bpf: Prepare bpf_prog_put() to be called from irq context.
authorAlexei Starovoitov <ast@kernel.org>
Thu, 15 Jul 2021 00:54:07 +0000 (17:54 -0700)
committerDaniel Borkmann <daniel@iogearbox.net>
Thu, 15 Jul 2021 20:31:10 +0000 (22:31 +0200)
commitd809e134be7a1fdd9f5b99ab3291c6da5c0b8240
treea777e2e1bfec485860b673d4b05fded73fd7c1fb
parentde587d564f957b685e47da1848d428b86173766d
bpf: Prepare bpf_prog_put() to be called from irq context.

Currently bpf_prog_put() is called from the task context only.
With addition of bpf timers the timer related helpers will start calling
bpf_prog_put() from irq-saved region and in rare cases might drop
the refcnt to zero.
To address this case, first, convert bpf_prog_free_id() to be irq-save
(this is similar to bpf_map_free_id), and, second, defer non irq
appropriate calls into work queue.
For example:
bpf_audit_prog() is calling kmalloc and wake_up_interruptible,
bpf_prog_kallsyms_del_all()->bpf_ksym_del()->spin_unlock_bh().
They are not safe with irqs disabled.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/20210715005417.78572-2-alexei.starovoitov@gmail.com
kernel/bpf/syscall.c