- 21 Jul, 2020 11 commits
-
-
Shannon Nelson authored
Use an offset to write the second half of the regs data into the second half of the buffer instead of overwriting the first half. Fixes: 4d03e00a ("ionic: Add initial ethtool support") Signed-off-by:
Shannon Nelson <snelson@pensando.io> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Murali Karicheri authored
skb_put_padto() can fail. So check for return type and return NULL for skb. Caller checks for skb and acts correctly if it is NULL. Fixes: 6d6148bc ("net: hsr: fix incorrect lsdu size in the tag of HSR frames for small frames") Signed-off-by:
Murali Karicheri <m-karicheri2@ti.com> Signed-off-by:
-
Guillaume Nault authored
bareudp.rst was written before iproute2 gained support for this new type of tunnel. Therefore, the sample command lines didn't match the final iproute2 implementation. Signed-off-by:
Guillaume Nault <gnault@redhat.com> Signed-off-by:
-
Liu Jian authored
When mlxsw_core_trap_register fails in mlxsw_emad_init, destroy_workqueue() shouled be called to destroy mlxsw_core->emad_wq. Fixes: d965465b ("mlxsw: core: Fix possible deadlock") Signed-off-by:
Liu Jian <liujian56@huawei.com> Reviewed-by:
Ido Schimmel <idosch@mellanox.com> Signed-off-by:
-
Liu Jian authored
When dma_coerce_mask_and_coherent() fails, the alloced netdev need to be freed. Fixes: 060ad66f ("dpaa_eth: change DMA device") Signed-off-by:
Madalin Bucur <madalin.bucur@oss.nxp.com> Signed-off-by:
-
David S. Miller authored
Karsten Graul says: ==================== net/smc: fixes 2020-07-20 Please apply the following patch series for smc to netdev's net tree. Patch 1 fixes a problem with a buffer that is not put back when the connection was killed in the meantime. Patch 2 fixes a wrong behaviour when the maximum dmb buffer count exceeded. ==================== Signed-off-by: -
Karsten Graul authored
There is a current limit of 1920 registered dmb buffers per ISM device for smc-d. One link group can contain 255 connections, each connection is using one dmb buffer. When the connection is closed then the registered buffer is held in a queue and is reused by the next connection. When a link group is 'full' then another link group is created and uses an own buffer pool. The link groups are added to a list using list_add() which puts a new link group to the first position in the list. In the situation that many connections are opened (>1920) and a few of them stay open while others are closed quickly we end up with at least 8 link groups. For a new connection a matching link group is looked up, iterating over the list of link groups. The trailing 7 link groups all have registered dmb buffers which could be reused, while the first link group has only a few dmb buffers and then hit the 1920 limit. Because the first link group is not full (255 connection limit not reached) it is chosen and finally the connection falls back to TCP because there is no dmb buffer available in this link group. There are multiple ways to fix that: using list_add_tail() allows to scan older link groups first for free buffers which ensures that buffers are reused first. This fixes the problem for smc-r link groups as well. For smc-d there is an even better way to address this problem because smc-d does not have the 255 connections per link group limit. So fix the problem for smc-d by allowing large link groups. Fixes: c6ba7c9b ("net/smc: add base infrastructure for SMC-D and ISM") Reviewed-by:
Ursula Braun <ubraun@linux.ibm.com> Signed-off-by:
Karsten Graul <kgraul@linux.ibm.com> Signed-off-by:
-
Karsten Graul authored
To get a send slot smc_wr_tx_get_free_slot() is called, which might wait for a free slot. When smc_wr_tx_get_free_slot() returns there is a check if the connection was killed in the meantime. In that case don't only return an error, but also put back the free slot. Fixes: b2900980 ("net/smc: cancel send and receive for terminated socket") Reviewed-by:
-
David Howells authored
rxrpc_sendmsg() returns EPIPE if there's an outstanding error, such as if rxrpc_recvmsg() indicating ENODATA if there's nothing for it to read. Change rxrpc_recvmsg() to return EAGAIN instead if there's nothing to read as this particular error doesn't get stored in ->sk_err by the networking core. Also change rxrpc_sendmsg() so that it doesn't fail with delayed receive errors (there's no way for it to report which call, if any, the error was caused by). Fixes: 17926a79 ("[AF_RXRPC]: Provide secure RxRPC sockets for use by userspace and kernel both") Signed-off-by:
David Howells <dhowells@redhat.com> Signed-off-by:
-
David S. Miller authored
Merge tag 'ieee802154-for-davem-2020-07-20' of git://git.kernel.org/pub/scm/linux/kernel/git/sschmidt/wpan Stefan Schmidt says: ==================== pull-request: ieee802154 for net 2020-07-20 An update from ieee802154 for your *net* tree. A potential memory leak fix for adf7242 from Liu Jian, and one more HTTPS link change from Alexander A. Klimov. ==================== Signed-off-by:
-
Zhang Changzhong authored
The driver forgets to call clk_disable_unprepare() in error path after a success calling for clk_prepare_enable(). Fix to goto err_clk_disable if clk_prepare_enable() is successful. Fixes: c80d36ff ("net: bcmgenet: Use devm_clk_get_optional() to get the clocks") Signed-off-by:
Zhang Changzhong <zhangchangzhong@huawei.com> Acked-by:
Doug Berger <opendmb@gmail.com> Acked-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
-
- 20 Jul, 2020 5 commits
-
-
Zhang Changzhong authored
The driver forgets to call clk_disable_unprepare() in error path after a success calling for clk_prepare_enable(). Fix to goto err_clk_disable if clk_prepare_enable() is successful. Fixes: 99d55638 ("net: bcmgenet: enable NETIF_F_HIGHDMA flag") Signed-off-by:
-
Alexander A. Klimov authored
Rationale: Reduces attack surface on kernel devs opening the links for MITM as HTTPS traffic is much harder to manipulate. Deterministic algorithm: For each file: If not .svg: For each line: If doesn't contain `\bxmlns\b`: For each link, `\bhttp://[^# \t\r\n]*(?:\w|/)`: If neither `\bgnu\.org/license`, nor `\bmozilla\.org/MPL\b`: If both the HTTP and HTTPS versions return 200 OK and serve the same content: Replace HTTP with HTTPS. Signed-off-by:Alexander A. Klimov <grandmaster@al2klimov.de> Link: https://lore.kernel.org/r/20200719113142.58304-1-grandmaster@al2klimov.deSigned-off-by:
Stefan Schmidt <stefan@datenfreihafen.org>
-
Taehee Yoo authored
If register_netdevice() is failed, net_device should not be used because variables are uninitialized or freed. So, the routine should be stopped immediately. But, bond_create() doesn't check return value of register_netdevice() immediately. That will result in a panic because of using uninitialized or freed memory. Test commands: modprobe netdev-notifier-error-inject echo -22 > /sys/kernel/debug/notifier-error-inject/netdev/\ actions/NETDEV_REGISTER/error modprobe bonding max_bonds=3 Splat looks like: [ 375.028492][ T193] general protection fault, probably for non-canonical address 0x6b6b6b6b6b6b6b6b: 0000 [#1] SMP DEBUG_PAGEALLOC PTI [ 375.033207][ T193] CPU: 2 PID: 193 Comm: kworker/2:2 Not tainted 5.8.0-rc4+ #645 [ 375.036068][ T193] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1 04/01/2014 [ 375.039673][ T193] Workqueue: events linkwatch_event [ 375.041557][ T193] RIP: 0010:dev_activate+0x4a/0x340 [ 375.043381][ T193] Code: 40 a8 04 0f 85 db 00 00 00 8b 83 08 04 00 00 85 c0 0f 84 0d 01 00 00 31 d2 89 d0 48 8d 04 40 48 c1 e0 07 48 03 83 00 04 00 00 <48> 8b 48 10 f6 41 10 01 75 08 f0 80 a1 a0 01 00 00 fd 48 89 48 08 [ 375.050267][ T193] RSP: 0018:ffff9f8facfcfdd8 EFLAGS: 00010202 [ 375.052410][ T193] RAX: 6b6b6b6b6b6b6b6b RBX: ffff9f8fae6ea000 RCX: 0000000000000006 [ 375.055178][ T193] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff9f8fae6ea000 [ 375.057762][ T193] RBP: ffff9f8fae6ea000 R08: 0000000000000000 R09: 0000000000000000 [ 375.059810][ T193] R10: 0000000000000001 R11: 0000000000000000 R12: ffff9f8facfcfe08 [ 375.061892][ T193] R13: ffffffff883587e0 R14: 0000000000000000 R15: ffff9f8fae6ea580 [ 375.063931][ T193] FS: 0000000000000000(0000) GS:ffff9f8fbae00000(0000) knlGS:0000000000000000 [ 375.066239][ T193] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 375.067841][ T193] CR2: 00007f2f542167a0 CR3: 000000012cee6002 CR4: 00000000003606e0 [ 375.069657][ T193] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 375.071471][ T193] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 375.073269][ T193] Call Trace: [ 375.074005][ T193] linkwatch_do_dev+0x4d/0x50 [ 375.075052][ T193] __linkwatch_run_queue+0x10b/0x200 [ 375.076244][ T193] linkwatch_event+0x21/0x30 [ 375.077274][ T193] process_one_work+0x252/0x600 [ 375.078379][ T193] ? process_one_work+0x600/0x600 [ 375.079518][ T193] worker_thread+0x3c/0x380 [ 375.080534][ T193] ? process_one_work+0x600/0x600 [ 375.081668][ T193] kthread+0x139/0x150 [ 375.082567][ T193] ? kthread_park+0x90/0x90 [ 375.083567][ T193] ret_from_fork+0x22/0x30 Fixes: e826eafa ("bonding: Call netif_carrier_off after register_netdevice") Signed-off-by:Taehee Yoo <ap420073@gmail.com> Signed-off-by:
-
Russell King authored
The commit below caused a regression for clearfog-gt-8k, where the link between the switch and the host does not come up. Investigation revealed two issues: - MV88E6xxx DSA no longer allows an in-band link to come up as the link is programmed to be forced down. Commit "net: dsa: mv88e6xxx: fix in-band AN link establishment" addresses this. - The dts configured dissimilar link modes at each end of the host to switch link; the host was configured using a fixed link (so has no in-band status) and the switch was configured to expect in-band status. With both issues fixed, the regression is resolved. Fixes: 34b5e6a3 ("net: dsa: mv88e6xxx: Configure MAC when using fixed link") Reported-by:
Martin Rowe <martin.p.rowe@gmail.com> Signed-off-by:
Russell King <rmk+kernel@armlinux.org.uk> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
Russell King authored
If in-band negotiation or fixed-link modes are specified for a DSA port, the DSA code will force the link down during initialisation. For fixed-link mode, this is fine, as phylink will manage the link state. However, for in-band mode, phylink expects the PCS to detect link, which will not happen if the link is forced down. There is a related issue that in in-band mode, the link could come up while we are making configuration changes, so we should force the link down prior to reconfiguring the interface mode. This patch addresses both issues. Fixes: 3be98b2d ("net: dsa: Down cpu/dsa ports phylink will control") Signed-off-by:
-
- 19 Jul, 2020 11 commits
-
-
David S. Miller authored
Karsten Graul says: ==================== net/smc: fixes 2020-07-16 Please apply the following patch series for smc to netdev's net tree. The patches address problems caused by late or unexpected link layer control packets, dma sync calls for unmapped memory, freed buffers that are not removed from the buffer list and a possible null pointer access that results in a crash. v1->v2: in patch 4, improve patch description and correct the comment for the new mutex ==================== Signed-off-by: -
Karsten Graul authored
When a listen socket is closed then all non-accepted sockets in its accept queue are to be released. Inside __smc_release() the helper smc_restore_fallback_changes() restores the changes done to the socket without to check if the clcsocket has a file set. This can result in a crash. Fix this by checking the file pointer first. Reviewed-by:
-
Karsten Graul authored
Two buffers are allocated for each SMC connection. Each buffer is added to a buffer list after creation. When the second buffer allocation fails, the first buffer is freed but not deleted from the list. This might result in crashes when another connection picks up the freed buffer later and starts to work with it. Reviewed-by:
-
Karsten Graul authored
The dma related ...sync_sg... functions check the link state before the dma function is actually called. But the check in smc_link_usable() allows links in ACTIVATING state which are not yet mapped to dma memory. Under high load it may happen that the sync_sg functions are called for such a link which results in an debug output like DMA-API: mlx5_core 0002:00:00.0: device driver tries to sync DMA memory it has not allocated [device address=0x0000000103370000] [size=65536 bytes] To fix that introduce a helper to check for the link state ACTIVE and use it where appropriate. And move the link state update to ACTIVATING to the end of smcr_link_init() when most initial setup is done. Reviewed-by:
-
Karsten Graul authored
As smc client the delete link requests are assigned to the flow when _any_ flow is active. This may break other flows that do not expect delete link requests during their handling. Fix that by assigning the request only when an add link flow is active. With that fix the code for smc client and smc server is the same, so remove the separate handling. Reviewed-by:
-
Karsten Graul authored
When a new ib device is up smc will send an add link invitation to the peer if needed. This is currently done with rudimentary flow control. Under high workload these add link invitations can disturb other llc flows because they arrive unexpected. Fix this by integrating the invitations into the normal llc event flow and handle them as a flow. While at it, check for already assigned requests in the flow before the new add link request is assigned. Reviewed-by:
-
Karsten Graul authored
To be save from unexpected or late llc response messages check if the arrived message fits to the current flow type and drop out-of-flow messages. And drop it when there is already a response assigned to the flow. Reviewed-by:
-
Karsten Graul authored
Before an smc ib device is used the first time for an smc link it is lazily initialized. When there are 2 active link groups and a new ib device is brought online then it might happen that 2 link creations run in parallel and enter smc_ib_setup_per_ibdev(). Both allocate new send and receive completion queues on the device, but only one set of them keeps assigned and the other leaks. Fix that by protecting the setup and cleanup code using a mutex. Reviewed-by:
-
Karsten Graul authored
For new rdma connections the SMC server assigns the link and sends the link data in the clc accept message. To match the correct link use not only the qp_num but also the gid and the mac of the links. If there are equal qp_nums for different links the wrong link would be chosen. Reviewed-by:
-
Karsten Graul authored
In a link-down condition we notify the SMC server and expect that the server will finally trigger the link clear processing on the client side. This could fail when anything along this notification path goes wrong. Clear the link as part of SMC client link-down processing to prevent dangling links. Reviewed-by:
-
Karsten Graul authored
A delete link could arrive during confirm link processing. Handle this situation directly in smc_llc_srv_conf_link() rather than using the logic in smc_llc_wait() to avoid the unexpected message handling there. Reviewed-by:
-
- 18 Jul, 2020 6 commits
-
-
Vadim Pasternak authored
Fix wrong reading of upper pages for SFP EEPROM. According to "Memory Organization" figure in SFF-8472 spec: When reading upper pages 1, 2 and 3 the offset should be set relative to zero and I2C high address 0x51 [1010001X (A2h)] is to be used. Fixes: a45bfb5a ("mlxsw: core: Extend QSFP EEPROM size for ethtool") Signed-off-by:
Vadim Pasternak <vadimp@mellanox.com> Reviewed-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
-
Murali Karicheri authored
Validate MAC address before copying the same to outgoing frame skb destination address. Since a node can have zero mac address for Link B until a valid frame is received over that link, this fix address the issue of a zero MAC address being in the packet. Signed-off-by:
-
Murali Karicheri authored
For small Ethernet frames with size less than minimum size 66 for HSR vs 60 for regular Ethernet frames, hsr driver currently doesn't pad the frame to make it minimum size. This results in incorrect LSDU size being populated in the HSR tag for these frames. Fix this by padding the frame to the minimum size applicable for HSR. Signed-off-by:
-
Randy Dunlap authored
Drop the doubled word "be" in a comment. Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Cc: Thomas Graf <tgraf@suug.ch> Cc: Herbert Xu <herbert@gondor.apana.org.au> Cc: netdev@vger.kernel.org Signed-off-by:
-
Alexandre Belloni authored
There is one RGMII check not using the phy_interface_mode_is_rgmii() helper. This prevents the driver from configuring the MAC properly when using a phy-mode that is not just rgmii, e.g. rgmii-rxid. This became an issue on sama5d3 xplained since the ksz9031 driver is hadling phy-mode properly and the phy-mode has to be set to rgmii-rxid. Fixes: bcf3440c ("net: phy: micrel: add phy-mode support for the KSZ9031 PHY") Signed-off-by:
Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by:
-
Nikita Danilov authored
This patch disables PTP on AQC111 and AQC112 due to a known HW issue, which can cause datapath issues. Ideally PTP block should have been disabled via PHY provisioning, but unfortunately many units have been shipped with enabled PTP block. Thus, we have to work around this in the driver. Fixes: dbcd6806 ("net: aquantia: add support for Phy access") Signed-off-by:
Nikita Danilov <ndanilov@marvell.com> Signed-off-by:
Mark Starovoytov <mstarovoitov@marvell.com> Signed-off-by:
Igor Russkikh <irusskikh@marvell.com> Signed-off-by:
-
- 17 Jul, 2020 7 commits
-
-
Wang Hai authored
When nfc_register_device fails in nci_register_device, destroy_workqueue() shouled be called to destroy ndev->tx_wq. Fixes: 3c1c0f5d ("NFC: NCI: Fix nci_register_device init sequence") Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
Wang Hai <wanghai38@huawei.com> Signed-off-by:
-
David S. Miller authored
Doug Berger says: ==================== net: bcmgenet: fix WAKE_FILTER resume from deep sleep The WAKE_FILTER logic can only wake the system from the standby power state. However, some systems that include the GENET IP support deeper power saving states and the driver should suspend and resume correctly from those states as well. This commit set squashes a few issues uncovered while testing suspend and resume from these deep sleep states. ==================== Signed-off-by: -
Doug Berger authored
The Hardware Filter Block RAM may not be preserved when the GENET block is reset during a deep sleep, so it is not sufficient to only backup and restore the enables. This commit clears out the HFB block and reprograms the rxnfc rules when the system resumes from a suspended state. To support this the bcmgenet_hfb_create_rxnfc_filter() function is modified to access the register space directly so that it can't fail due to memory allocation issues. Fixes: f50932cc ("net: bcmgenet: add WAKE_FILTER support") Signed-off-by:
-
Doug Berger authored
When the GENET driver resumes from deep sleep the UMAC_CMD register may not be accessible and therefore should not be accessed from bcmgenet_wol_power_up_cfg() if the GENET has been reset. This commit adds a check of the RBUF_ACPI_EN flag when Wake on Filter is enabled. A clear flag indicates that the GENET hardware must have been reset so the remainder of the hardware programming is bypassed. Fixes: f50932cc ("net: bcmgenet: add WAKE_FILTER support") Signed-off-by:
-
Doug Berger authored
When the GENET driver resumes from deep sleep the UMAC_CMD register may not be accessible and therefore should not be accessed from bcmgenet_wol_power_up_cfg() if the GENET has been reset. This commit adds a check of the MPD_EN flag when Wake on Magic Packet is enabled. A clear flag indicates that the GENET hardware must have been reset so the remainder of the hardware programming is bypassed. Fixes: 1a1d5106 ("net: bcmgenet: move clk_wol management to bcmgenet_wol") Signed-off-by:
-
Paolo Pisati authored
Fix ip_defrag.sh when CONFIG_NF_DEFRAG_IPV6=m: $ sudo ./ip_defrag.sh + set -e + mktemp -u XXXXXX + readonly NETNS=ns-rGlXcw + trap cleanup EXIT + setup + ip netns add ns-rGlXcw + ip -netns ns-rGlXcw link set lo up + ip netns exec ns-rGlXcw sysctl -w net.ipv4.ipfrag_high_thresh=9000000 + ip netns exec ns-rGlXcw sysctl -w net.ipv4.ipfrag_low_thresh=7000000 + ip netns exec ns-rGlXcw sysctl -w net.ipv4.ipfrag_time=1 + ip netns exec ns-rGlXcw sysctl -w net.ipv6.ip6frag_high_thresh=9000000 + ip netns exec ns-rGlXcw sysctl -w net.ipv6.ip6frag_low_thresh=7000000 + ip netns exec ns-rGlXcw sysctl -w net.ipv6.ip6frag_time=1 + ip netns exec ns-rGlXcw sysctl -w net.netfilter.nf_conntrack_frag6_high_thresh=9000000 + cleanup + ip netns del ns-rGlXcw $ ls -la /proc/sys/net/netfilter/nf_conntrack_frag6_high_thresh ls: cannot access '/proc/sys/net/netfilter/nf_conntrack_frag6_high_thresh': No such file or directory $ sudo modprobe nf_defrag_ipv6 $ ls -la /proc/sys/net/netfilter/nf_conntrack_frag6_high_thresh -rw-r--r-- 1 root root 0 Jul 14 12:34 /proc/sys/net/netfilter/nf_conntrack_frag6_high_thresh Signed-off-by:
Paolo Pisati <paolo.pisati@canonical.com> Reviewed-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Wang Hai authored
If try_toggle_control_gpio() failed in smc_drv_probe(), free_netdev(ndev) should be called to free the ndev created earlier. Otherwise, a memleak will occur. Fixes: 7d2911c4 ("net: smc91x: Fix gpios for device tree based booting") Reported-by:
-
