fkie_cve-2022-49872
Vulnerability from fkie_nvd
Published
2025-05-01 15:16
Modified
2025-05-02 13:52
Severity ?
Summary
In the Linux kernel, the following vulnerability has been resolved:
net: gso: fix panic on frag_list with mixed head alloc types
Since commit 3dcbdb134f32 ("net: gso: Fix skb_segment splat when
splitting gso_size mangled skb having linear-headed frag_list"), it is
allowed to change gso_size of a GRO packet. However, that commit assumes
that "checking the first list_skb member suffices; i.e if either of the
list_skb members have non head_frag head, then the first one has too".
It turns out this assumption does not hold. We've seen BUG_ON being hit
in skb_segment when skbs on the frag_list had differing head_frag with
the vmxnet3 driver. This happens because __netdev_alloc_skb and
__napi_alloc_skb can return a skb that is page backed or kmalloced
depending on the requested size. As the result, the last small skb in
the GRO packet can be kmalloced.
There are three different locations where this can be fixed:
(1) We could check head_frag in GRO and not allow GROing skbs with
different head_frag. However, that would lead to performance
regression on normal forward paths with unmodified gso_size, where
!head_frag in the last packet is not a problem.
(2) Set a flag in bpf_skb_net_grow and bpf_skb_net_shrink indicating
that NETIF_F_SG is undesirable. That would need to eat a bit in
sk_buff. Furthermore, that flag can be unset when all skbs on the
frag_list are page backed. To retain good performance,
bpf_skb_net_grow/shrink would have to walk the frag_list.
(3) Walk the frag_list in skb_segment when determining whether
NETIF_F_SG should be cleared. This of course slows things down.
This patch implements (3). To limit the performance impact in
skb_segment, the list is walked only for skbs with SKB_GSO_DODGY set
that have gso_size changed. Normal paths thus will not hit it.
We could check only the last skb but since we need to walk the whole
list anyway, let's stay on the safe side.
References
Impacted products
| Vendor | Product | Version |
|---|
{
"cveTags": [],
"descriptions": [
{
"lang": "en",
"value": "In the Linux kernel, the following vulnerability has been resolved:\n\nnet: gso: fix panic on frag_list with mixed head alloc types\n\nSince commit 3dcbdb134f32 (\"net: gso: Fix skb_segment splat when\nsplitting gso_size mangled skb having linear-headed frag_list\"), it is\nallowed to change gso_size of a GRO packet. However, that commit assumes\nthat \"checking the first list_skb member suffices; i.e if either of the\nlist_skb members have non head_frag head, then the first one has too\".\n\nIt turns out this assumption does not hold. We\u0027ve seen BUG_ON being hit\nin skb_segment when skbs on the frag_list had differing head_frag with\nthe vmxnet3 driver. This happens because __netdev_alloc_skb and\n__napi_alloc_skb can return a skb that is page backed or kmalloced\ndepending on the requested size. As the result, the last small skb in\nthe GRO packet can be kmalloced.\n\nThere are three different locations where this can be fixed:\n\n(1) We could check head_frag in GRO and not allow GROing skbs with\n different head_frag. However, that would lead to performance\n regression on normal forward paths with unmodified gso_size, where\n !head_frag in the last packet is not a problem.\n\n(2) Set a flag in bpf_skb_net_grow and bpf_skb_net_shrink indicating\n that NETIF_F_SG is undesirable. That would need to eat a bit in\n sk_buff. Furthermore, that flag can be unset when all skbs on the\n frag_list are page backed. To retain good performance,\n bpf_skb_net_grow/shrink would have to walk the frag_list.\n\n(3) Walk the frag_list in skb_segment when determining whether\n NETIF_F_SG should be cleared. This of course slows things down.\n\nThis patch implements (3). To limit the performance impact in\nskb_segment, the list is walked only for skbs with SKB_GSO_DODGY set\nthat have gso_size changed. Normal paths thus will not hit it.\n\nWe could check only the last skb but since we need to walk the whole\nlist anyway, let\u0027s stay on the safe side."
},
{
"lang": "es",
"value": "En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: net: gso: correcci\u00f3n del p\u00e1nico en frag_list con tipos de asignaci\u00f3n de encabezado mixtos. Desde el commit 3dcbdb134f32 (\"net: gso: correcci\u00f3n del error skb_segment al dividir gso_size de un skb alterado con frag_list de encabezado lineal\"), se permite cambiar gso_size de un paquete GRO. Sin embargo, esta confirmaci\u00f3n asume que basta con comprobar el primer miembro de list_skb; es decir, si alguno de los miembros de list_skb tiene un encabezado distinto de head_frag, el primero tambi\u00e9n lo tendr\u00e1. Resulta que esta suposici\u00f3n no se cumple. Hemos observado que se ha alcanzado BUG_ON en skb_segment cuando los skbs de frag_list ten\u00edan un head_frag distinto con el controlador vmxnet3. Esto se debe a que __netdev_alloc_skb y __napi_alloc_skb pueden devolver un skb con respaldo de p\u00e1gina o asignado a km, seg\u00fan el tama\u00f1o solicitado. Como resultado, el \u00faltimo skb peque\u00f1o en el paquete GRO puede ser asignado a km. Hay tres ubicaciones diferentes donde esto puede ser corregido: (1) Podr\u00edamos revisar head_frag en GRO y no permitir que GROing skbs con diferente head_frag. Sin embargo, eso llevar\u00eda a una regresi\u00f3n del rendimiento en rutas de avance normales con gso_size sin modificar, donde !head_frag en el \u00faltimo paquete no es un problema. (2) Establecer un indicador en bpf_skb_net_grow y bpf_skb_net_shrink que indique que NETIF_F_SG no es deseable. Esto requerir\u00eda consumir un poco en sk_buff. Adem\u00e1s, ese indicador puede ser desactivado cuando todos los skbs en la frag_list est\u00e1n en retroceso de p\u00e1gina. Para mantener un buen rendimiento, bpf_skb_net_grow/shrink tendr\u00eda que recorrer la frag_list. (3) Recorrer la frag_list en skb_segment al determinar si NETIF_F_SG debe ser borrado. Esto, por supuesto, ralentiza el proceso. Este parche implementa (3). Para limitar el impacto en el rendimiento de skb_segment, la lista solo se recorre para skb con SKB_GSO_DODGY configurado y gso_size modificado. Por lo tanto, las rutas normales no la alcanzar\u00e1n. Podr\u00edamos revisar solo el \u00faltimo skb, pero como necesitamos recorrer la lista completa de todos modos, mejor optemos por lo seguro."
}
],
"id": "CVE-2022-49872",
"lastModified": "2025-05-02T13:52:51.693",
"metrics": {},
"published": "2025-05-01T15:16:12.133",
"references": [
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/0a9f56e525ea871d3950b90076912f5c7494f00f"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/50868de7dc4e7f0fcadd6029f32bf4387c102ee6"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/5876b7f249a1ecbbcc8e35072c3828d6526d1c3a"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/598d9e30927b15731e83797fbd700ecf399f42dd"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/65ad047fd83502447269fda8fd26c99077a9af47"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/9e4b7a99a03aefd37ba7bb1f022c8efab5019165"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/ad25a115f50800c6847e0d841c5c7992a9f7c1b3"
},
{
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"url": "https://git.kernel.org/stable/c/bd5362e58721e4d0d1a37796593bd6e51536ce7a"
}
],
"sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"vulnStatus": "Awaiting Analysis"
}
Loading…
Loading…
Sightings
| Author | Source | Type | Date |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or seen somewhere by the user.
- Confirmed: The vulnerability is confirmed from an analyst perspective.
- Exploited: This vulnerability was exploited and seen by the user reporting the sighting.
- Patched: This vulnerability was successfully patched by the user reporting the sighting.
- Not exploited: This vulnerability was not exploited or seen by the user reporting the sighting.
- Not confirmed: The user expresses doubt about the veracity of the vulnerability.
- Not patched: This vulnerability was not successfully patched by the user reporting the sighting.
Loading…
Loading…