet contents. When console is enabled, it will costs much more cpu time to dump packet and soft lockup will be detected. net_ratelimit mechanism can be used to limit the dumping rate. PID: 33036 TASK: ffff949da6f20000 CPU: 23 COMMAND: "vhost-32980" #0 [fffffe00003fce50] crash_nmi_callback at ffffffff89249253 #1 [fffffe00003fce58] nmi_handle at ffffffff89225fa3 #2 [fffffe00003fceb0] default_do_nmi at ffffffff8922642e #3 [fffffe00003fced0] do_nmi at ffffffff8922660d #4 [fffffe00003fcef0] end_repeat_nmi at ffffffff89c01663 [exception RIP: io_serial_in+20] RIP: ffffffff89792594 RSP: ffffa655314979e8 RFLAGS: 00000002 RAX: ffffffff89792500 RBX: ffffffff8af428a0 RCX: 0000000000000000 RDX: 00000000000003fd RSI: 0000000000000005 RDI: ffffffff8af428a0 RBP: 0000000000002710 R8: 0000000000000004 R9: 000000000000000f R10: 0000000000000000 R11: ffffffff8acbf64f R12: 0000000000000020 R13: ffffffff8acbf698 R14: 0000000000000058 R15: 0000000000000000 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 #5 [ffffa655314979e8] io_serial_in at ffffffff89792594 #6 [ffffa655314979e8] wait_for_xmitr at ffffffff89793470 #7 [ffffa65531497a08] serial8250_console_putchar at ffffffff897934f6 #8 [ffffa65531497a20] uart_console_write at ffffffff8978b605 #9 [ffffa65531497a48] serial8250_console_write at ffffffff89796558 #10 [ffffa65531497ac8] console_unlock at ffffffff89316124 #11 [ffffa65531497b10] vprintk_emit at ffffffff89317c07 #12 [ffffa65531497b68] printk at ffffffff89318306 #13 [ffffa65531497bc8] print_hex_dump at ffffffff89650765 #14 [ffffa65531497ca8] tun_do_read at ffffffffc0b06c27 [tun] #15 [ffffa65531497d38] tun_recvmsg at ffffffffc0b06e34 [tun] #16 [ffffa65531497d68] handle_rx at ffffffffc0c5d682 [vhost_net] #17 [ffffa65531497ed0] vhost_worker at ffffffffc0c644dc [vhost] #18 [ffffa65531497f10] kthread at ffffffff892d2e72 #19 [ffffa65531497f50] ret_from_fork at ffffffff89c0022f Fixes: ef3db4a ("tun: avoid BUG, dump packet on GSO errors") Signed-off-by: Lei Chen <lei.chen@smartx.com> Reviewed-by: Willem de Bruijn <willemb@google.com> Acked-by: Jason Wang <jasowang@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Link: https://lore.kernel.org/r/20240415020247.2207781-1-lei.chen@smartx.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>

…rnel/git/netfilter/nf-next Pablo Neira Ayuso says: ==================== Netfilter updates for net-next The following patchset contains Netfilter updates for net-next: Patch #1 skips transaction if object type provides no .update interface. Patch #2 skips NETDEV_CHANGENAME which is unused. Patch #3 enables conntrack to handle Multicast Router Advertisements and Multicast Router Solicitations from the Multicast Router Discovery protocol (RFC4286) as untracked opposed to invalid packets. From Linus Luessing. Patch #4 updates DCCP conntracker to mark invalid as invalid, instead of dropping them, from Jason Xing. Patch #5 uses NF_DROP instead of -NF_DROP since NF_DROP is 0, also from Jason. Patch #6 removes reference in netfilter's sysctl documentation on pickup entries which were already removed by Florian Westphal. Patch #7 removes check for IPS_OFFLOAD flag to disable early drop which allows to evict entries from the conntrack table, also from Florian. Patches #8 to #16 updates nf_tables pipapo set backend to allocate the datastructure copy on-demand from preparation phase, to better deal with OOM situations where .commit step is too late to fail. Series from Florian Westphal. Patch #17 adds a selftest with packetdrill to cover conntrack TCP state transitions, also from Florian. Patch #18 use GFP_KERNEL to clone elements from control plane to avoid quick atomic reserves exhaustion with large sets, reporter refers to million entries magnitude. * tag 'nf-next-24-05-12' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next: netfilter: nf_tables: allow clone callbacks to sleep selftests: netfilter: add packetdrill based conntrack tests netfilter: nft_set_pipapo: remove dirty flag netfilter: nft_set_pipapo: move cloning of match info to insert/removal path netfilter: nft_set_pipapo: prepare pipapo_get helper for on-demand clone netfilter: nft_set_pipapo: merge deactivate helper into caller netfilter: nft_set_pipapo: prepare walk function for on-demand clone netfilter: nft_set_pipapo: prepare destroy function for on-demand clone netfilter: nft_set_pipapo: make pipapo_clone helper return NULL netfilter: nft_set_pipapo: move prove_locking helper around netfilter: conntrack: remove flowtable early-drop test netfilter: conntrack: documentation: remove reference to non-existent sysctl netfilter: use NF_DROP instead of -NF_DROP netfilter: conntrack: dccp: try not to drop skb in conntrack netfilter: conntrack: fix ct-state for ICMPv6 Multicast Router Discovery netfilter: nf_tables: remove NETDEV_CHANGENAME from netdev chain event handler netfilter: nf_tables: skip transaction if update object is not implemented ==================== Link: https://lore.kernel.org/r/20240512161436.168973-1-pablo@netfilter.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>

ui_browser__show() is capturing the input title that is stack allocated memory in hist_browser__run(). Avoid a use after return by strdup-ing the string. Committer notes: Further explanation from Ian Rogers: My command line using tui is: $ sudo bash -c 'rm /tmp/asan.log*; export ASAN_OPTIONS="log_path=/tmp/asan.log"; /tmp/perf/perf mem record -a sleep 1; /tmp/perf/perf mem report' I then go to the perf annotate view and quit. This triggers the asan error (from the log file): ``` ==1254591==ERROR: AddressSanitizer: stack-use-after-return on address 0x7f2813331920 at pc 0x7f28180 65991 bp 0x7fff0a21c750 sp 0x7fff0a21bf10 READ of size 80 at 0x7f2813331920 thread T0 #0 0x7f2818065990 in __interceptor_strlen ../../../../src/libsanitizer/sanitizer_common/sanitizer_common_interceptors.inc:461 #1 0x7f2817698251 in SLsmg_write_wrapped_string (/lib/x86_64-linux-gnu/libslang.so.2+0x98251) #2 0x7f28176984b9 in SLsmg_write_nstring (/lib/x86_64-linux-gnu/libslang.so.2+0x984b9) #3 0x55c94045b365 in ui_browser__write_nstring ui/browser.c:60 #4 0x55c94045c558 in __ui_browser__show_title ui/browser.c:266 #5 0x55c94045c776 in ui_browser__show ui/browser.c:288 #6 0x55c94045c06d in ui_browser__handle_resize ui/browser.c:206 #7 0x55c94047979b in do_annotate ui/browsers/hists.c:2458 #8 0x55c94047fb17 in evsel__hists_browse ui/browsers/hists.c:3412 #9 0x55c940480a0c in perf_evsel_menu__run ui/browsers/hists.c:3527 #10 0x55c940481108 in __evlist__tui_browse_hists ui/browsers/hists.c:3613 #11 0x55c9404813f7 in evlist__tui_browse_hists ui/browsers/hists.c:3661 #12 0x55c93ffa253f in report__browse_hists tools/perf/builtin-report.c:671 #13 0x55c93ffa58ca in __cmd_report tools/perf/builtin-report.c:1141 #14 0x55c93ffaf159 in cmd_report tools/perf/builtin-report.c:1805 #15 0x55c94000c05c in report_events tools/perf/builtin-mem.c:374 #16 0x55c94000d96d in cmd_mem tools/perf/builtin-mem.c:516 #17 0x55c9400e44ee in run_builtin tools/perf/perf.c:350 #18 0x55c9400e4a5a in handle_internal_command tools/perf/perf.c:403 #19 0x55c9400e4e22 in run_argv tools/perf/perf.c:447 #20 0x55c9400e53ad in main tools/perf/perf.c:561 #21 0x7f28170456c9 in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58 #22 0x7f2817045784 in __libc_start_main_impl ../csu/libc-start.c:360 #23 0x55c93ff544c0 in _start (/tmp/perf/perf+0x19a4c0) (BuildId: 84899b0e8c7d3a3eaa67b2eb35e3d8b2f8cd4c93) Address 0x7f2813331920 is located in stack of thread T0 at offset 32 in frame #0 0x55c94046e85e in hist_browser__run ui/browsers/hists.c:746 This frame has 1 object(s): [32, 192) 'title' (line 747) <== Memory access at offset 32 is inside this variable HINT: this may be a false positive if your program uses some custom stack unwind mechanism, swapcontext or vfork ``` hist_browser__run isn't on the stack so the asan error looks legit. There's no clean init/exit on struct ui_browser so I may be trading a use-after-return for a memory leak, but that seems look a good trade anyway. Fixes: 05e8b08 ("perf ui browser: Stop using 'self'") Signed-off-by: Ian Rogers <irogers@google.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com> Cc: Ben Gainey <ben.gainey@arm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: James Clark <james.clark@arm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kajol Jain <kjain@linux.ibm.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: K Prateek Nayak <kprateek.nayak@amd.com> Cc: Li Dong <lidong@vivo.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oliver Upton <oliver.upton@linux.dev> Cc: Paran Lee <p4ranlee@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@amd.com> Cc: Sun Haiyong <sunhaiyong@loongson.cn> Cc: Tim Chen <tim.c.chen@linux.intel.com> Cc: Yanteng Si <siyanteng@loongson.cn> Cc: Yicong Yang <yangyicong@hisilicon.com> Link: https://lore.kernel.org/r/20240507183545.1236093-2-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

The code in ocfs2_dio_end_io_write() estimates number of necessary transaction credits using ocfs2_calc_extend_credits(). This however does not take into account that the IO could be arbitrarily large and can contain arbitrary number of extents. Extent tree manipulations do often extend the current transaction but not in all of the cases. For example if we have only single block extents in the tree, ocfs2_mark_extent_written() will end up calling ocfs2_replace_extent_rec() all the time and we will never extend the current transaction and eventually exhaust all the transaction credits if the IO contains many single block extents. Once that happens a WARN_ON(jbd2_handle_buffer_credits(handle) <= 0) is triggered in jbd2_journal_dirty_metadata() and subsequently OCFS2 aborts in response to this error. This was actually triggered by one of our customers on a heavily fragmented OCFS2 filesystem. To fix the issue make sure the transaction always has enough credits for one extent insert before each call of ocfs2_mark_extent_written(). Heming Zhao said: ------ PANIC: "Kernel panic - not syncing: OCFS2: (device dm-1): panic forced after error" PID: xxx TASK: xxxx CPU: 5 COMMAND: "SubmitThread-CA" #0 machine_kexec at ffffffff8c069932 #1 __crash_kexec at ffffffff8c1338fa #2 panic at ffffffff8c1d69b9 #3 ocfs2_handle_error at ffffffffc0c86c0c [ocfs2] #4 __ocfs2_abort at ffffffffc0c88387 [ocfs2] #5 ocfs2_journal_dirty at ffffffffc0c51e98 [ocfs2] #6 ocfs2_split_extent at ffffffffc0c27ea3 [ocfs2] #7 ocfs2_change_extent_flag at ffffffffc0c28053 [ocfs2] #8 ocfs2_mark_extent_written at ffffffffc0c28347 [ocfs2] #9 ocfs2_dio_end_io_write at ffffffffc0c2bef9 [ocfs2] #10 ocfs2_dio_end_io at ffffffffc0c2c0f5 [ocfs2] #11 dio_complete at ffffffff8c2b9fa7 #12 do_blockdev_direct_IO at ffffffff8c2bc09f #13 ocfs2_direct_IO at ffffffffc0c2b653 [ocfs2] #14 generic_file_direct_write at ffffffff8c1dcf14 #15 __generic_file_write_iter at ffffffff8c1dd07b #16 ocfs2_file_write_iter at ffffffffc0c49f1f [ocfs2] #17 aio_write at ffffffff8c2cc72e #18 kmem_cache_alloc at ffffffff8c248dde #19 do_io_submit at ffffffff8c2ccada #20 do_syscall_64 at ffffffff8c004984 #21 entry_SYSCALL_64_after_hwframe at ffffffff8c8000ba Link: https://lkml.kernel.org/r/20240617095543.6971-1-jack@suse.cz Link: https://lkml.kernel.org/r/20240614145243.8837-1-jack@suse.cz Fixes: c15471f ("ocfs2: fix sparse file & data ordering issue in direct io") Signed-off-by: Jan Kara <jack@suse.cz> Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com> Reviewed-by: Heming Zhao <heming.zhao@suse.com> Cc: Mark Fasheh <mark@fasheh.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Junxiao Bi <junxiao.bi@oracle.com> Cc: Changwei Ge <gechangwei@live.cn> Cc: Gang He <ghe@suse.com> Cc: Jun Piao <piaojun@huawei.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

…rnel/git/netfilter/nf-next into main Pablo Neira Ayuso says: ==================== Netfilter/IPVS updates for net-next The following patchset contains Netfilter/IPVS updates for net-next: Patch #1 to #11 to shrink memory consumption for transaction objects: struct nft_trans_chain { /* size: 120 (-32), cachelines: 2, members: 10 */ struct nft_trans_elem { /* size: 72 (-40), cachelines: 2, members: 4 */ struct nft_trans_flowtable { /* size: 80 (-48), cachelines: 2, members: 5 */ struct nft_trans_obj { /* size: 72 (-40), cachelines: 2, members: 4 */ struct nft_trans_rule { /* size: 80 (-32), cachelines: 2, members: 6 */ struct nft_trans_set { /* size: 96 (-24), cachelines: 2, members: 8 */ struct nft_trans_table { /* size: 56 (-40), cachelines: 1, members: 2 */ struct nft_trans_elem can now be allocated from kmalloc-96 instead of kmalloc-128 slab. Series from Florian Westphal. For the record, I have mangled patch #1 to add nft_trans_container_*() and use if for every transaction object. I have also added BUILD_BUG_ON to ensure struct nft_trans always comes at the beginning of the container transaction object. And few minor cleanups, any new bugs are of my own. Patch #12 simplify check for SCTP GSO in IPVS, from Ismael Luceno. Patch #13 nf_conncount key length remains in the u32 bound, from Yunjian Wang. Patch #14 removes unnecessary check for CTA_TIMEOUT_L3PROTO when setting default conntrack timeouts via nfnetlink_cttimeout API, from Lin Ma. Patch #15 updates NFT_SECMARK_CTX_MAXLEN to 4096, SELinux could use larger secctx names than the existing 256 bytes length. Patch #16 adds a selftest to exercise nfnetlink_queue listeners leaving nfnetlink_queue, from Florian Westphal. Patch #17 increases hitcount from 255 to 65535 in xt_recent, from Phil Sutter. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>

A sysfs reader can race with a device reset or removal, attempting to read device state when the device is not actually present. eg: [exception RIP: qed_get_current_link+17] #8 [ffffb9e4f2907c48] qede_get_link_ksettings at ffffffffc07a994a [qede] #9 [ffffb9e4f2907cd8] __rh_call_get_link_ksettings at ffffffff992b01a3 #10 [ffffb9e4f2907d38] __ethtool_get_link_ksettings at ffffffff992b04e4 #11 [ffffb9e4f2907d90] duplex_show at ffffffff99260300 #12 [ffffb9e4f2907e38] dev_attr_show at ffffffff9905a01c #13 [ffffb9e4f2907e50] sysfs_kf_seq_show at ffffffff98e0145b #14 [ffffb9e4f2907e68] seq_read at ffffffff98d902e3 #15 [ffffb9e4f2907ec8] vfs_read at ffffffff98d657d1 #16 [ffffb9e4f2907f00] ksys_read at ffffffff98d65c3f #17 [ffffb9e4f2907f38] do_syscall_64 at ffffffff98a052fb crash> struct net_device.state ffff9a9d21336000 state = 5, state 5 is __LINK_STATE_START (0b1) and __LINK_STATE_NOCARRIER (0b100). The device is not present, note lack of __LINK_STATE_PRESENT (0b10). This is the same sort of panic as observed in commit 4224cfd ("net-sysfs: add check for netdevice being present to speed_show"). There are many other callers of __ethtool_get_link_ksettings() which don't have a device presence check. Move this check into ethtool to protect all callers. Fixes: d519e17 ("net: export device speed and duplex via sysfs") Fixes: 4224cfd ("net-sysfs: add check for netdevice being present to speed_show") Signed-off-by: Jamie Bainbridge <jamie.bainbridge@gmail.com> Link: https://patch.msgid.link/8bae218864beaa44ed01628140475b9bf641c5b0.1724393671.git.jamie.bainbridge@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>

In binder_add_freeze_work() we iterate over the proc->nodes with the proc->inner_lock held. However, this lock is temporarily dropped in order to acquire the node->lock first (lock nesting order). This can race with binder_node_release() and trigger a use-after-free: ================================================================== BUG: KASAN: slab-use-after-free in _raw_spin_lock+0xe4/0x19c Write of size 4 at addr ffff53c04c29dd04 by task freeze/640 CPU: 5 UID: 0 PID: 640 Comm: freeze Not tainted 6.11.0-07343-ga727812a8d45 #17 Hardware name: linux,dummy-virt (DT) Call trace: _raw_spin_lock+0xe4/0x19c binder_add_freeze_work+0x148/0x478 binder_ioctl+0x1e70/0x25ac __arm64_sys_ioctl+0x124/0x190 Allocated by task 637: __kmalloc_cache_noprof+0x12c/0x27c binder_new_node+0x50/0x700 binder_transaction+0x35ac/0x6f74 binder_thread_write+0xfb8/0x42a0 binder_ioctl+0x18f0/0x25ac __arm64_sys_ioctl+0x124/0x190 Freed by task 637: kfree+0xf0/0x330 binder_thread_read+0x1e88/0x3a68 binder_ioctl+0x16d8/0x25ac __arm64_sys_ioctl+0x124/0x190 ================================================================== Fix the race by taking a temporary reference on the node before releasing the proc->inner lock. This ensures the node remains alive while in use. Fixes: d579b04 ("binder: frozen notification") Cc: stable@vger.kernel.org Reviewed-by: Alice Ryhl <aliceryhl@google.com> Acked-by: Todd Kjos <tkjos@google.com> Signed-off-by: Carlos Llamas <cmllamas@google.com> Link: https://lore.kernel.org/r/20240926233632.821189-2-cmllamas@google.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Current loop calls vfs_statfs() while holding the q->limits_lock. If FS takes some locking in vfs_statfs callback, this may lead to ABBA locking bug (at least, FAT fs has this issue actually). So this patch calls vfs_statfs() outside q->limits_locks instead, because looks like no reason to hold q->limits_locks while getting discord configs. Chain exists of: &sbi->fat_lock --> &q->q_usage_counter(io)#17 --> &q->limits_lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&q->limits_lock); lock(&q->q_usage_counter(io)#17); lock(&q->limits_lock); lock(&sbi->fat_lock); *** DEADLOCK *** Reported-by: syzbot+a5d8c609c02f508672cc@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=a5d8c609c02f508672cc Reviewed-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Signed-off-by: Jens Axboe <axboe@kernel.dk>

alobakin requested a review from walking-machine February 24, 2023 13:06

alobakin force-pushed the iavf-xdp branch 2 times, most recently from 41f3642 to 8862fe7 Compare March 1, 2023 16:57

alobakin changed the base branch from bpf-next to net-next March 1, 2023 17:00

alobakin force-pushed the iavf-xdp branch 2 times, most recently from 777041a to 6e1d7f5 Compare March 3, 2023 18:04

alobakin force-pushed the net-next branch from 4d42cd6 to db47fa2 Compare March 9, 2023 13:39

alobakin force-pushed the iavf-xdp branch 4 times, most recently from fe96a55 to e327e47 Compare March 10, 2023 09:39

alobakin force-pushed the iavf-xdp branch from e327e47 to 966873b Compare March 24, 2023 18:43

alobakin and others added 18 commits March 24, 2023 20:33

iavf: optimize Rx hotpath a bunch -- vol. 2

3916621

Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

walking-machine and others added 7 commits March 24, 2023 21:00

iavf: Implement AF_XDP RX processing

d5b13b4

Implement RX packet processing specific to AF_XDP ZC. All actions except XDP_PASS are supported, the skb path will be implemented in later patches. Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>

iavf: Implement XDP_PASS path in AF_XDP processing

112110e

Construct skb and fill in its fields, when AF_XDP is enabled on the ring, if XDP program returns XDP_PASS. (will be fixed up). Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>

iavf: fixup for optimize vol. 2

0ec8e9a

Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>

alobakin force-pushed the iavf-xdp branch from 966873b to 0ec8e9a Compare March 24, 2023 20:02

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

iavf: XDP Page Pool edition #17

iavf: XDP Page Pool edition #17

alobakin commented Feb 24, 2023

iavf: XDP Page Pool edition #17

Are you sure you want to change the base?

iavf: XDP Page Pool edition #17

Conversation

alobakin commented Feb 24, 2023