- 15 Aug, 2023 11 commits
-
-
Florian Westphal authored
When flushing, individual set elements are disabled in the next generation via the ->flush callback. Catchall elements are not disabled. This is incorrect and may lead to double-deactivations of catchall elements which then results in memory leaks: WARNING: CPU: 1 PID: 3300 at include/net/netfilter/nf_tables.h:1172 nft_map_deactivate+0x549/0x730 CPU: 1 PID: 3300 Comm: nft Not tainted 6.5.0-rc5+ #60 RIP: 0010:nft_map_deactivate+0x549/0x730 [..] ? nft_map_deactivate+0x549/0x730 nf_tables_delset+0xb66/0xeb0 (the warn is due to nft_use_dec() detecting underflow). Fixes: aaa31047 ("netfilter: nftables: add catch-all set element support") Reported-by:
lonial con <kongln9170@gmail.com> Signed-off-by:
Florian Westphal <fw@strlen.de>
-
Florian Westphal authored
Jakub Kicinski says: We've got some new kdoc warnings here: net/netfilter/nft_set_pipapo.c:1557: warning: Function parameter or member '_set' not described in 'pipapo_gc' net/netfilter/nft_set_pipapo.c:1557: warning: Excess function parameter 'set' description in 'pipapo_gc' include/net/netfilter/nf_tables.h:577: warning: Function parameter or member 'dead' not described in 'nft_set' Fixes: 5f68718b ("netfilter: nf_tables: GC transaction API to avoid race with control plane") Fixes: f6c383b8 ("netfilter: nf_tables: adapt set backend to use GC transaction API") Reported-by:
Jakub Kicinski <kuba@kernel.org> Closes: https://lore.kernel.org/netdev/20230810104638.746e46f1@kernel.org/Signed-off-by:
-
Florian Westphal authored
->abort invocation may cause splat on debug kernels: WARNING: suspicious RCU usage net/netfilter/nft_set_pipapo.c:1697 suspicious rcu_dereference_check() usage! [..] rcu_scheduler_active = 2, debug_locks = 1 1 lock held by nft/133554: [..] (nft_net->commit_mutex){+.+.}-{3:3}, at: nf_tables_valid_genid [..] lockdep_rcu_suspicious+0x1ad/0x260 nft_pipapo_abort+0x145/0x180 __nf_tables_abort+0x5359/0x63d0 nf_tables_abort+0x24/0x40 nfnetlink_rcv+0x1a0a/0x22c0 netlink_unicast+0x73c/0x900 netlink_sendmsg+0x7f0/0xc20 ____sys_sendmsg+0x48d/0x760 Transaction mutex is held, so parallel updates are not possible. Switch to _protected and check mutex is held for lockdep enabled builds. Fixes: 212ed75d ("netfilter: nf_tables: integrate pipapo into commit protocol") Signed-off-by: -
Jason Xing authored
In the real workload, I encountered an issue which could cause the RTO timer to retransmit the skb per 1ms with linear option enabled. The amount of lost-retransmitted skbs can go up to 1000+ instantly. The root cause is that if the icsk_rto happens to be zero in the 6th round (which is the TCP_THIN_LINEAR_RETRIES value), then it will always be zero due to the changed calculation method in tcp_retransmit_timer() as follows: icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX); Above line could be converted to icsk->icsk_rto = min(0 << 1, TCP_RTO_MAX) = 0 Therefore, the timer expires so quickly without any doubt. I read through the RFC 6298 and found that the RTO value can be rounded up to a certain value, in Linux, say TCP_RTO_MIN as default, which is regarded as the lower bound in this patch as suggested by Eric. Fixes: 36e31b0a ("net: TCP thin linear timeouts") Suggested-by:
Eric Dumazet <edumazet@google.com> Signed-off-by:
Jason Xing <kernelxing@tencent.com> Reviewed-by:
David S. Miller <davem@davemloft.net>
-
Liang Chen authored
The failure handling procedure destroys page pools for all queues, including those that haven't had their page pool created yet. this patch introduces necessary adjustments to prevent potential risks and inconsistency with the error handling behavior. Fixes: 0ebab78c ("net: veth: add page_pool for page recycling") Acked-by:
Jesper Dangaard Brouer <hawk@kernel.org> Signed-off-by:
Liang Chen <liangchen.linux@gmail.com> Link: https://lore.kernel.org/r/20230812023016.10553-1-liangchen.linux@gmail.comSigned-off-by:
-
Jakub Kicinski authored
Michal Schmidt says: ==================== octeon_ep: fixes for error and remove paths I have an Octeon card that's misconfigured in a way that exposes a couple of bugs in the octeon_ep driver's error paths. It can reproduce the issues that patches 1 & 4 are fixing. Patches 2 & 3 are a result of reviewing the nearby code. ==================== Link: https://lore.kernel.org/r/20230810150114.107765-1-mschmidt@redhat.comSigned-off-by:
-
Michal Schmidt authored
If it fails to get the devices's MAC address, octep_probe exits while leaving the delayed work intr_poll_task queued. When the work later runs, it's a use after free. Move the cancelation of intr_poll_task from octep_remove into octep_device_cleanup. This does not change anything in the octep_remove flow, but octep_device_cleanup is called also in the octep_probe error path, where the cancelation is needed. Note that the cancelation of ctrl_mbox_task has to follow intr_poll_task's, because the ctrl_mbox_task may be queued by intr_poll_task. Fixes: 24d43332 ("octeon_ep: poll for control messages") Signed-off-by:
Michal Schmidt <mschmidt@redhat.com> Link: https://lore.kernel.org/r/20230810150114.107765-5-mschmidt@redhat.comSigned-off-by:
-
Michal Schmidt authored
intr_poll_task may queue ctrl_mbox_task. The function octep_poll_non_ioq_interrupts_cn93_pf does this. When removing the driver and canceling these two works, cancel ctrl_mbox_task last to guarantee it does not run anymore. Fixes: 24d43332 ("octeon_ep: poll for control messages") Signed-off-by:
-
Michal Schmidt authored
tx_timeout_task is canceled too early when removing the driver. Nothing prevents .ndo_tx_timeout from triggering and queuing the work again. Better cancel it after the netdev is unregistered. It's harmless for octep_tx_timeout_task to run in the window between the unregistration and cancelation, because it checks netif_running. Fixes: 862cd659 ("octeon_ep: Add driver framework and device initialization") Signed-off-by:
-
Michal Schmidt authored
The intention was to wait up to 500 ms for the mbox response. The third argument to wait_event_interruptible_timeout() is supposed to be the timeout duration. The driver mistakenly passed absolute time instead. Fixes: 577f0d1b ("octeon_ep: add separate mailbox command and response queues") Signed-off-by:
Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20230810150114.107765-2-mschmidt@redhat.comSigned-off-by:
-
Radhey Shyam Pandey authored
On Zynq UltraScale+ MPSoC ubuntu platform when systemctl issues suspend, network manager bring down the interface and goes into suspend. When it wakes up it again enables the interface. This leads to xilinx-psgtr "PLL lock timeout" on interface bringup, as the power management controller power down the entire FPD (including SERDES) if none of the FPD devices are in use and serdes is not initialized on resume. $ sudo rtcwake -m no -s 120 -v $ sudo systemctl suspend <this does ifconfig eth1 down> $ ifconfig eth1 up xilinx-psgtr fd400000.phy: lane 0 (type 10, protocol 5): PLL lock timeout phy phy-fd400000.phy.0: phy poweron failed --> -110 macb driver is called in this way: 1. macb_close: Stop network interface. In this function, it reset MACB IP and disables PHY and network interface. 2. macb_suspend: It is called in kernel suspend flow. But because network interface has been disabled(netif_running(ndev) is false), it does nothing and returns directly; 3. System goes into suspend state. Some time later, system is waken up by RTC wakeup device; 4. macb_resume: It does nothing because network interface has been disabled; 5. macb_open: It is called to enable network interface again. ethernet interface is initialized in this API but serdes which is power-off by PMUFW during FPD-off suspend is not initialized again and so we hit GT PLL lock issue on open. To resolve this PLL timeout issue always do PS GTR initialization when ethernet device is configured as non-wakeup source. Fixes: f22bd29b ("net: macb: Fix ZynqMP SGMII non-wakeup source resume failure") Fixes: 8b73fa3a ("net: macb: Added ZynqMP-specific initialization") Signed-off-by:
Radhey Shyam Pandey <radhey.shyam.pandey@amd.com> Link: https://lore.kernel.org/r/1691414091-2260697-1-git-send-email-radhey.shyam.pandey@amd.comSigned-off-by:
-
- 14 Aug, 2023 2 commits
-
-
Petr Machata authored
This test verifies whether the encapsulated packets have the correct configured TTL. It does so by sending ICMP packets through the test topology and mirroring them to a gretap netdevice. On a busy host however, more than just the test ICMP packets may end up flowing through the topology, get mirrored, and counted. This leads to potential spurious failures as the test observes much more mirrored packets than the sent test packets, and assumes a bug. Fix this by tightening up the mirror action match. Change it from matchall to a flower classifier matching on ICMP packets specifically. Fixes: 45315673 ("selftests: forwarding: Test changes in mirror-to-gretap") Signed-off-by:
Petr Machata <petrm@nvidia.com> Tested-by:
Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr> Reviewed-by:
Ido Schimmel <idosch@nvidia.com> Reviewed-by:
-
Vlad Buslov authored
This reverts commit 718cb09a. The commit triggers multiple syzbot issues, probably due to possibility of manually creating VLAN 0 on netdevice which will cause the code to delete it since it can't distinguish such VLAN from implicit VLAN 0 automatically created for devices with NETIF_F_HW_VLAN_CTAG_FILTER feature. Reported-by: syzbot+662f783a5cdf3add2719@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/00000000000090196d0602a6167d@google.com/ Reported-by: syzbot+4b4f06495414e92701d5@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/00000000000096ae870602a61602@google.com/ Reported-by: syzbot+d810d3cd45ed1848c3f7@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/0000000000009f0f9c0602a616ce@google.com/ Fixes: 718cb09a ("vlan: Fix VLAN 0 memory leak") Signed-off-by:
Vlad Buslov <vladbu@nvidia.com> Reviewed-by:
-
- 13 Aug, 2023 1 commit
-
-
Russell King (Oracle) authored
Uwe reports: "Most PHYs signal WoL using an interrupt. So disabling interrupts [at shutdown] breaks WoL at least on PHYs covered by the marvell driver." Discussing with Ioana, the problem which was trying to be solved was: "The board in question is a LS1021ATSN which has two AR8031 PHYs that share an interrupt line. In case only one of the PHYs is probed and there are pending interrupts on the PHY#2 an IRQ storm will happen since there is no entity to clear the interrupt from PHY#2's registers. PHY#1's driver will get stuck in .handle_interrupt() indefinitely." Further confirmation that "the two AR8031 PHYs are on the same MDIO bus." With WoL using interrupts to wake the system, in such a case, the system will begin booting with an asserted interrupt. Thus, we need to cope with an interrupt asserted during boot. Solve this instead by disabling interrupts during PHY probe. This will ensure in Ioana's situation that both PHYs of the same type sharing an interrupt line on a common MDIO bus will have their interrupt outputs disabled when the driver probes the device, but before we hook in any interrupt handlers - thus avoiding the interrupt storm. A better fix would be for platform firmware to disable the interrupting devices at source during boot, before control is handed to the kernel. Fixes: e2f016cf ("net: phy: add a shutdown procedure") Link: 20230804071757.383971-1-u.kleine-koenig@pengutronix.de Reported-by:
Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by:
Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Florian Fainelli <florian.fainelli@broadcom.com> Signed-off-by:
-
- 11 Aug, 2023 3 commits
-
-
Xiang Yang authored
The reference of pdev->dev is taken by of_find_device_by_node, so it should be released when not need anymore. Fixes: 7dc54d3b ("net: pcs: add Renesas MII converter driver") Signed-off-by:
Xiang Yang <xiangyang3@huawei.com> Reviewed-by:
Vladimir Oltean <olteanv@gmail.com> Signed-off-by:
-
Jason Wang authored
Commit 25266128 ("virtio-net: fix race between set queues and probe") tries to fix the race between set queues and probe by calling _virtnet_set_queues() before DRIVER_OK is set. This violates virtio spec. Fixing this by setting queues after virtio_device_ready(). Note that rtnl needs to be held for userspace requests to change the number of queues. So we are serialized in this way. Fixes: 25266128 ("virtio-net: fix race between set queues and probe") Reported-by:
Dragos Tatulea <dtatulea@nvidia.com> Acked-by:
Michael S. Tsirkin <mst@redhat.com> Signed-off-by:
Jason Wang <jasowang@redhat.com> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipJakub Kicinski authored
Cross merge x86 fixes to fix clang linking errors: ld.lld: error: ./arch/x86/kernel/vmlinux.lds:221: at least one side of the expression must be absolute These will hopefully be downstream by the time we ship the next batch of fixes. * 'x86/bugs' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86: Move gds_ucode_mitigated() declaration to header x86/speculation: Add cpu_show_gds() prototype driver core: cpu: Make cpu_show_not_affected() static x86/srso: Fix build breakage with the LLVM linker Documentation/srso: Document IBPB aspect and fix formatting driver core: cpu: Unify redundant silly stubs Documentation/hw-vuln: Unify filename specification in index Link: https://lore.kernel.org/all/CAHk-=wj_b+FGTnevQSBAtCWuhCk=0oQ_THvthBW2hzqpOTLFmg@mail.gmail.com/Signed-off-by:
-
- 10 Aug, 2023 23 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netLinus Torvalds authored
Pull networking fixes from Jakub Kicinski: "Including fixes from netfilter, wireless and bpf. Still trending up in size but the good news is that the "current" regressions are resolved, AFAIK. We're getting weirdly many fixes for Wake-on-LAN and suspend/resume handling on embedded this week (most not merged yet), not sure why. But those are all for older bugs. Current release - regressions: - tls: set MSG_SPLICE_PAGES consistently when handing encrypted data over to TCP Current release - new code bugs: - eth: mlx5: correct IDs on VFs internal to the device (IPU) Previous releases - regressions: - phy: at803x: fix WoL support / reporting on AR8032 - bonding: fix incorrect deletion of ETH_P_8021AD protocol VID from slaves, leading to BUG_ON() - tun: prevent tun_build_skb() from exceeding the packet size limit - wifi: rtw89: fix 8852AE disconnection caused by RX full flags - eth/PCI: enetc: fix probing after 6fffbc7a ("PCI: Honor firmware's device disabled status"), keep PCI devices around even if they are disabled / not going to be probed to be able to apply quirks on them - eth: prestera: fix handling IPv4 routes with nexthop IDs Previous releases - always broken: - netfilter: re-work garbage collection to avoid races between user-facing API and timeouts - tunnels: fix generating ipv4 PMTU error on non-linear skbs - nexthop: fix infinite nexthop bucket dump when using maximum nexthop ID - wifi: nl80211: fix integer overflow in nl80211_parse_mbssid_elems() Misc: - unix: use consistent error code in SO_PEERPIDFD - ipv6: adjust ndisc_is_useropt() to include PREFIX_INFO, in prep for upcoming IETF RFC" * tag 'net-6.5-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (94 commits) net: hns3: fix strscpy causing content truncation issue net: tls: set MSG_SPLICE_PAGES consistently ibmvnic: Ensure login failure recovery is safe from other resets ibmvnic: Do partial reset on login failure ibmvnic: Handle DMA unmapping of login buffs in release functions ibmvnic: Unmap DMA login rsp buffer on send login fail ibmvnic: Enforce stronger sanity checks on login response net: mana: Fix MANA VF unload when hardware is unresponsive netfilter: nf_tables: remove busy mark and gc batch API netfilter: nft_set_hash: mark set element as dead when deleting from packet path netfilter: nf_tables: adapt set backend to use GC transaction API netfilter: nf_tables: GC transaction API to avoid race with control plane selftests/bpf: Add sockmap test for redirecting partial skb data selftests/bpf: fix a CI failure caused by vsock sockmap test bpf, sockmap: Fix bug that strp_done cannot be called bpf, sockmap: Fix map type error in sock_map_del_link xsk: fix refcount underflow in error path ipv6: adjust ndisc_is_useropt() to also return true for PIO selftests: forwarding: bridge_mdb: Make test more robust selftests: forwarding: bridge_mdb_max: Fix failing test with old libnet ... -
Hao Chen authored
hns3_dbg_fill_content()/hclge_dbg_fill_content() is aim to integrate some items to a string for content, and we add '\n' and '\0' in the last two bytes of content. strscpy() will add '\0' in the last byte of destination buffer(one of items), it result in finishing content print ahead of schedule and some dump content truncation. One Error log shows as below: cat mac_list/uc UC MAC_LIST: Expected: UC MAC_LIST: FUNC_ID MAC_ADDR STATE pf 00:2b:19:05:03:00 ACTIVE The destination buffer is length-bounded and not required to be NUL-terminated, so just change strscpy() to memcpy() to fix it. Fixes: 1cf3d556 ("net: hns3: fix strncpy() not using dest-buf length as length issue") Signed-off-by:
Hao Chen <chenhao418@huawei.com> Signed-off-by:
Jijie Shao <shaojijie@huawei.com> Link: https://lore.kernel.org/r/20230809020902.1941471-1-shaojijie@huawei.comSigned-off-by:
-
Jakub Kicinski authored
We used to change the flags for the last segment, because non-last segments had the MSG_SENDPAGE_NOTLAST flag set. That flag is no longer a thing so remove the setting. Since flags most likely don't have MSG_SPLICE_PAGES set this avoids passing parts of the sg as splice and parts as non-splice. Before commit under Fixes we'd have called tcp_sendpage() which would add the MSG_SPLICE_PAGES. Why this leads to trouble remains unclear but Tariq reports hitting the WARN_ON(!sendpage_ok()) due to page refcount of 0. Fixes: e117dcfd ("tls: Inline do_tcp_sendpages()") Reported-by:
Tariq Toukan <tariqt@nvidia.com> Link: https://lore.kernel.org/all/4c49176f-147a-4283-f1b1-32aac7b4b996@gmail.com/Tested-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengineLinus Torvalds authored
Pull dmaengine fixes from Vinod Koul: - HAS_IOMEM fixes for fsl edma and intel idma - return-value fix, interrupt vector setting and typo fix for xilinx xdma - email updates for codeaurora email domain move - correct pause status for pl330 driver - idxd clear flag on disable fix - function documentation fix for owl dma - potential un-allocated memory fix for mcf driver * tag 'dmaengine-fix-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine: dmaengine: xilinx: xdma: Fix typo dmaengine: xilinx: xdma: Fix interrupt vector setting dmaengine: owl-dma: Modify mismatched function name dmaengine: idxd: Clear PRS disable flag when disabling IDXD device dmaengine: pl330: Return DMA_PAUSED when transaction is paused dmaengine: qcom_hidma: Update codeaurora email domain dmaengine: mcf-edma: Fix a potential un-allocated memory access dmaengine: xilinx: xdma: Fix Judgment of the return value idmaengine: make FSL_EDMA and INTEL_IDMA64 depends on HAS_IOMEM
-
git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nfJakub Kicinski authored
Pablo Neira Ayuso says: ==================== Netfilter fixes for net The existing attempt to resolve races between control plane and GC work is error prone, as reported by Bien Pham <phamnnb@sea.com>, some places forgot to call nft_set_elem_mark_busy(), leading to double-deactivation of elements. This series contains the following patches: 1) Do not skip expired elements during walk otherwise elements might never decrement the reference counter on data, leading to memleak. 2) Add a GC transaction API to replace the former attempt to deal with races between control plane and GC. GC worker sets on NFT_SET_ELEM_DEAD_BIT on elements and it creates a GC transaction to remove the expired elements, GC transaction could abort in case of interference with control plane and retried later (GC async). Set backends such as rbtree and pipapo also perform GC from control plane (GC sync), in such case, element deactivation and removal is safe because mutex is held then collected elements are released via call_rcu(). 3) Adapt existing set backends to use the GC transaction API. 4) Update rhash set backend to set on _DEAD bit to report deleted elements from datapath for GC. 5) Remove old GC batch API and the NFT_SET_ELEM_BUSY_BIT. * tag 'nf-23-08-10' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf: netfilter: nf_tables: remove busy mark and gc batch API netfilter: nft_set_hash: mark set element as dead when deleting from packet path netfilter: nf_tables: adapt set backend to use GC transaction API netfilter: nf_tables: GC transaction API to avoid race with control plane netfilter: nf_tables: don't skip expired elements during walk ==================== Link: https://lore.kernel.org/r/20230810070830.24064-1-pablo@netfilter.orgSigned-off-by:
-
https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpfJakub Kicinski authored
Martin KaFai Lau says: ==================== pull-request: bpf 2023-08-09 We've added 5 non-merge commits during the last 7 day(s) which contain a total of 6 files changed, 102 insertions(+), 8 deletions(-). The main changes are: 1) A bpf sockmap memleak fix and a fix in accessing the programs of a sockmap under the incorrect map type from Xu Kuohai. 2) A refcount underflow fix in xsk from Magnus Karlsson. * tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf: selftests/bpf: Add sockmap test for redirecting partial skb data selftests/bpf: fix a CI failure caused by vsock sockmap test bpf, sockmap: Fix bug that strp_done cannot be called bpf, sockmap: Fix map type error in sock_map_del_link xsk: fix refcount underflow in error path ==================== Link: https://lore.kernel.org/r/20230810055303.120917-1-martin.lau@linux.devSigned-off-by:
-
Nick Child authored
If a login request fails, the recovery process should be protected against parallel resets. It is a known issue that freeing and registering CRQ's in quick succession can result in a failover CRQ from the VIOS. Processing a failover during login recovery is dangerous for two reasons: 1. This will result in two parallel initialization processes, this can cause serious issues during login. 2. It is possible that the failover CRQ is received but never executed. We get notified of a pending failover through a transport event CRQ. The reset is not performed until a INIT CRQ request is received. Previously, if CRQ init fails during login recovery, then the ibmvnic irq is freed and the login process returned error. If failover_pending is true (a transport event was received), then the ibmvnic device would never be able to process the reset since it cannot receive the CRQ_INIT request due to the irq being freed. This leaved the device in a inoperable state. Therefore, the login failure recovery process must be hardened against these possible issues. Possible failovers (due to quick CRQ free and init) must be avoided and any issues during re-initialization should be dealt with instead of being propagated up the stack. This logic is similar to that of ibmvnic_probe(). Fixes: dff515a3 ("ibmvnic: Harden device login requests") Signed-off-by:
Nick Child <nnac123@linux.ibm.com> Reviewed-by:
-
Nick Child authored
Perform a partial reset before sending a login request if any of the following are true: 1. If a previous request times out. This can be dangerous because the VIOS could still receive the old login request at any point after the timeout. Therefore, it is best to re-register the CRQ's and sub-CRQ's before retrying. 2. If the previous request returns an error that is not described in PAPR. PAPR provides procedures if the login returns with partial success or aborted return codes (section L.5.1) but other values do not have a defined procedure. Previously, these conditions just returned error from the login function rather than trying to resolve the issue. This can cause further issues since most callers of the login function are not prepared to handle an error when logging in. This improper cleanup can lead to the device being permanently DOWN'd. For example, if the VIOS believes that the device is already logged in then it will return INVALID_STATE (-7). If we never re-register CRQ's then it will always think that the device is already logged in. This leaves the device inoperable. The partial reset involves freeing the sub-CRQs, freeing the CRQ then registering and initializing a new CRQ and sub-CRQs. This essentially restarts all communication with VIOS to allow for a fresh login attempt that will be unhindered by any previous failed attempts. Fixes: dff515a3 ("ibmvnic: Harden device login requests") Signed-off-by:
-
Nick Child authored
Rather than leaving the DMA unmapping of the login buffers to the login response handler, move this work into the login release functions. Previously, these functions were only used for freeing the allocated buffers. This could lead to issues if there are more than one outstanding login buffer requests, which is possible if a login request times out. If a login request times out, then there is another call to send login. The send login function makes a call to the login buffer release function. In the past, this freed the buffers but did not DMA unmap. Therefore, the VIOS could still write to the old login (now freed) buffer. It is for this reason that it is a good idea to leave the DMA unmap call to the login buffers release function. Since the login buffer release functions now handle DMA unmapping, remove the duplicate DMA unmapping in handle_login_rsp(). Fixes: dff515a3 ("ibmvnic: Harden device login requests") Signed-off-by:
-
Nick Child authored
If the LOGIN CRQ fails to send then we must DMA unmap the response buffer. Previously, if the CRQ failed then the memory was freed without DMA unmapping. Fixes: c98d9cc4 ("ibmvnic: send_login should check for crq errors") Signed-off-by:
-
Nick Child authored
Ensure that all offsets in a login response buffer are within the size of the allocated response buffer. Any offsets or lengths that surpass the allocation are likely the result of an incomplete response buffer. In these cases, a full reset is necessary. When attempting to login, the ibmvnic device will allocate a response buffer and pass a reference to the VIOS. The VIOS will then send the ibmvnic device a LOGIN_RSP CRQ to signal that the buffer has been filled with data. If the ibmvnic device does not get a response in 20 seconds, the old buffer is freed and a new login request is sent. With 2 outstanding requests, any LOGIN_RSP CRQ's could be for the older login request. If this is the case then the login response buffer (which is for the newer login request) could be incomplete and contain invalid data. Therefore, we must enforce strict sanity checks on the response buffer values. Testing has shown that the `off_rxadd_buff_size` value is filled in last by the VIOS and will be the smoking gun for these circumstances. Until VIOS can implement a mechanism for tracking outstanding response buffers and a method for mapping a LOGIN_RSP CRQ to a particular login response buffer, the best ibmvnic can do in this situation is perform a full reset. Fixes: dff515a3 ("ibmvnic: Harden device login requests") Signed-off-by:
-
Souradeep Chakrabarti authored
When unloading the MANA driver, mana_dealloc_queues() waits for the MANA hardware to complete any inflight packets and set the pending send count to zero. But if the hardware has failed, mana_dealloc_queues() could wait forever. Fix this by adding a timeout to the wait. Set the timeout to 120 seconds, which is a somewhat arbitrary value that is more than long enough for functional hardware to complete any sends. Cc: stable@vger.kernel.org Fixes: ca9c54d2 ("net: mana: Add a driver for Microsoft Azure Network Adapter (MANA)") Signed-off-by:
Souradeep Chakrabarti <schakrabarti@linux.microsoft.com> Link: https://lore.kernel.org/r/1691576525-24271-1-git-send-email-schakrabarti@linux.microsoft.comSigned-off-by:
-
Arnd Bergmann authored
The declaration got placed in the .c file of the caller, but that causes a warning for the definition: arch/x86/kernel/cpu/bugs.c:682:6: error: no previous prototype for 'gds_ucode_mitigated' [-Werror=missing-prototypes] Move it to a header where both sides can observe it instead. Fixes: 81ac7e5d ("KVM: Add GDS_NO support to KVM") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
Dave Hansen <dave.hansen@linux.intel.com> Tested-by:
Daniel Sneddon <daniel.sneddon@linux.intel.com> Cc: stable@kernel.org Link: https://lore.kernel.org/all/20230809130530.1913368-2-arnd%40kernel.org
-
Arnd Bergmann authored
The newly added function has two definitions but no prototypes: drivers/base/cpu.c:605:16: error: no previous prototype for 'cpu_show_gds' [-Werror=missing-prototypes] Add a declaration next to the other ones for this file to avoid the warning. Fixes: 8974eb58 ("x86/speculation: Add Gather Data Sampling mitigation") Signed-off-by:
-
Borislav Petkov (AMD) authored
Fix a -Wmissing-prototypes warning and add the gather_data_sampling() stub macro call for real. Fixes: 0fddfe33 ("driver core: cpu: Unify redundant silly stubs") Closes: https://lore.kernel.org/oe-kbuild-all/202308101956.oRj1ls7s-lkp@intel.comReported-by:
kernel test robot <lkp@intel.com> Signed-off-by:
Borislav Petkov (AMD) <bp@alien8.de> Link: https://lore.kernel.org/r/202308101956.oRj1ls7s-lkp@intel.com
-
Nick Desaulniers authored
The assertion added to verify the difference in bits set of the addresses of srso_untrain_ret_alias() and srso_safe_ret_alias() would fail to link in LLVM's ld.lld linker with the following error: ld.lld: error: ./arch/x86/kernel/vmlinux.lds:210: at least one side of the expression must be absolute ld.lld: error: ./arch/x86/kernel/vmlinux.lds:211: at least one side of the expression must be absolute Use ABSOLUTE to evaluate the expression referring to at least one of the symbols so that LLD can evaluate the linker script. Also, add linker version info to the comment about XOR being unsupported in either ld.bfd or ld.lld until somewhat recently. Fixes: fb3bd914 ("x86/srso: Add a Speculative RAS Overflow mitigation") Closes: https://lore.kernel.org/llvm/CA+G9fYsdUeNu-gwbs0+T6XHi4hYYk=Y9725-wFhZ7gJMspLDRA@mail.gmail.com/Reported-by:
Nathan Chancellor <nathan@kernel.org> Reported-by:
Daniel Kolesa <daniel@octaforge.org> Reported-by:
Naresh Kamboju <naresh.kamboju@linaro.org> Suggested-by:
Sven Volkinsfeld <thyrc@gmx.net> Signed-off-by:
Nick Desaulniers <ndesaulniers@google.com> Signed-off-by:
-
Borislav Petkov (AMD) authored
Add a note about the dependency of the User->User mitigation on the previous Spectre v2 IBPB selection. Make the layout moar pretty. Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Link: https://lore.kernel.org/r/20230809102700.29449-4-bp@alien8.de
-
Borislav Petkov (AMD) authored
Make them all a weak function, aliasing to a single function which issues the "Not affected" string. No functional changes. Suggested-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
Nikolay Borisov <nik.borisov@suse.com> Link: https://lore.kernel.org/r/20230809102700.29449-3-bp@alien8.de
-
Borislav Petkov (AMD) authored
Most of the index.rst files in Documentation/ refer to other rst files without their file extension in the name. Do that here too. No functional changes. Reported-by:
-
Pablo Neira Ayuso authored
Ditch it, it has been replace it by the GC transaction API and it has no clients anymore. Signed-off-by:Pablo Neira Ayuso <pablo@netfilter.org>
-
Pablo Neira Ayuso authored
Set on the NFT_SET_ELEM_DEAD_BIT flag on this element, instead of performing element removal which might race with an ongoing transaction. Enable gc when dynamic flag is set on since dynset deletion requires garbage collection after this patch. Fixes: d0a8d877 ("netfilter: nft_dynset: support for element deletion") Signed-off-by:
-
Pablo Neira Ayuso authored
Use the GC transaction API to replace the old and buggy gc API and the busy mark approach. No set elements are removed from async garbage collection anymore, instead the _DEAD bit is set on so the set element is not visible from lookup path anymore. Async GC enqueues transaction work that might be aborted and retried later. rbtree and pipapo set backends does not set on the _DEAD bit from the sync GC path since this runs in control plane path where mutex is held. In this case, set elements are deactivated, removed and then released via RCU callback, sync GC never fails. Fixes: 3c4287f6 ("nf_tables: Add set type for arbitrary concatenation of ranges") Fixes: 8d8540c4 ("netfilter: nft_set_rbtree: add timeout support") Fixes: 9d098292 ("netfilter: nft_hash: add support for timeouts") Signed-off-by:
-
Pablo Neira Ayuso authored
The set types rhashtable and rbtree use a GC worker to reclaim memory. From system work queue, in periodic intervals, a scan of the table is done. The major caveat here is that the nft transaction mutex is not held. This causes a race between control plane and GC when they attempt to delete the same element. We cannot grab the netlink mutex from the work queue, because the control plane has to wait for the GC work queue in case the set is to be removed, so we get following deadlock: cpu 1 cpu2 GC work transaction comes in , lock nft mutex `acquire nft mutex // BLOCKS transaction asks to remove the set set destruction calls cancel_work_sync() cancel_work_sync will now block forever, because it is waiting for the mutex the caller already owns. This patch adds a new API that deals with garbage collection in two steps: 1) Lockless GC of expired elements sets on the NFT_SET_ELEM_DEAD_BIT so they are not visible via lookup. Annotate current GC sequence in the GC transaction. Enqueue GC transaction work as soon as it is full. If ruleset is updated, then GC transaction is aborted and retried later. 2) GC work grabs the mutex. If GC sequence has changed then this GC transaction lost race with control plane, abort it as it contains stale references to objects and let GC try again later. If the ruleset is intact, then this GC transaction deactivates and removes the elements and it uses call_rcu() to destroy elements. Note that no elements are removed from GC lockless path, the _DEAD bit is set and pointers are collected. GC catchall does not remove the elements anymore too. There is a new set->dead flag that is set on to abort the GC transaction to deal with set->ops->destroy() path which removes the remaining elements in the set from commit_release, where no mutex is held. To deal with GC when mutex is held, which allows safe deactivate and removal, add sync GC API which releases the set element object via call_rcu(). This is used by rbtree and pipapo backends which also perform garbage collection from control plane path. Since element removal from sets can happen from control plane and element garbage collection/timeout, it is necessary to keep the set structure alive until all elements have been deactivated and destroyed. We cannot do a cancel_work_sync or flush_work in nft_set_destroy because its called with the transaction mutex held, but the aforementioned async work queue might be blocked on the very mutex that nft_set_destroy() callchain is sitting on. This gives us the choice of ABBA deadlock or UaF. To avoid both, add set->refs refcount_t member. The GC API can then increment the set refcount and release it once the elements have been free'd. Set backends are adapted to use the GC transaction API in a follow up patch entitled: ("netfilter: nf_tables: use gc transaction API in set backends") This is joint work with Florian Westphal. Fixes: cfed7e1b ("netfilter: nf_tables: add set garbage collection helpers") Signed-off-by:
-
