summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2024-05-06bcachefs: Fix assert in bch2_alloc_v4_invalid()Kent Overstreet
Reported-by: syzbot+10827fa6b176e1acf1d0@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: fix overflow in fiemapReed Riley
filefrag (and potentially other utilities that call fiemap) sometimes pass ULONG_MAX as the length. fiemap_prep clamps excessively large lengths - but the calculation of end can overflow if it occurs before calling fiemap_prep. When this happens, filefrag assumes it has read to the end and exits. Signed-off-by: Reed Riley <reed@riley.engineer> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Add a better limit for maximum number of bucketsKent Overstreet
The bucket_gens array is a single array allocation (one byte per bucket), and kernel allocations are still limited to INT_MAX. Check this limit to avoid failing the bucket_gens array allocation. Reported-by: syzbot+b29f436493184ea42e2b@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Fix lifetime issue in device iterator helpersKent Overstreet
bch2_get_next_dev() and bch2_get_next_online_dev() iterate over devices, dropping and taking refs as they go; we can't access the previous device (for ca->dev_idx) after we've dropped our ref to it, unless we take rcu_read_lock() first. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Fix bch2_dev_lookup() refcountingKent Overstreet
bch2_dev_lookup() is supposed to take a ref on the device it returns, but for_each_member_device() takes refs as it iterates, for_each_member_device_rcu() does not. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Initialize bch_write_op->failed in inline data pathKent Overstreet
Normally this is initialized in __bch2_write(), which is executed in a loop, but the inline data path skips this. Reported-by: syzbot+fd3ccb331eb21f05d13b@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Fix refcount put in sb_field_resize error pathKent Overstreet
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Inodes need extra padding for varint_decode_fast()Kent Overstreet
Reported-by: syzbot+66b9b74f6520068596a9@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: Fix early error path in bch2_fs_btree_key_cache_exit()Kent Overstreet
Reported-by: syzbot+a35cdb62ec34d44fb062@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: bucket_pos_to_bp_noerror()Kent Overstreet
We don't want the assert when we're checking if the backpointer is valid. Reported-by: syzbot+bf7215c0525098e7747a@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06bcachefs: don't free error pointersKent Overstreet
Reported-by: syzbot+3333603f569fc2ef258c@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06wifi: iwlwifi: Ensure prph_mac dump includes all addressesDaniel Gabay
In prph_mac_iter, ensure that all required addresses are dumped even if a read fails. Currently, if a read fails, the region dump is stopped, preventing the creation of prph_mac.lst. By dumping all addresses even if a read fails, we can accurately determine which addresses were successfully read and which were not. Signed-off-by: Daniel Gabay <daniel.gabay@intel.com> Reviewed-by: Eilon Rinat <eilon.rinat@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.31fa9ce91a1c.Ia0c86f70c7a6874c15ffc6f8235aa88530208546@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: don't request statistics in restartMiri Korenblit
During restart mac80211 notifies the driver about the association, (if we was associated before the restart) which causes the driver to request statistics from the FW. This causes to an immediate exit from EMLSR after the restart is done, when the statistics notif is handled. (too low TPT). There is no point in requesting statistics wnyway, since the FW just started and don't have any. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Reviewed-by: Johannes Berg <johannes.berg@intel.com> Link: https://msgid.link/20240506095953.16638dec9f7b.I093514312179bae566ad8d73ffb0355c6eee288a@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: exit EMLSR if secondary link is not usedMiri Korenblit
Exit EMLSR mode if the secondary link is not used enough for Rx/Tx Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Reviewed-by: Johannes Berg <johannes.berg@intel.com> Link: https://msgid.link/20240506095953.99ad1d71e9b9.Ide825433488ec809773efdc36937e3089d0012df@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: add beacon template version 14striebit
In version 14 tim_size became the offset of the broadcast TWT IE. Signed-off-by: striebit <shaul.triebitz@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.76957de93810.I2c718b0d648f2559fe1337df39915c5e772856bc@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: align UATS naming with firmwareJohannes Berg
The firmware has different names for this, which is confusing as even the convention of having the firmware name in a comment after the struct definition wasn't met here. Fix the naming, but keep UATS in some of it since that's the BIOS name. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.b0dfe17d5f44.I8f5f5a831c7b934ce3140f838315827c018103bb@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: Force SCU_ACTIVE for specific platformsDaniel Gabay
Firmware 0x2F7 assert observed in Dell platforms when using GL HW. This issue is mitigated by setting SCU_FORCE_ACTIVE during platform low power states. Driver shall indicate firmware to force SCU active by setting bit 29 in context info prph scratch control flags. This mitigation is limited to Dell platforms with GL HW only. Signed-off-by: Daniel Gabay <daniel.gabay@intel.com> Reviewed-by: Ofer Kimelman <ofer.kimelman@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.3d0c56c2bb1a.I97d9da402890d2085b5698666cceffc417b6b6df@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: record and return channel survey informationBenjamin Berg
While doing a passive scan, the firmware will report per-channel survey information. This information is primarily useful for hostapd when doing an ACS (Automatic Channel Selection). Collect this information and add it to the result set when getting the survey information. Signed-off-by: Benjamin Berg <benjamin.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.9287591a5999.I54a3f9f6480d3694e67eea1cb4f5853beace2780@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: add the firmware API for channel surveyBenjamin Berg
When requested, the firmware can return per-channel survey information generally used for ACS (automatic channel selection). Add the API for this, which consists of a flag and a new channel survey notification. Signed-off-by: Benjamin Berg <benjamin.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.1facde532676.I3864ac4bc0fecb7fd5136e85c07585ab7100234b@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Fix race in scan completionIlan Peer
The move of the scan complete notification handling to the wiphy worker introduced a race between scan complete notification and scan abort: - The wiphy lock is held, e.g., for rfkill handling etc. - Scan complete notification is received but not handled yet. - Scan abort is triggered, and scan abort is sent to the FW. Once the scan abort command is sent successfully, the flow synchronously waits for the scan complete notification. However, as the scan complete notification was already received but not processed yet, this hangs for a second and continues leaving the scan status in an inconsistent state. - Once scan complete handling is started (when the wiphy lock is not held) since the scan status is not an inconsistent state, a warning is issued and the scan complete notification is not handled. To fix this issue, switch back the scan complete notification to be asynchronously handling, and only move the link selection logic to a worker (which was the original reason for the move to use wiphy lock). While at it, refactor some prints to improve debug data. Fixes: 07bf5297d392 ("wifi: iwlwifi: mvm: Implement new link selection algorithm") Signed-off-by: Ilan Peer <ilan.peer@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.1f484a86324b.I63ed445a47f144546948c74ae6df85587fdb4ce3@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Add a print for invalid link pair due to bandwidthYedidya Benshimol
When validating a link pair for EMLSR, add a print for invalid link pair due to bandwidth Signed-off-by: Yedidya Benshimol <yedidya.ben.shimol@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.9e57ad898cf4.Id8edfd5e3774ea6475d5f4178ab7ea75a870ef95@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: add a debugfs for reading EMLSR blocking reasonsYedidya Benshimol
Add a reading for all active EMLSR blocking reasons for testing purposes. Signed-off-by: Yedidya Benshimol <yedidya.ben.shimol@intel.com> Reviewed-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.6d494a335e81.Ic0fa6a9636e3c1a3b1420e85e704a19d4a56e8d9@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Add active EMLSR blocking reasons printsYedidya Benshimol
Upon adding/removing an EMLSR blocking reason add to the print the EMLSR disabling mask Signed-off-by: Yedidya Benshimol <yedidya.ben.shimol@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.1e34fe2c3e51.Ia7db0392d81818ceb70a7b199d3f5fa8a4ad198d@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: bump FW API to 90 for BZ/SC devicesMiri Korenblit
Start supporting API version 90 for new devices. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.4e4b19128b56.I2f9196191f1ea78e96e92f9db8ecb3cc9bbfd9b3@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: fix primary link settingMiri Korenblit
mvmvif::primary link holds the ID and not a bitmap. Fix this Fixes: 07bf5297d392 ("wifi: iwlwifi: mvm: Implement new link selection algorithm") Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Reviewed-by: Johannes Berg <johannes.berg@intel.com> Link: https://msgid.link/20240506095953.779bf6949053.Ia9297991ff2fdc82ae7c730e0069e2dd6e5f2902@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: use already determined cmd_idJohannes Berg
In iwl_mvm_rs_fw_rate_init() we have a variable cmd_id that holds the command ID, so we can just use that instead of the various calculations of it. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240506095953.f894ede03b26.I18f03c272b1c0807767f2713f3ffbb2941c57d9b@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: don't reset link selection during restartJohannes Berg
After restart, we might want to end up with the same config as before, even for multi-link/EMLSR. Therefore, don't reset the stored link selection result in that case. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.e81db303f1dc.Ie8267082f623d14376a2052d222e18da6545f34b@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: Print EMLSR states nameDaniel Gabay
This is useful for debug instead of looking for the hex value. Signed-off-by: Daniel Gabay <daniel.gabay@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.f3509cf652f2.Ic086b6b2132ffe249b3c4bdd24c673ce7fd1b614@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Block EMLSR when a p2p/softAP vif is activeYedidya Benshimol
When there's an active link in a non-station vif, the station vif is not allowed to enter EMLSR Note that blocking EMLSR by calling iwl_mvm_block_esr() we will schedule an exit from EMLSR worker, but the worker cannot run before the activation of the non-BSS link, as ieee80211_remain_on_channel already holds the wiphy mutex. Handle that by explicitly calling ieee80211_set_active_links() to leave EMLSR, and then doing iwl_mvm_block_esr() only for consistency and to avoid re-entering it before ready. Note that a call to ieee80211_set_active_links requires to release the mvm mutex, but that's ok since we still hold the wiphy lock. The only thing that might race here is the ESR_MODE_NOTIF, so this changes its handler to run under the wiphy lock. Signed-off-by: Yedidya Benshimol <yedidya.ben.shimol@intel.com> Co-developed-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.916193759f8a.Idf3a3caf5cdc3e69c81710b7ceb57e87f2de87e4@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: fix typo in debug printMiri Korenblit
Change EMSLR to EMLSR Fixes: 6cf7df9f013f ("wifi: iwlwifi: mvm: Add helper functions to update EMLSR status") Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.db629302bfdc.I135e28b89fab3b614ad8758c0305834934f8c0af@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: exit EMLSR when CSA happensJohannes Berg
If CSA is happening, then exit EMLSR to keep the better link, which is the primary link unless that's doing the CSA with quiet. This is done because we can't transmit the OMN frame on a quiet link, but want to exit EMLSR during CSA for better beacon reception, so we can follow the switch accurately. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.3ffff9577f08.I2620971fa5aef789e0d4a588def4c2621e8bed5b@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Disable/enable EMLSR due to link's bandwidth/bandYedidya Benshimol
Enable EMLSR when bandwidth settings meet the criteria in both band and width, otherwise disable. Signed-off-by: Yedidya Benshimol <yedidya.ben.shimol@intel.com> Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.4e473d4f7f5c.I3adf5619b60bfba8af0cd7eae9dac947419603b6@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: avoid always prefering single-linkMiri Korenblit
The new link selection algorithm uses defaults values for BSS load if the BSS Load element was not published by the AP. For 6 GHz, that value is 0. So if the best link is 6 GHz, the EMLSR grade to always be equal to the grade of the best link, and then the best link grade is getting a bonus of 10 percent, meaning that we will never activate EMLSR. Change the logic to not give a bonus for the best link. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.4614e6891dbd.Ie40eae0dd99d82ba60dea5b6dbcd42dcdf16b90d@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: trigger link selection upon TTLM start/endMiri Korenblit
When non default TTLM is applied, mac80211 may force us to use a specific link (For example, if the only active link becomes a dormant link, mac80211 will pick the first usable link and set it as active). When default TTLM is applied, we have new usable links that we might want to select. Therefore, trigger MLO scan and link selection upon change in TTLM. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Reviewed-by: Ilan Peer <ilan.peer@intel.com> Link: https://msgid.link/20240505091420.ed2b386566a8.I0168e61da86b2027633743aaf5d97e483991f0dc@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: consider FWs recommendation for EMLSRMiri Korenblit
FW sends a notification indicating whether activating EMLSR mode is recommended or not. Support the notification and enter EMLSR only if recommended. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.2fd3387882eb.I7a8a5b24658744ed732bfc03b1872c9298483d62@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Activate EMLSR based on traffic volumeMiri Korenblit
Adjust EMLSR activation to account for traffic levels. By tracking the number of RX/TX MPDUs, EMLSR will be activated only when traffic volume meets the required threshold. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.9480f99ac8fc.If9eb946e929a39e10fe5f4638bc8bc3f8976edf1@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: don't always unblock EMLSRMiri Korenblit
When an event occurs to unblock EMLSR, the code attempts to re-enable EMLSR. However, the current implementation always tries to activate EMLSR, regardless of whether the blocker was set before the unblocking event or not. If EMLSR was already unblocked, there is no need to re-activate it. Fixes: 6cf7df9f013f ("wifi: iwlwifi: mvm: Add helper functions to update EMLSR status") Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.eb861402dac9.I6a1d9f774f5551cfab60ea37b71a62640496af9b@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: Always allow entering EMLSR from debugfsMiri Korenblit
EMLSR can't be activated from mac80211. Except for the debugfs, which is intended for testing purposes. Currently we don't allow entering EMLSR from debugfs if EMLSR is blocked, i.e. if mvmvif::esr_disable_reason is not 0. But we need a way to activate EMLSR regardless of the vif being blocked, for testing. Remove the check of esr_disable_reason Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.bc3c24d9e0e6.Iad60e22a0d7e2b2b989051e1140b6dc98bef7bcc@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: add a debugfs for (un)blocking EMLSRMiri Korenblit
This is needed for testing purposes. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.eba2b6f0664c.I5f058e02abda11bf2eccfd2bcb59ca26bae87a3a@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: mvm: trigger link selection after exiting EMLSRMiri Korenblit
If the reason for exiting EMLSR was a blocking reason, wait for the corresponding unblocking event: - if there is an ongoing scan - do nothing. Link selection will be triggered at the end of it. - If more than 30 seconds passed since the exit, trigger MLO scan, which will trigger link selection - If less then 30 seconds passed since exit, reuse the latest link selection result If the reason for exiting EMLSR was an exit reason (IWL_MVM_EXIT_*), schedule MLO scan in 30 seconds. Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Reviewed-by: Ilan Peer <ilan.peer@intel.com> Link: https://msgid.link/20240505091420.6a808c4ae8f5.Ia79605838eb6deee9358bec633ef537f2653db92@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06wifi: iwlwifi: cleanup EMLSR when BT is active handlingMiri Korenblit
BT Coex disables EMLSR only for a 2.4 GHz link, but doesn't block the vif from using EMLSR with a different link pair. In addition, storing it in mvmvif:disable_esr_reason requires extracting the BT Coex bit before checking if EMLSR is blocked or not for a specific vif. Therefore, change the BT Coex bit to be an exit reason and not a blocker. On link selection, EMLSR mode will be re-calculated for the 2.4 GHz link instead of checking that bit. While at it, move the relevant function declarations to the EMLSR functions area in mvm.h Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com> Link: https://msgid.link/20240505091420.a2e93b67c895.I183a0039ef076613144648cc46fbe9ab3d47c574@changeid Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06Merge wireless into wireless-nextJohannes Berg
Given how late we are in the cycle, merge the two fixes from wireless into wireless-next as they don't see that urgent. This way, the wireless tree won't need rebasing later. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2024-05-06bcachefs: Fix a scheduler splat in __bch2_next_write_buffer_flush_journal_buf()Kent Overstreet
We're using mutex_lock() inside a wait_event() conditional - prepare_to_wait() has already flipped task state, so potentially blocking ops need annotation. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-06net: fix out-of-bounds access in ops_initThadeu Lima de Souza Cascardo
net_alloc_generic is called by net_alloc, which is called without any locking. It reads max_gen_ptrs, which is changed under pernet_ops_rwsem. It is read twice, first to allocate an array, then to set s.len, which is later used to limit the bounds of the array access. It is possible that the array is allocated and another thread is registering a new pernet ops, increments max_gen_ptrs, which is then used to set s.len with a larger than allocated length for the variable array. Fix it by reading max_gen_ptrs only once in net_alloc_generic. If max_gen_ptrs is later incremented, it will be caught in net_assign_generic. Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com> Fixes: 073862ba5d24 ("netns: fix net_alloc_generic()") Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20240502132006.3430840-1-cascardo@igalia.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-05-06Merge branch 'add-tcp-fraglist-gro-support'Paolo Abeni
Felix Fietkau says: ==================== Add TCP fraglist GRO support When forwarding TCP after GRO, software segmentation is very expensive, especially when the checksum needs to be recalculated. One case where that's currently unavoidable is when routing packets over PPPoE. Performance improves significantly when using fraglist GRO implemented in the same way as for UDP. When NETIF_F_GRO_FRAGLIST is enabled, perform a lookup for an established socket in the same netns as the receiving device. While this may not cover all relevant use cases in multi-netns configurations, it should be good enough for most configurations that need this. Here's a measurement of running 2 TCP streams through a MediaTek MT7622 device (2-core Cortex-A53), which runs NAT with flow offload enabled from one ethernet port to PPPoE on another ethernet port + cake qdisc set to 1Gbps. rx-gro-list off: 630 Mbit/s, CPU 35% idle rx-gro-list on: 770 Mbit/s, CPU 40% idle Changes since v4: - add likely() to prefer the non-fraglist path in check Changes since v3: - optimize __tcpv4_gso_segment_csum - add unlikely() - reorder dev_net/skb_gro_network_header calls after NETIF_F_GRO_FRAGLIST check - add support for ipv6 nat - drop redundant pskb_may_pull check Changes since v2: - create tcp_gro_header_pull helper function to pull tcp header only once - optimize __tcpv4_gso_segment_list_csum, drop obsolete flags check Changes since v1: - revert bogus tcp flags overwrite on segmentation - fix kbuild issue with !CONFIG_IPV6 - only perform socket lookup for the first skb in the GRO train Changes since RFC: - split up patches - handle TCP flags mutations ==================== Link: https://lore.kernel.org/r/20240502084450.44009-1-nbd@nbd.name Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-05-06net: add heuristic for enabling TCP fraglist GROFelix Fietkau
When forwarding TCP after GRO, software segmentation is very expensive, especially when the checksum needs to be recalculated. One case where that's currently unavoidable is when routing packets over PPPoE. Performance improves significantly when using fraglist GRO implemented in the same way as for UDP. When NETIF_F_GRO_FRAGLIST is enabled, perform a lookup for an established socket in the same netns as the receiving device. While this may not cover all relevant use cases in multi-netns configurations, it should be good enough for most configurations that need this. Here's a measurement of running 2 TCP streams through a MediaTek MT7622 device (2-core Cortex-A53), which runs NAT with flow offload enabled from one ethernet port to PPPoE on another ethernet port + cake qdisc set to 1Gbps. rx-gro-list off: 630 Mbit/s, CPU 35% idle rx-gro-list on: 770 Mbit/s, CPU 40% idle Acked-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-05-06net: create tcp_gro_header_pull helper functionFelix Fietkau
Pull the code out of tcp_gro_receive in order to access the tcp header from tcp4/6_gro_receive. Acked-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-05-06net: create tcp_gro_lookup helper functionFelix Fietkau
This pulls the flow port matching out of tcp_gro_receive, so that it can be reused for the next change, which adds the TCP fraglist GRO heuristic. Acked-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-05-06net: add code for TCP fraglist GROFelix Fietkau
This implements fraglist GRO similar to how it's handled in UDP, however no functional changes are added yet. The next change adds a heuristic for using fraglist GRO instead of regular GRO. Acked-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-05-06net: add support for segmenting TCP fraglist GSO packetsFelix Fietkau
Preparation for adding TCP fraglist GRO support. It expects packets to be combined in a similar way as UDP fraglist GSO packets. For IPv4 packets, NAT is handled in the same way as UDP fraglist GSO. Acked-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
generated by cgit (git 2.34.1) at 2025-09-16 01:37:29 +0000