ghsa-p672-9qr7-4cmf
Vulnerability from github
In the Linux kernel, the following vulnerability has been resolved:
interconnect: Don't access req_list while it's being manipulated
The icc_lock mutex was split into separate icc_lock and icc_bw_lock mutexes in [1] to avoid lockdep splats. However, this didn't adequately protect access to icc_node::req_list.
The icc_set_bw() function will eventually iterate over req_list while only holding icc_bw_lock, but req_list can be modified while only holding icc_lock. This causes races between icc_set_bw(), of_icc_get(), and icc_put().
Example A:
CPU0 CPU1
---- ----
icc_set_bw(path_a)
mutex_lock(&icc_bw_lock);
icc_put(path_b)
mutex_lock(&icc_lock);
aggregate_requests()
hlist_for_each_entry(r, ...
hlist_del(...
Example B:
CPU0 CPU1
---- ----
icc_set_bw(path_a)
mutex_lock(&icc_bw_lock);
path_b = of_icc_get()
of_icc_get_by_index()
mutex_lock(&icc_lock);
path_find()
path_init()
aggregate_requests()
hlist_for_each_entry(r, ...
hlist_add_head(...
Fix this by ensuring icc_bw_lock is always held before manipulating icc_node::req_list. The additional places icc_bw_lock is held don't perform any memory allocations, so we should still be safe from the original lockdep splats that motivated the separate locks.
[1] commit af42269c3523 ("interconnect: Fix locking for runpm vs reclaim")
{
"affected": [],
"aliases": [
"CVE-2024-27005"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-05-01T06:15:18Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\ninterconnect: Don\u0027t access req_list while it\u0027s being manipulated\n\nThe icc_lock mutex was split into separate icc_lock and icc_bw_lock\nmutexes in [1] to avoid lockdep splats. However, this didn\u0027t adequately\nprotect access to icc_node::req_list.\n\nThe icc_set_bw() function will eventually iterate over req_list while\nonly holding icc_bw_lock, but req_list can be modified while only\nholding icc_lock. This causes races between icc_set_bw(), of_icc_get(),\nand icc_put().\n\nExample A:\n\n CPU0 CPU1\n ---- ----\n icc_set_bw(path_a)\n mutex_lock(\u0026icc_bw_lock);\n icc_put(path_b)\n mutex_lock(\u0026icc_lock);\n aggregate_requests()\n hlist_for_each_entry(r, ...\n hlist_del(...\n \u003cr = invalid pointer\u003e\n\nExample B:\n\n CPU0 CPU1\n ---- ----\n icc_set_bw(path_a)\n mutex_lock(\u0026icc_bw_lock);\n path_b = of_icc_get()\n of_icc_get_by_index()\n mutex_lock(\u0026icc_lock);\n path_find()\n path_init()\n aggregate_requests()\n hlist_for_each_entry(r, ...\n hlist_add_head(...\n \u003cr = invalid pointer\u003e\n\nFix this by ensuring icc_bw_lock is always held before manipulating\nicc_node::req_list. The additional places icc_bw_lock is held don\u0027t\nperform any memory allocations, so we should still be safe from the\noriginal lockdep splats that motivated the separate locks.\n\n[1] commit af42269c3523 (\"interconnect: Fix locking for runpm vs reclaim\")",
"id": "GHSA-p672-9qr7-4cmf",
"modified": "2024-05-03T03:30:47Z",
"published": "2024-05-01T06:31:43Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-27005"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/4c65507121ea8e0b47fae6d2049c8688390d46b6"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/d0d04efa2e367921654b5106cc5c05e3757c2b42"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/de1bf25b6d771abdb52d43546cf57ad775fb68a1"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/message/4EZ6PJW7VOZ224TD7N4JZNU6KV32ZJ53"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/message/DAMSOZXJEPUOXW33WZYWCVAY7Z5S7OOY"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/message/GCBZZEC7L7KTWWAS2NLJK6SO3IZIL4WW"
}
],
"schema_version": "1.4.0",
"severity": []
}