linux.git/kernel/bpf/cgroup.c, branch v6.6.131

bpf: Allow pre-ordering for bpf cgroup progs

2025-06-04T12:41:58+00:00

[ Upstream commit 4b82b181a26cff8bf7adc3a85a88d121d92edeaf ]

Currently for bpf progs in a cgroup hierarchy, the effective prog array
is computed from bottom cgroup to upper cgroups (post-ordering). For
example, the following cgroup hierarchy
    root cgroup: p1, p2
        subcgroup: p3, p4
have BPF_F_ALLOW_MULTI for both cgroup levels.
The effective cgroup array ordering looks like
    p3 p4 p1 p2
and at run time, progs will execute based on that order.

But in some cases, it is desirable to have root prog executes earlier than
children progs (pre-ordering). For example,
  - prog p1 intends to collect original pkt dest addresses.
  - prog p3 will modify original pkt dest addresses to a proxy address for
    security reason.
The end result is that prog p1 gets proxy address which is not what it
wants. Putting p1 to every child cgroup is not desirable either as it
will duplicate itself in many child cgroups. And this is exactly a use case
we are encountering in Meta.

To fix this issue, let us introduce a flag BPF_F_PREORDER. If the flag
is specified at attachment time, the prog has higher priority and the
ordering with that flag will be from top to bottom (pre-ordering).
For example, in the above example,
    root cgroup: p1, p2
        subcgroup: p3, p4
Let us say p2 and p4 are marked with BPF_F_PREORDER. The final
effective array ordering will be
    p2 p4 p3 p1

Suggested-by: Andrii Nakryiko 
Acked-by: Andrii Nakryiko 
Signed-off-by: Yonghong Song 
Link: https://lore.kernel.org/r/20250224230116.283071-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov 
Signed-off-by: Sasha Levin

cgroup/bpf: use a dedicated workqueue for cgroup bpf destruction

2024-11-08T15:28:24+00:00

[ Upstream commit 117932eea99b729ee5d12783601a4f7f5fd58a23 ]

A hung_task problem shown below was found:

INFO: task kworker/0:0:8 blocked for more than 327 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
Workqueue: events cgroup_bpf_release
Call Trace:
 
 __schedule+0x5a2/0x2050
 ? find_held_lock+0x33/0x100
 ? wq_worker_sleeping+0x9e/0xe0
 schedule+0x9f/0x180
 schedule_preempt_disabled+0x25/0x50
 __mutex_lock+0x512/0x740
 ? cgroup_bpf_release+0x1e/0x4d0
 ? cgroup_bpf_release+0xcf/0x4d0
 ? process_scheduled_works+0x161/0x8a0
 ? cgroup_bpf_release+0x1e/0x4d0
 ? mutex_lock_nested+0x2b/0x40
 ? __pfx_delay_tsc+0x10/0x10
 mutex_lock_nested+0x2b/0x40
 cgroup_bpf_release+0xcf/0x4d0
 ? process_scheduled_works+0x161/0x8a0
 ? trace_event_raw_event_workqueue_execute_start+0x64/0xd0
 ? process_scheduled_works+0x161/0x8a0
 process_scheduled_works+0x23a/0x8a0
 worker_thread+0x231/0x5b0
 ? __pfx_worker_thread+0x10/0x10
 kthread+0x14d/0x1c0
 ? __pfx_kthread+0x10/0x10
 ret_from_fork+0x59/0x70
 ? __pfx_kthread+0x10/0x10
 ret_from_fork_asm+0x1b/0x30
 

This issue can be reproduced by the following pressuse test:
1. A large number of cpuset cgroups are deleted.
2. Set cpu on and off repeatly.
3. Set watchdog_thresh repeatly.
The scripts can be obtained at LINK mentioned above the signature.

The reason for this issue is cgroup_mutex and cpu_hotplug_lock are
acquired in different tasks, which may lead to deadlock.
It can lead to a deadlock through the following steps:
1. A large number of cpusets are deleted asynchronously, which puts a
   large number of cgroup_bpf_release works into system_wq. The max_active
   of system_wq is WQ_DFL_ACTIVE(256). Consequently, all active works are
   cgroup_bpf_release works, and many cgroup_bpf_release works will be put
   into inactive queue. As illustrated in the diagram, there are 256 (in
   the acvtive queue) + n (in the inactive queue) works.
2. Setting watchdog_thresh will hold cpu_hotplug_lock.read and put
   smp_call_on_cpu work into system_wq. However step 1 has already filled
   system_wq, 'sscs.work' is put into inactive queue. 'sscs.work' has
   to wait until the works that were put into the inacvtive queue earlier
   have executed (n cgroup_bpf_release), so it will be blocked for a while.
3. Cpu offline requires cpu_hotplug_lock.write, which is blocked by step 2.
4. Cpusets that were deleted at step 1 put cgroup_release works into
   cgroup_destroy_wq. They are competing to get cgroup_mutex all the time.
   When cgroup_metux is acqured by work at css_killed_work_fn, it will
   call cpuset_css_offline, which needs to acqure cpu_hotplug_lock.read.
   However, cpuset_css_offline will be blocked for step 3.
5. At this moment, there are 256 works in active queue that are
   cgroup_bpf_release, they are attempting to acquire cgroup_mutex, and as
   a result, all of them are blocked. Consequently, sscs.work can not be
   executed. Ultimately, this situation leads to four processes being
   blocked, forming a deadlock.

system_wq(step1)		WatchDog(step2)			cpu offline(step3)	cgroup_destroy_wq(step4)
...
2000+ cgroups deleted asyn
256 actives + n inactives
				__lockup_detector_reconfigure
				P(cpu_hotplug_lock.read)
				put sscs.work into system_wq
256 + n + 1(sscs.work)
sscs.work wait to be executed
				warting sscs.work finish
								percpu_down_write
								P(cpu_hotplug_lock.write)
								...blocking...
											css_killed_work_fn
											P(cgroup_mutex)
											cpuset_css_offline
											P(cpu_hotplug_lock.read)
											...blocking...
256 cgroup_bpf_release
mutex_lock(&cgroup_mutex);
..blocking...

To fix the problem, place cgroup_bpf_release works on a dedicated
workqueue which can break the loop and solve the problem. System wqs are
for misc things which shouldn't create a large number of concurrent work
items. If something is going to generate >WQ_DFL_ACTIVE(256) concurrent
work items, it should use its own dedicated workqueue.

Fixes: 4bfc0bb2c60e ("bpf: decouple the lifetime of cgroup_bpf from cgroup itself")
Cc: stable@vger.kernel.org # v5.3+
Link: https://lore.kernel.org/cgroups/e90c32d2-2a85-4f28-9154-09c7d320cb60@huawei.com/T/#t
Tested-by: Vishal Chourasia 
Signed-off-by: Chen Ridong 
Signed-off-by: Tejun Heo 
Signed-off-by: Sasha Levin

bpf: Add sockptr support for setsockopt

2024-09-12T09:11:34+00:00

[ Upstream commit 3f31e0d14d44ad491a81b7c1f83f32fbc300a867 ]

The whole network stack uses sockptr, and while it doesn't move to
something more modern, let's use sockptr in setsockptr BPF hooks, so, it
could be used by other callers.

The main motivation for this change is to use it in the io_uring
{g,s}etsockopt(), which will use a userspace pointer for *optval, but, a
kernel value for optlen.

Link: https://lore.kernel.org/all/ZSArfLaaGcfd8LH8@gmail.com/

Signed-off-by: Breno Leitao 
Acked-by: Martin KaFai Lau 
Link: https://lore.kernel.org/r/20231016134750.1381153-3-leitao@debian.org
Signed-off-by: Jens Axboe 
Stable-dep-of: 33f339a1ba54 ("bpf, net: Fix a potential race in do_sock_getsockopt()")
Signed-off-by: Sasha Levin

bpf: Add sockptr support for getsockopt

2024-09-12T09:11:34+00:00

[ Upstream commit a615f67e1a426f35366b8398c11f31c148e7df48 ]

The whole network stack uses sockptr, and while it doesn't move to
something more modern, let's use sockptr in getsockptr BPF hooks, so, it
could be used by other callers.

The main motivation for this change is to use it in the io_uring
{g,s}etsockopt(), which will use a userspace pointer for *optval, but, a
kernel value for optlen.

Link: https://lore.kernel.org/all/ZSArfLaaGcfd8LH8@gmail.com/

Signed-off-by: Breno Leitao 
Acked-by: Martin KaFai Lau 
Link: https://lore.kernel.org/r/20231016134750.1381153-2-leitao@debian.org
Signed-off-by: Jens Axboe 
Stable-dep-of: 33f339a1ba54 ("bpf, net: Fix a potential race in do_sock_getsockopt()")
Signed-off-by: Sasha Levin

bpf: Propagate modified uaddrlen from cgroup sockaddr programs

2024-02-01T00:19:04+00:00

[ Upstream commit fefba7d1ae198dcbf8b3b432de46a4e29f8dbd8c ]

As prep for adding unix socket support to the cgroup sockaddr hooks,
let's propagate the sockaddr length back to the caller after running
a bpf cgroup sockaddr hook program. While not important for AF_INET or
AF_INET6, the sockaddr length is important when working with AF_UNIX
sockaddrs as the size of the sockaddr cannot be determined just from the
address family or the sockaddr's contents.

__cgroup_bpf_run_filter_sock_addr() is modified to take the uaddrlen as
an input/output argument. After running the program, the modified sockaddr
length is stored in the uaddrlen pointer.

Signed-off-by: Daan De Meyer 
Link: https://lore.kernel.org/r/20231011185113.140426-3-daan.j.demeyer@gmail.com
Signed-off-by: Martin KaFai Lau 
Stable-dep-of: c5114710c8ce ("xsk: fix usage of multi-buffer BPF helpers for ZC XDP")
Signed-off-by: Sasha Levin

bpf, cgroup: fix multiple kernel-doc warnings

2023-09-12T20:19:07+00:00

Fix missing or extra function parameter kernel-doc warnings
in cgroup.c:

kernel/bpf/cgroup.c:1359: warning: Excess function parameter 'type' description in '__cgroup_bpf_run_filter_skb'
kernel/bpf/cgroup.c:1359: warning: Function parameter or member 'atype' not described in '__cgroup_bpf_run_filter_skb'
kernel/bpf/cgroup.c:1439: warning: Excess function parameter 'type' description in '__cgroup_bpf_run_filter_sk'
kernel/bpf/cgroup.c:1439: warning: Function parameter or member 'atype' not described in '__cgroup_bpf_run_filter_sk'
kernel/bpf/cgroup.c:1467: warning: Excess function parameter 'type' description in '__cgroup_bpf_run_filter_sock_addr'
kernel/bpf/cgroup.c:1467: warning: Function parameter or member 'atype' not described in '__cgroup_bpf_run_filter_sock_addr'
kernel/bpf/cgroup.c:1512: warning: Excess function parameter 'type' description in '__cgroup_bpf_run_filter_sock_ops'
kernel/bpf/cgroup.c:1512: warning: Function parameter or member 'atype' not described in '__cgroup_bpf_run_filter_sock_ops'
kernel/bpf/cgroup.c:1685: warning: Excess function parameter 'type' description in '__cgroup_bpf_run_filter_sysctl'
kernel/bpf/cgroup.c:1685: warning: Function parameter or member 'atype' not described in '__cgroup_bpf_run_filter_sysctl'
kernel/bpf/cgroup.c:795: warning: Excess function parameter 'type' description in '__cgroup_bpf_replace'
kernel/bpf/cgroup.c:795: warning: Function parameter or member 'new_prog' not described in '__cgroup_bpf_replace'

Signed-off-by: Randy Dunlap 
Cc: Martin KaFai Lau 
Cc: bpf@vger.kernel.org
Cc: Alexei Starovoitov 
Cc: Daniel Borkmann 
Cc: Andrii Nakryiko 
Link: https://lore.kernel.org/r/20230912060812.1715-1-rdunlap@infradead.org
Signed-off-by: Alexei Starovoitov

Merge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next

2023-05-17T02:50:05+00:00

Daniel Borkmann says:

====================
pull-request: bpf-next 2023-05-16

We've added 57 non-merge commits during the last 19 day(s) which contain
a total of 63 files changed, 3293 insertions(+), 690 deletions(-).

The main changes are:

1) Add precision propagation to verifier for subprogs and callbacks,
   from Andrii Nakryiko.

2) Improve BPF's {g,s}setsockopt() handling with wrong option lengths,
   from Stanislav Fomichev.

3) Utilize pahole v1.25 for the kernel's BTF generation to filter out
   inconsistent function prototypes, from Alan Maguire.

4) Various dyn-pointer verifier improvements to relax restrictions,
   from Daniel Rosenberg.

5) Add a new bpf_task_under_cgroup() kfunc for designated task,
   from Feng Zhou.

6) Unblock tests for arm64 BPF CI after ftrace supporting direct call,
   from Florent Revest.

7) Add XDP hint kfunc metadata for RX hash/timestamp for igc,
   from Jesper Dangaard Brouer.

8) Add several new dyn-pointer kfuncs to ease their usability,
   from Joanne Koong.

9) Add in-depth LRU internals description and dot function graph,
   from Joe Stringer.

10) Fix KCSAN report on bpf_lru_list when accessing node->ref,
    from Martin KaFai Lau.

11) Only dump unprivileged_bpf_disabled log warning upon write,
    from Kui-Feng Lee.

12) Extend test_progs to directly passing allow/denylist file,
    from Stephen Veiss.

13) Fix BPF trampoline memleak upon failure attaching to fentry,
    from Yafang Shao.

14) Fix emitting struct bpf_tcp_sock type in vmlinux BTF,
    from Yonghong Song.

* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (57 commits)
  bpf: Fix memleak due to fentry attach failure
  bpf: Remove bpf trampoline selector
  bpf, arm64: Support struct arguments in the BPF trampoline
  bpftool: JIT limited misreported as negative value on aarch64
  bpf: fix calculation of subseq_idx during precision backtracking
  bpf: Remove anonymous union in bpf_kfunc_call_arg_meta
  bpf: Document EFAULT changes for sockopt
  selftests/bpf: Correctly handle optlen > 4096
  selftests/bpf: Update EFAULT {g,s}etsockopt selftests
  bpf: Don't EFAULT for {g,s}setsockopt with wrong optlen
  libbpf: fix offsetof() and container_of() to work with CO-RE
  bpf: Address KCSAN report on bpf_lru_list
  bpf: Add --skip_encoding_btf_inconsistent_proto, --btf_gen_optimized to pahole flags for v1.25
  selftests/bpf: Accept mem from dynptr in helper funcs
  bpf: verifier: Accept dynptr mem as mem in helpers
  selftests/bpf: Check overflow in optional buffer
  selftests/bpf: Test allowing NULL buffer in dynptr slice
  bpf: Allow NULL buffers in bpf_dynptr_slice(_rw)
  selftests/bpf: Add testcase for bpf_task_under_cgroup
  bpf: Add bpf_task_under_cgroup() kfunc
  ...
====================

Link: https://lore.kernel.org/r/20230515225603.27027-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski

bpf: Don't EFAULT for {g,s}setsockopt with wrong optlen

2023-05-13T23:20:15+00:00

With the way the hooks implemented right now, we have a special
condition: optval larger than PAGE_SIZE will expose only first 4k into
BPF; any modifications to the optval are ignored. If the BPF program
doesn't handle this condition by resetting optlen to 0,
the userspace will get EFAULT.

The intention of the EFAULT was to make it apparent to the
developers that the program is doing something wrong.
However, this inadvertently might affect production workloads
with the BPF programs that are not too careful (i.e., returning EFAULT
for perfectly valid setsockopt/getsockopt calls).

Let's try to minimize the chance of BPF program screwing up userspace
by ignoring the output of those BPF programs (instead of returning
EFAULT to the userspace). pr_info_once those cases to
the dmesg to help with figuring out what's going wrong.

Fixes: 0d01da6afc54 ("bpf: implement getsockopt and setsockopt hooks")
Suggested-by: Martin KaFai Lau 
Signed-off-by: Stanislav Fomichev 
Link: https://lore.kernel.org/r/20230511170456.1759459-2-sdf@google.com
Signed-off-by: Martin KaFai Lau

Merge tag 'cgroup-for-6.4' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup

2023-04-29T17:05:22+00:00

Pull cgroup updates from Tejun Heo:

 - cpuset changes including the fix for an incorrect interaction with
   CPU hotplug and an optimization

 - Other doc and cosmetic changes

* tag 'cgroup-for-6.4' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  docs: cgroup-v1/cpusets: update libcgroup project link
  cgroup/cpuset: Minor updates to test_cpuset_prs.sh
  cgroup/cpuset: Include offline CPUs when tasks' cpumasks in top_cpuset are updated
  cgroup/cpuset: Skip task update if hotplug doesn't affect current cpuset
  cpuset: Clean up cpuset_node_allowed
  cgroup: bpf: use cgroup_lock()/cgroup_unlock() wrappers

bpf: Don't EFAULT for getsockopt with optval=NULL

2023-04-21T15:09:53+00:00

Some socket options do getsockopt with optval=NULL to estimate the size
of the final buffer (which is returned via optlen). This breaks BPF
getsockopt assumptions about permitted optval buffer size. Let's enforce
these assumptions only when non-NULL optval is provided.

Fixes: 0d01da6afc54 ("bpf: implement getsockopt and setsockopt hooks")
Reported-by: Martin KaFai Lau 
Signed-off-by: Stanislav Fomichev 
Signed-off-by: Daniel Borkmann 
Link: https://lore.kernel.org/bpf/ZD7Js4fj5YyI2oLd@google.com/T/#mb68daf700f87a9244a15d01d00c3f0e5b08f49f7
Link: https://lore.kernel.org/bpf/20230418225343.553806-2-sdf@google.com