CVSS: -EPSS: 0%CPEs: 4EXPL: 0CVE-2022-48648 – sfc: fix null pointer dereference in efx_hard_start_xmit
https://notcve.org/view.php?id=CVE-2022-48648
In the Linux kernel, the following vulnerability has been resolved: sfc: fix null pointer dereference in efx_hard_start_xmit Trying to get the channel from the tx_queue variable here is wrong because we can only be here if tx_queue is NULL, so we shouldn't dereference it. As the above comment in the code says, this is very unlikely to happen, but it's wrong anyway so let's fix it. I hit this issue because of a different bug that caused tx_queue to be NULL. If that happens, this is the error message that we get here: BUG: unable to handle kernel NULL pointer dereference at 0000000000000020 [...] RIP: 0010:efx_hard_start_xmit+0x153/0x170 [sfc] En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: sfc: corrige la desreferencia del puntero nulo en efx_hard_start_xmit Intentar obtener el canal de la variable tx_queue aquí es incorrecto porque solo podemos estar aquí si tx_queue es NULL, por lo que no debemos desreferenciarlo. Como dice el comentario anterior en el código, es muy poco probable que esto suceda, pero de todos modos está mal, así que solucionémoslo. Encontré este problema debido a un error diferente que provocó que tx_queue fuera NULL. • https://git.kernel.org/stable/c/12804793b17c0e19115a90d98f2f3df0cb79e233 https://git.kernel.org/stable/c/b3b41d4d95d3822b2e459ecbc80d030ea6aec5e7 https://git.kernel.org/stable/c/8547c7bfc0617e7184e4da65b9b96681fcfe9998 https://git.kernel.org/stable/c/b3b952168ee1f220ba729fa100fd9d5aa752eb03 https://git.kernel.org/stable/c/0a242eb2913a4aa3d6fbdb86559f27628e9466f3 •
CVSS: -EPSS: 0%CPEs: 6EXPL: 0CVE-2022-48647 – sfc: fix TX channel offset when using legacy interrupts
https://notcve.org/view.php?id=CVE-2022-48647
In the Linux kernel, the following vulnerability has been resolved: sfc: fix TX channel offset when using legacy interrupts In legacy interrupt mode the tx_channel_offset was hardcoded to 1, but that's not correct if efx_sepparate_tx_channels is false. In that case, the offset is 0 because the tx queues are in the single existing channel at index 0, together with the rx queue. Without this fix, as soon as you try to send any traffic, it tries to get the tx queues from an uninitialized channel getting these errors: WARNING: CPU: 1 PID: 0 at drivers/net/ethernet/sfc/tx.c:540 efx_hard_start_xmit+0x12e/0x170 [sfc] [...] RIP: 0010:efx_hard_start_xmit+0x12e/0x170 [sfc] [...] Call Trace: <IRQ> dev_hard_start_xmit+0xd7/0x230 sch_direct_xmit+0x9f/0x360 __dev_queue_xmit+0x890/0xa40 [...] BUG: unable to handle kernel NULL pointer dereference at 0000000000000020 [...] RIP: 0010:efx_hard_start_xmit+0x153/0x170 [sfc] [...] Call Trace: <IRQ> dev_hard_start_xmit+0xd7/0x230 sch_direct_xmit+0x9f/0x360 __dev_queue_xmit+0x890/0xa40 [...] En el kernel de Linux, se resolvió la siguiente vulnerabilidad: sfc: corrige el desplazamiento del canal TX cuando se usan interrupciones heredadas En el modo de interrupción heredado, tx_channel_offset estaba codificado en 1, pero eso no es correcto si efx_sepparate_tx_channels es falso. En ese caso, el desplazamiento es 0 porque las colas de transmisión están en el único canal existente en el índice 0, junto con la cola de recepción. Sin esta solución, tan pronto como intenta enviar tráfico, intenta obtener las colas de transmisión de un canal no inicializado y obtiene estos errores: ADVERTENCIA: CPU: 1 PID: 0 en drivers/net/ethernet/sfc/tx.c: 540 efx_hard_start_xmit+0x12e/0x170 [sfc] [...] • https://git.kernel.org/stable/c/bf2af9b24313553f3f0b30443220ab0ac8595d2d https://git.kernel.org/stable/c/06cb7e134f8f4a11b66f1dbeb5f237412a0aeedc https://git.kernel.org/stable/c/c308dfd1b43ef0d4c3e57b741bb3462eb7a7f4a2 https://git.kernel.org/stable/c/1a7c0b3ef93d1d1dae7b0a85dab3ac6e3ad0ef61 https://git.kernel.org/stable/c/b2d60329a0b88c4e35017436ee29c43be59d46a5 https://git.kernel.org/stable/c/b4afd3878f961d3517f27b3213730fceef77945c https://git.kernel.org/stable/c/360910b88d1466a90644a4e0533803b594344a2b https://git.kernel.org/stable/c/5f623a77cfc2d501d72bcb4f9ee71721e •
CVSS: 6.2EPSS: 0%CPEs: 2EXPL: 0CVE-2022-48646 – sfc/siena: fix null pointer dereference in efx_hard_start_xmit
https://notcve.org/view.php?id=CVE-2022-48646
In the Linux kernel, the following vulnerability has been resolved: sfc/siena: fix null pointer dereference in efx_hard_start_xmit Like in previous patch for sfc, prevent potential (but unlikely) NULL pointer dereference. En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: sfc/siena: corrige la desreferencia del puntero nulo en efx_hard_start_xmit Al igual que en el parche anterior para sfc, evita una posible (pero poco probable) desreferencia del puntero NULL. • https://git.kernel.org/stable/c/12804793b17c0e19115a90d98f2f3df0cb79e233 https://git.kernel.org/stable/c/a4eadca702dff0768dd01be6789bbec2a18e5b0a https://git.kernel.org/stable/c/589c6eded10c77a12b7b2cf235b6b19a2bdb91fa • CWE-476: NULL Pointer Dereference •
CVSS: -EPSS: 0%CPEs: 3EXPL: 0CVE-2022-48645 – net: enetc: deny offload of tc-based TSN features on VF interfaces
https://notcve.org/view.php?id=CVE-2022-48645
In the Linux kernel, the following vulnerability has been resolved: net: enetc: deny offload of tc-based TSN features on VF interfaces TSN features on the ENETC (taprio, cbs, gate, police) are configured through a mix of command BD ring messages and port registers: enetc_port_rd(), enetc_port_wr(). Port registers are a region of the ENETC memory map which are only accessible from the PCIe Physical Function. They are not accessible from the Virtual Functions. Moreover, attempting to access these registers crashes the kernel: $ echo 1 > /sys/bus/pci/devices/0000\:00\:00.0/sriov_numvfs pci 0000:00:01.0: [1957:ef00] type 00 class 0x020001 fsl_enetc_vf 0000:00:01.0: Adding to iommu group 15 fsl_enetc_vf 0000:00:01.0: enabling device (0000 -> 0002) fsl_enetc_vf 0000:00:01.0 eno0vf0: renamed from eth0 $ tc qdisc replace dev eno0vf0 root taprio num_tc 8 map 0 1 2 3 4 5 6 7 \ queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 base-time 0 \ sched-entry S 0x7f 900000 sched-entry S 0x80 100000 flags 0x2 Unable to handle kernel paging request at virtual address ffff800009551a08 Internal error: Oops: 96000007 [#1] PREEMPT SMP pc : enetc_setup_tc_taprio+0x170/0x47c lr : enetc_setup_tc_taprio+0x16c/0x47c Call trace: enetc_setup_tc_taprio+0x170/0x47c enetc_setup_tc+0x38/0x2dc taprio_change+0x43c/0x970 taprio_init+0x188/0x1e0 qdisc_create+0x114/0x470 tc_modify_qdisc+0x1fc/0x6c0 rtnetlink_rcv_msg+0x12c/0x390 Split enetc_setup_tc() into separate functions for the PF and for the VF drivers. Also remove enetc_qos.o from being included into enetc-vf.ko, since it serves absolutely no purpose there. En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: net: enetc: denegar la descarga de funciones TSN basadas en tc en interfaces VF Las funciones TSN en ENETC (taprio, cbs, gate, policial) se configuran mediante una combinación de comandos BD ring mensajes y registros de puertos: enetc_port_rd(), enetc_port_wr(). Los registros de puerto son una región del mapa de memoria ENETC a la que solo se puede acceder desde la función física PCIe. • https://git.kernel.org/stable/c/34c6adf1977b611fca3b824ad12a2a415e1e420e https://git.kernel.org/stable/c/510e703e4ed0e011db860bc21228aff48fc9eea7 https://git.kernel.org/stable/c/23022b74b1a23bed044f6bc96cf92f6ca5f3e75f https://git.kernel.org/stable/c/5641c751fe2f92d3d9e8a8e03c1263ac8caa0b42 •
CVSS: -EPSS: 0%CPEs: 5EXPL: 0CVE-2022-48644 – net/sched: taprio: avoid disabling offload when it was never enabled
https://notcve.org/view.php?id=CVE-2022-48644
In the Linux kernel, the following vulnerability has been resolved: net/sched: taprio: avoid disabling offload when it was never enabled In an incredibly strange API design decision, qdisc->destroy() gets called even if qdisc->init() never succeeded, not exclusively since commit 87b60cfacf9f ("net_sched: fix error recovery at qdisc creation"), but apparently also earlier (in the case of qdisc_create_dflt()). The taprio qdisc does not fully acknowledge this when it attempts full offload, because it starts off with q->flags = TAPRIO_FLAGS_INVALID in taprio_init(), then it replaces q->flags with TCA_TAPRIO_ATTR_FLAGS parsed from netlink (in taprio_change(), tail called from taprio_init()). But in taprio_destroy(), we call taprio_disable_offload(), and this determines what to do based on FULL_OFFLOAD_IS_ENABLED(q->flags). But looking at the implementation of FULL_OFFLOAD_IS_ENABLED() (a bitwise check of bit 1 in q->flags), it is invalid to call this macro on q->flags when it contains TAPRIO_FLAGS_INVALID, because that is set to U32_MAX, and therefore FULL_OFFLOAD_IS_ENABLED() will return true on an invalid set of flags. As a result, it is possible to crash the kernel if user space forces an error between setting q->flags = TAPRIO_FLAGS_INVALID, and the calling of taprio_enable_offload(). This is because drivers do not expect the offload to be disabled when it was never enabled. The error that we force here is to attach taprio as a non-root qdisc, but instead as child of an mqprio root qdisc: $ tc qdisc add dev swp0 root handle 1: \ mqprio num_tc 8 map 0 1 2 3 4 5 6 7 \ queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 hw 0 $ tc qdisc replace dev swp0 parent 1:1 \ taprio num_tc 8 map 0 1 2 3 4 5 6 7 \ queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 base-time 0 \ sched-entry S 0x7f 990000 sched-entry S 0x80 100000 \ flags 0x0 clockid CLOCK_TAI Unable to handle kernel paging request at virtual address fffffffffffffff8 [fffffffffffffff8] pgd=0000000000000000, p4d=0000000000000000 Internal error: Oops: 96000004 [#1] PREEMPT SMP Call trace: taprio_dump+0x27c/0x310 vsc9959_port_setup_tc+0x1f4/0x460 felix_port_setup_tc+0x24/0x3c dsa_slave_setup_tc+0x54/0x27c taprio_disable_offload.isra.0+0x58/0xe0 taprio_destroy+0x80/0x104 qdisc_create+0x240/0x470 tc_modify_qdisc+0x1fc/0x6b0 rtnetlink_rcv_msg+0x12c/0x390 netlink_rcv_skb+0x5c/0x130 rtnetlink_rcv+0x1c/0x2c Fix this by keeping track of the operations we made, and undo the offload only if we actually did it. I've added "bool offloaded" inside a 4 byte hole between "int clockid" and "atomic64_t picos_per_byte". Now the first cache line looks like below: $ pahole -C taprio_sched net/sched/sch_taprio.o struct taprio_sched { struct Qdisc * * qdiscs; /* 0 8 */ struct Qdisc * root; /* 8 8 */ u32 flags; /* 16 4 */ enum tk_offsets tk_offset; /* 20 4 */ int clockid; /* 24 4 */ bool offloaded; /* 28 1 */ /* XXX 3 bytes hole, try to pack */ atomic64_t picos_per_byte; /* 32 0 */ /* XXX 8 bytes hole, try to pack */ spinlock_t current_entry_lock; /* 40 0 */ /* XXX 8 bytes hole, try to pack */ struct sched_entry * current_entry; /* 48 8 */ struct sched_gate_list * oper_sched; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: net/sched: taprio: evita deshabilitar la descarga cuando nunca estuvo habilitada. En una decisión de diseño de API increíblemente extraña, se llama a qdisc->destroy() incluso si qdisc->init() nunca tuvo éxito, no exclusivamente desde el commit 87b60cfacf9f ("net_sched: corregir recuperación de error en la creación de qdisc"), sino aparentemente también antes (en el caso de qdisc_create_dflt()). La qdisc taprio no reconoce completamente esto cuando intenta una descarga completa, porque comienza con q->flags = TAPRIO_FLAGS_INVALID en taprio_init(), luego reemplaza q->flags con TCA_TAPRIO_ATTR_FLAGS analizado desde netlink (en taprio_change(), cola llamada de taprio_init()). • https://git.kernel.org/stable/c/9c66d15646760eb8982242b4531c4d4fd36118fd https://git.kernel.org/stable/c/d12a1eb07003e597077329767c6aa86a7e972c76 https://git.kernel.org/stable/c/586def6ebed195f3594a4884f7c5334d0e1ad1bb https://git.kernel.org/stable/c/f58e43184226e5e9662088ccf1389e424a3a4cbd https://git.kernel.org/stable/c/c7c9c7eb305ab8b4e93e4e4e1b78d8cfcbc26323 https://git.kernel.org/stable/c/db46e3a88a09c5cf7e505664d01da7238cd56c92 •