ghsa-8x8h-8p2w-8g4f
Vulnerability from github
In the Linux kernel, the following vulnerability has been resolved:
bpf: Fix racing between bpf_timer_cancel_and_free and bpf_timer_cancel
The following race is possible between bpf_timer_cancel_and_free and bpf_timer_cancel. It will lead a UAF on the timer->timer.
bpf_timer_cancel(); spin_lock(); t = timer->time; spin_unlock();
bpf_timer_cancel_and_free();
spin_lock();
t = timer->timer;
timer->timer = NULL;
spin_unlock();
hrtimer_cancel(&t->timer);
kfree(t);
/* UAF on t */
hrtimer_cancel(&t->timer);
In bpf_timer_cancel_and_free, this patch frees the timer->timer after a rcu grace period. This requires a rcu_head addition to the "struct bpf_hrtimer". Another kfree(t) happens in bpf_timer_init, this does not need a kfree_rcu because it is still under the spin_lock and timer->timer has not been visible by others yet.
In bpf_timer_cancel, rcu_read_lock() is added because this helper can be used in a non rcu critical section context (e.g. from a sleepable bpf prog). Other timer->timer usages in helpers.c have been audited, bpf_timer_cancel() is the only place where timer->timer is used outside of the spin_lock.
Another solution considered is to mark a t->flag in bpf_timer_cancel and clear it after hrtimer_cancel() is done. In bpf_timer_cancel_and_free, it busy waits for the flag to be cleared before kfree(t). This patch goes with a straight forward solution and frees timer->timer after a rcu grace period.
{
"affected": [],
"aliases": [
"CVE-2024-26737"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-04-03T17:15:51Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nbpf: Fix racing between bpf_timer_cancel_and_free and bpf_timer_cancel\n\nThe following race is possible between bpf_timer_cancel_and_free\nand bpf_timer_cancel. It will lead a UAF on the timer-\u003etimer.\n\nbpf_timer_cancel();\n\tspin_lock();\n\tt = timer-\u003etime;\n\tspin_unlock();\n\n\t\t\t\t\tbpf_timer_cancel_and_free();\n\t\t\t\t\t\tspin_lock();\n\t\t\t\t\t\tt = timer-\u003etimer;\n\t\t\t\t\t\ttimer-\u003etimer = NULL;\n\t\t\t\t\t\tspin_unlock();\n\t\t\t\t\t\thrtimer_cancel(\u0026t-\u003etimer);\n\t\t\t\t\t\tkfree(t);\n\n\t/* UAF on t */\n\thrtimer_cancel(\u0026t-\u003etimer);\n\nIn bpf_timer_cancel_and_free, this patch frees the timer-\u003etimer\nafter a rcu grace period. This requires a rcu_head addition\nto the \"struct bpf_hrtimer\". Another kfree(t) happens in bpf_timer_init,\nthis does not need a kfree_rcu because it is still under the\nspin_lock and timer-\u003etimer has not been visible by others yet.\n\nIn bpf_timer_cancel, rcu_read_lock() is added because this helper\ncan be used in a non rcu critical section context (e.g. from\na sleepable bpf prog). Other timer-\u003etimer usages in helpers.c\nhave been audited, bpf_timer_cancel() is the only place where\ntimer-\u003etimer is used outside of the spin_lock.\n\nAnother solution considered is to mark a t-\u003eflag in bpf_timer_cancel\nand clear it after hrtimer_cancel() is done. In bpf_timer_cancel_and_free,\nit busy waits for the flag to be cleared before kfree(t). This patch\ngoes with a straight forward solution and frees timer-\u003etimer after\na rcu grace period.",
"id": "GHSA-8x8h-8p2w-8g4f",
"modified": "2024-04-03T18:30:42Z",
"published": "2024-04-03T18:30:42Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-26737"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/0281b919e175bb9c3128bd3872ac2903e9436e3f"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/5268bb02107b9eedfdcd51db75b407d10043368c"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/7d80a9e745fa5b47da3bca001f186c02485c7c33"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/8327ed12e8ebc5436bfaa1786c49988894f9c8a6"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/addf5e297e6cbf5341f9c07720693ca9ba0057b5"
}
],
"schema_version": "1.4.0",
"severity": []
}