- 12 Oct, 2021 31 commits
-
-
Jakub Kicinski authored
Leon Romanovsky says: ==================== devlink reload simplification Simplify devlink reload APIs. ==================== Link: https://lore.kernel.org/r/cover.1634044267.git.leonro@nvidia.comSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Leon Romanovsky authored
Commit a0c76345 ("devlink: disallow reload operation during device cleanup") added devlink_reload_{enable,disable}() APIs to prevent reload operation from racing with device probe/dismantle. After recent changes to move devlink_register() to the end of device probe and devlink_unregister() to the beginning of device dismantle, these races can no longer happen. Reload operations will be denied if the devlink instance is unregistered and devlink_unregister() will block until all in-flight operations are done. Therefore, remove these devlink_reload_{enable,disable}() APIs. Signed-off-by:
Leon Romanovsky <leonro@nvidia.com> Signed-off-by:
-
Leon Romanovsky authored
Mulitport slave device doesn't support devlink reload, so instead of complicating initialization flow with devlink_reload_enable() which will be removed in next patch, don't set DEVLINK_F_RELOAD feature bit for such devices. This fixes an error when reload counters exposed (and equal zero) for the mode that is not supported at all. Fixes: d89ddaae ("net/mlx5: Disable devlink reload for multi port slave device") Signed-off-by:
-
Leon Romanovsky authored
Introduce new devlink call to set feature mask to control devlink behavior during device initialization phase after devlink_alloc() is already called. This allows us to set reload ops based on device property which is not known at the beginning of driver initialization. For the sake of simplicity, this API lacks any type of locking and needs to be called before devlink_register() to make sure that no parallel access to the ops is possible at this stage. Signed-off-by:
-
Leon Romanovsky authored
Initial annotation patch to separate calls that needs to be executed before or after devlink_register(). Signed-off-by:
-
Leon Romanovsky authored
Both netdev_to_devlink and netdev_to_devlink_port are used in devlink.c only, so move them in order to reduce their scope. Signed-off-by:
-
Leon Romanovsky authored
The declaration of struct devlink in general header provokes the situation where internal fields can be accidentally used by the driver authors. In order to reduce such possible situations, let's reduce the namespace exposure of struct devlink. Signed-off-by:
-
Jakub Kicinski authored
I recently added a variable called addr to tulip_init_one() but for sparc there's already a variable called that half way thru the function. Rename it to fix build. Fixes: ca879317 ("ethernet: tulip: remove direct netdev->dev_addr writes") Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Hao Chen authored
Add a file node "page_pool_info" for debugfs, then cat this file node to dump page pool info as below: QUEUE_ID ALLOCATE_CNT FREE_CNT POOL_SIZE(PAGE_NUM) ORDER NUMA_ID MAX_LEN 0 512 0 512 0 2 4K 1 512 0 512 0 2 4K 2 512 0 512 0 2 4K 3 512 0 512 0 2 4K 4 512 0 512 0 2 4K Signed-off-by:
Hao Chen <chenhao288@hisilicon.com> Signed-off-by:
Guangbin Huang <huangguangbin2@huawei.com> Signed-off-by:
-
Jakub Kicinski authored
I missed removing i from the array index when converting from a loop to a direct copy. Fixes: ca879317 ("ethernet: tulip: remove direct netdev->dev_addr writes") Reported-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
David S. Miller authored
Daniel Borkmann says: ==================== Managed Neighbor Entries This series adds a couple of fixes related to NTF_EXT_LEARNED and NTF_USE neighbor flags, extends the UAPI with a new NDA_FLAGS_EXT netlink attribute in order to be able to add new neighbor flags from user space given all current struct ndmsg / ndm_flags bits are used up. Finally, the core of this series adds a new NTF_EXT_MANAGED flag to neighbors, which allows user space control planes to add 'managed' neighbor entries. Meaning, user space may either transition existing entries or can push down new L3 entries without lladdr into the kernel where the latter will periodically try to keep such NTF_EXT_MANAGED managed entries in reachable state. Main use case for this series are XDP / tc BPF load-balancers which make use of the bpf_fib_lookup() helper for backends. For more details, please see individual patches. Thanks! ==================== Signed-off-by: -
Daniel Borkmann authored
Allow a user space control plane to insert entries with a new NTF_EXT_MANAGED flag. The flag then indicates to the kernel that the neighbor entry should be periodically probed for keeping the entry in NUD_REACHABLE state iff possible. The use case for this is targeting XDP or tc BPF load-balancers which use the bpf_fib_lookup() BPF helper in order to piggyback on neighbor resolution for their backends. Given they cannot be resolved in fast-path, a control plane inserts the L3 (without L2) entries manually into the neighbor table and lets the kernel do the neighbor resolution either on the gateway or on the backend directly in case the latter resides in the same L2. This avoids to deal with L2 in the control plane and to rebuild what the kernel already does best anyway. NTF_EXT_MANAGED can be combined with NTF_EXT_LEARNED in order to avoid GC eviction. The kernel then adds NTF_MANAGED flagged entries to a per-neighbor table which gets triggered by the system work queue to periodically call neigh_event_send() for performing the resolution. The implementation allows migration from/to NTF_MANAGED neighbor entries, so that already existing entries can be converted by the control plane if needed. Potentially, we could make the interval for periodically calling neigh_event_send() configurable; right now it's set to DELAY_PROBE_TIME which is also in line with mlxsw which has similar driver-internal infrastructure c723c735 ("mlxsw: spectrum_router: Periodically update the kernel's neigh table"). In future, the latter could possibly reuse the NTF_MANAGED neighbors as well. Example: # ./ip/ip n replace 192.168.178.30 dev enp5s0 managed extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a managed extern_learn REACHABLE [...] Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Acked-by:
Roopa Prabhu <roopa@nvidia.com> Link: https://linuxplumbersconf.org/event/11/contributions/953/Signed-off-by:
-
Roopa Prabhu authored
Currently, all bits in struct ndmsg's ndm_flags are used up with the most recent addition of 435f2e7c ("net: bridge: add support for sticky fdb entries"). This makes it impossible to extend the neighboring subsystem with new NTF_* flags: struct ndmsg { __u8 ndm_family; __u8 ndm_pad1; __u16 ndm_pad2; __s32 ndm_ifindex; __u16 ndm_state; __u8 ndm_flags; __u8 ndm_type; }; There are ndm_pad{1,2} attributes which are not used. However, due to uncareful design, the kernel does not enforce them to be zero upon new neighbor entry addition, and given they've been around forever, it is not possible to reuse them today due to risk of breakage. One option to overcome this limitation is to add a new NDA_FLAGS_EXT attribute for extended flags. In struct neighbour, there is a 3 byte hole between protocol and ha_lock, which allows neigh->flags to be extended from 8 to 32 bits while still being on the same cacheline as before. This also allows for all future NTF_* flags being in neigh->flags rather than yet another flags field. Unknown flags in NDA_FLAGS_EXT will be rejected by the kernel. Co-developed-by:
-
Daniel Borkmann authored
Currently, it is not possible to migrate a neighbor entry between NUD_PERMANENT state and NTF_USE flag with a dynamic NUD state from a user space control plane. Similarly, it is not possible to add/remove NTF_EXT_LEARNED flag from an existing neighbor entry in combination with NTF_USE flag. This is due to the latter directly calling into neigh_event_send() without any meta data updates as happening in __neigh_update(). Thus, to enable this use case, extend the latter with a NEIGH_UPDATE_F_USE flag where we break the NUD_PERMANENT state in particular so that a latter neigh_event_send() is able to re-resolve a neighbor entry. Before fix, NUD_PERMANENT -> NUD_* & NTF_USE: # ./ip/ip n replace 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a PERMANENT [...] # ./ip/ip n replace 192.168.178.30 dev enp5s0 use extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a PERMANENT [...] As can be seen, despite the admin-triggered replace, the entry remains in the NUD_PERMANENT state. After fix, NUD_PERMANENT -> NUD_* & NTF_USE: # ./ip/ip n replace 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a PERMANENT [...] # ./ip/ip n replace 192.168.178.30 dev enp5s0 use extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a extern_learn REACHABLE [...] # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a extern_learn STALE [...] # ./ip/ip n replace 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a PERMANENT [...] After the fix, the admin-triggered replace switches to a dynamic state from the NTF_USE flag which triggered a new neighbor resolution. Likewise, we can transition back from there, if needed, into NUD_PERMANENT. Similar before/after behavior can be observed for below transitions: Before fix, NTF_USE -> NTF_USE | NTF_EXT_LEARNED -> NTF_USE: # ./ip/ip n replace 192.168.178.30 dev enp5s0 use # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a REACHABLE [...] # ./ip/ip n replace 192.168.178.30 dev enp5s0 use extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a REACHABLE [...] After fix, NTF_USE -> NTF_USE | NTF_EXT_LEARNED -> NTF_USE: # ./ip/ip n replace 192.168.178.30 dev enp5s0 use # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a REACHABLE [...] # ./ip/ip n replace 192.168.178.30 dev enp5s0 use extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a extern_learn REACHABLE [...] # ./ip/ip n replace 192.168.178.30 dev enp5s0 use # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a REACHABLE [..] Signed-off-by:
-
Daniel Borkmann authored
The NTF_EXT_LEARNED neigh flag is usually propagated back to user space upon dump of the neighbor table. However, when used in combination with NTF_USE flag this is not the case despite exempting the entry from the garbage collector. This results in inconsistent state since entries are typically marked in neigh->flags with NTF_EXT_LEARNED, but here they are not. Fix it by propagating the creation flag to ___neigh_create(). Before fix: # ./ip/ip n replace 192.168.178.30 dev enp5s0 use extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a REACHABLE [...] After fix: # ./ip/ip n replace 192.168.178.30 dev enp5s0 use extern_learn # ./ip/ip n 192.168.178.30 dev enp5s0 lladdr f4:8c:50:5e:71:9a extern_learn REACHABLE [...] Fixes: 9ce33e46 ("neighbour: support for NTF_EXT_LEARNED flag") Signed-off-by:
-
Len Baker authored
As noted in the "Deprecated Interfaces, Language Features, Attributes, and Conventions" documentation [1], size calculations (especially multiplication) should not be performed in memory allocator (or similar) function arguments due to the risk of them overflowing. This could lead to values wrapping around and a smaller allocation being made than the caller was expecting. Using those allocations could lead to linear overflows of heap memory and other misbehaviors. So, take the opportunity to refactor the hnae_handle structure to switch the last member to flexible array, changing the code accordingly. Also, fix the comment in the hnae_vf_cb structure to inform that the ae_handle member must be the last member. Then, use the struct_size() helper to do the arithmetic instead of the argument "size + count * size" in the kzalloc() function. This code was detected with the help of Coccinelle and audited and fixed manually. [1] https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-argumentsSigned-off-by:
Len Baker <len.baker@gmx.com> Signed-off-by:
-
David S. Miller authored
Ido Schimmel says: ==================== mlxsw: Add support for ECN mirroring Petr says: Patches in this set have been floating around for some time now together with trap_fwd support. That will however need more work, time for which is nowhere to be found, apparently. Instead, this patchset enables offload of only packet mirroring on RED mark qevent, enabling mirroring of ECN-marked packets. Formally it enables offload of filters added to blocks bound to the RED qevent mark if: - The switch ASIC is Spectrum-2 or above. - Only a single filter is attached at the block, at chain 0 (the default), and its classifier is matchall. - The filter has hw_stats set to disabled. - The filter has a single action, which is mirror. This differs from early_drop qevent offload, which supports mirroring and trapping. However trapping in context of ECN-marked packets is not suitable, because the HW does not drop the packet, as the trap action implies. And there is as of now no way to express only the part of trapping that transfers the packet to the SW datapath, sans the HW-datapath drop. The patchset progresses as follows: Patch #1 is an extack propagation. Mirroring of ECN-marked packets is configured in the ASIC through an ECN trigger, which is considered "egress", unlike the EARLY_DROP trigger. In patch #2, add a helper to classify triggers as ingress. As clarified above, traps cannot be offloaded on mark qevent. Similarly, given a trap_fwd action, it would not be offloadable on early_drop qevent. In patch #3, introduce support for tracking actions permissible on a given block. Patch #4 actually adds the mark qevent offload. In patch #5, fix a small style issue in one of the selftests, and in patch #6 add mark offload selftests. ==================== Signed-off-by: -
Petr Machata authored
Add do_mark_test(), which is to do_ecn_test() like do_drop_test() is to do_red_test(): meant to test that actions on the RED mark qevent block are offloaded, and executed on ECN-marked packets. The test splits install_qdisc() into its constituents, install_root_qdisc() and install_qdisc_tcX(). This is in order to test that when mirroring is enabled on one TC, the other TC does not mirror. Signed-off-by:
Petr Machata <petrm@nvidia.com> Signed-off-by:
Ido Schimmel <idosch@nvidia.com> Signed-off-by:
-
Petr Machata authored
These variables are cut'n'pasted from other functions in the file and not actually used. Signed-off-by:
-
Petr Machata authored
The RED "mark" qevent can be offloaded under similar conditions as the RED "early_drop" qevent. Therefore recognize its binding type in the TC_SETUP_BLOCK handler and translate to the right SPAN trigger, with the right set of supported actions. Signed-off-by:
-
Petr Machata authored
One block can be bound to several qevents. The qevent type that the block is bound to determines which actions make sense in a given context. In the particular case of mlxsw, trap cannot be offloaded on a RED mark qevent, because the trap contract specifies that the packet is dropped in the HW datapath, and the HW trigger that the action is offloaded to is always forwarding the packet (in addition to marking in). Therefore keep track of which actions are permissible at each binding block. When an attempt is made to bind a certain action at a binding point where it is not supported, bounce the request. Signed-off-by:
-
Petr Machata authored
The following patches will configure the MLXSW_SP_SPAN_TRIGGER_ECN mirroring trigger. This trigger is considered "egress", unlike the previously-offloaded _EARLY_DROP. Add a helper to spectrum_span, mlxsw_sp_span_trigger_is_ingress(), to classify triggers to ingress and egress. Pass result of this instead of hardcoding true when calling mlxsw_sp_span_analyzed_port_get()/_put(). Signed-off-by:
-
Petr Machata authored
This function will report a new failure in the following patches. Pass extack so that the failure is explicable. Signed-off-by:
-
Jakub Kicinski authored
Krzysztof Kozlowski says: ==================== nfc: minor printk cleanup v2: Correct SPDX license in patch 2/7 (as Joe pointed out). v1: Remove unused variable in pn533 (reported by kbuild). ==================== Link: https://lore.kernel.org/r/20211011133835.236347-1-krzysztof.kozlowski@canonical.comSigned-off-by:
-
Krzysztof Kozlowski authored
ftrace is a preferred and standard way to debug entering and exiting functions so drop useless debug prints. Signed-off-by:
Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Signed-off-by:
-
Krzysztof Kozlowski authored
ftrace is a preferred and standard way to debug entering and exiting functions so drop useless debug prints. Signed-off-by:
Mark Greer <mgreer@animalcreek.com> Signed-off-by:
-
Krzysztof Kozlowski authored
ftrace is a preferred and standard way to debug entering and exiting functions so drop useless debug prints. Signed-off-by:
-
Krzysztof Kozlowski authored
ftrace is a preferred and standard way to debug entering and exiting functions so drop useless debug prints. Signed-off-by:
-
Krzysztof Kozlowski authored
Simplify the code dereferencing several pointers to reach the struct device. Signed-off-by:
-
Krzysztof Kozlowski authored
Replace standard GPLv2 license text with SPDX tag. Although the comment mentions GPLv2-only, it refers to the full license file which allows later GPL versions. Signed-off-by:
-
Krzysztof Kozlowski authored
ftrace is a preferred and standard way to debug entering and exiting functions so drop useless debug prints. Signed-off-by:
-
- 11 Oct, 2021 9 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/nexDavid S. Miller authored
t-queue Tony Nguyen says: ==================== 100GbE Intel Wired LAN Driver Updates 2021-10-11 Wojciech Drewek says: This series adds support for adding/removing advanced switch filters in ice driver. Advanced filters are building blocks for HW acceleration of TC orchestration. Add ndo_setup_tc callback implementation for PF and VF port representors (when device is configured in switchdev mode). Define dummy packet headers to allow adding advanced rules in HW. Supported headers, and thus filters, are: - MAC + IPv4 + UDP - MAC + VLAN + IPv4 + UDP - MAC + IPv4 + TCP - MAC + VLAN + IPv4 + TCP - MAC + IPv6 + UDP - MAC + VLAN + IPv6 + UDP - MAC + IPv6 + TCP - MAC + VLAN + IPv6 + TCP ==================== Signed-off-by: -
David S. Miller authored
Jeroen de Borst says: ==================== gve: minor code and performance improvements This patchset contains a number of independent minor code and performance improvements. ==================== Signed-off-by: -
Catherine Sullivan authored
The rx_buf_alloc_fail counter wasn't getting updated. Fixes: 433e274b ("gve: Add stats for gve.") Signed-off-by:
Catherine Sullivan <csully@google.com> Signed-off-by:
Jeroen de Borst <jeroendb@google.com> Signed-off-by:
-
Jordan Kim authored
Half pages are just used for small enough packets. This change allows this to also apply for systems with pages larger than 4 KB. Fixes: 02b0e0c1 ("gve: Rx Buffer Recycling") Signed-off-by:
Jordan Kim <jrkim@google.com> Signed-off-by:
-
Catherine Sullivan authored
Configure XPS when adding tx queues to the notification blocks. Fixes: dbdaa675 ("gve: Move some static functions to a common file") Signed-off-by:
David Awogbemila <awogbemila@google.com> Signed-off-by:
-
John Fraker authored
Don't always reset the driver on a TX timeout. Attempt to recover by kicking the queue in case an IRQ was missed. Fixes: 9e5f7d26 ("gve: Add workqueue and reset support") Signed-off-by:
John Fraker <jfraker@google.com> Signed-off-by:
-
Tao Liu authored
When TX queue is full, attemt to process enough TX completions to avoid stalling the queue. Fixes: f5cedc84 ("gve: Add transmit and receive support") Signed-off-by:
Tao Liu <xliutaox@google.com> Signed-off-by:
-
Catherine Sullivan authored
Add a pagecnt bias field to rx buffer info struct to eliminate needing to increment the atomic page ref count on every pass in the rx hotpath. Also prefetch two packet pages ahead. Fixes: ede3fcf5 ("gve: Add support for raw addressing to the rx path") Signed-off-by:
Yanchun Fu <yangchun@google.com> Signed-off-by:
Nathan Lewis <npl@google.com> Signed-off-by:
-
Yangchun Fu authored
Use napi_complete_done to allow for the use of gro_flush_timeout. Fixes: f5cedc84 ("gve: Add transmit and receive support") Signed-off-by:
-
