linux-stable.git/kernel, branch v6.6.92

ftrace: Fix preemption accounting for stacktrace filter command

2025-05-22T12:12:21+00:00

commit 11aff32439df6ca5b3b891b43032faf88f4a6a29 upstream.

The preemption count of the stacktrace filter command to trace ksys_read
is consistently incorrect:

$ echo ksys_read:stacktrace > set_ftrace_filter

   <...>-453     [004] ...1.    38.308956: 
=> ksys_read
=> do_syscall_64
=> entry_SYSCALL_64_after_hwframe

The root cause is that the trace framework disables preemption when
invoking the filter command callback in function_trace_probe_call:

   preempt_disable_notrace();
   probe_ops->func(ip, parent_ip, probe_opsbe->tr, probe_ops, probe->data);
   preempt_enable_notrace();

Use tracing_gen_ctx_dec() to account for the preempt_disable_notrace(),
which will output the correct preemption count:

$ echo ksys_read:stacktrace > set_ftrace_filter

   <...>-410     [006] .....    31.420396: 
=> ksys_read
=> do_syscall_64
=> entry_SYSCALL_64_after_hwframe

Cc: stable@vger.kernel.org
Fixes: 36590c50b2d07 ("tracing: Merge irqflags + preempt counter.")
Link: https://lore.kernel.org/20250512094246.1167956-2-dolinux.peng@gmail.com
Signed-off-by: pengdonglin 
Signed-off-by: Steven Rostedt (Google) 
Signed-off-by: Greg Kroah-Hartman

ftrace: Fix preemption accounting for stacktrace trigger command

2025-05-22T12:12:21+00:00

commit e333332657f615ac2b55aa35565c4a882018bbe9 upstream.

When using the stacktrace trigger command to trace syscalls, the
preemption count was consistently reported as 1 when the system call
event itself had 0 (".").

For example:

root@ubuntu22-vm:/sys/kernel/tracing/events/syscalls/sys_enter_read
$ echo stacktrace > trigger
$ echo 1 > enable

    sshd-416     [002] .....   232.864910: sys_read(fd: a, buf: 556b1f3221d0, count: 8000)
    sshd-416     [002] ...1.   232.864913: 
 => ftrace_syscall_enter
 => syscall_trace_enter
 => do_syscall_64
 => entry_SYSCALL_64_after_hwframe

The root cause is that the trace framework disables preemption in __DO_TRACE before
invoking the trigger callback.

Use the tracing_gen_ctx_dec() that will accommodate for the increase of
the preemption count in __DO_TRACE when calling the callback. The result
is the accurate reporting of:

    sshd-410     [004] .....   210.117660: sys_read(fd: 4, buf: 559b725ba130, count: 40000)
    sshd-410     [004] .....   210.117662: 
 => ftrace_syscall_enter
 => syscall_trace_enter
 => do_syscall_64
 => entry_SYSCALL_64_after_hwframe

Cc: stable@vger.kernel.org
Fixes: ce33c845b030c ("tracing: Dump stacktrace trigger to the corresponding instance")
Link: https://lore.kernel.org/20250512094246.1167956-1-dolinux.peng@gmail.com
Signed-off-by: pengdonglin 
Signed-off-by: Steven Rostedt (Google) 
Signed-off-by: Greg Kroah-Hartman

tracing: probes: Fix a possible race in trace_probe_log APIs

2025-05-22T12:12:12+00:00

[ Upstream commit fd837de3c9cb1a162c69bc1fb1f438467fe7f2f5 ]

Since the shared trace_probe_log variable can be accessed and
modified via probe event create operation of kprobe_events,
uprobe_events, and dynamic_events, it should be protected.
In the dynamic_events, all operations are serialized by
`dyn_event_ops_mutex`. But kprobe_events and uprobe_events
interfaces are not serialized.

To solve this issue, introduces dyn_event_create(), which runs
create() operation under the mutex, for kprobe_events and
uprobe_events. This also uses lockdep to check the mutex is
held when using trace_probe_log* APIs.

Link: https://lore.kernel.org/all/174684868120.551552.3068655787654268804.stgit@devnote2/

Reported-by: Paul Cacheux 
Closes: https://lore.kernel.org/all/20250510074456.805a16872b591e2971a4d221@kernel.org/
Fixes: ab105a4fb894 ("tracing: Use tracing error_log with probe events")
Signed-off-by: Masami Hiramatsu (Google) 
Signed-off-by: Sasha Levin

cgroup/cpuset: Extend kthread_is_per_cpu() check to all PF_NO_SETAFFINITY tasks

2025-05-22T12:12:12+00:00

[ Upstream commit 39b5ef791d109dd54c7c2e6e87933edfcc0ad1ac ]

Commit ec5fbdfb99d1 ("cgroup/cpuset: Enable update_tasks_cpumask()
on top_cpuset") enabled us to pull CPUs dedicated to child partitions
from tasks in top_cpuset by ignoring per cpu kthreads. However, there
can be other kthreads that are not per cpu but have PF_NO_SETAFFINITY
flag set to indicate that we shouldn't mess with their CPU affinity.
For other kthreads, their affinity will be changed to skip CPUs dedicated
to child partitions whether it is an isolating or a scheduling one.

As all the per cpu kthreads have PF_NO_SETAFFINITY set, the
PF_NO_SETAFFINITY tasks are essentially a superset of per cpu kthreads.
Fix this issue by dropping the kthread_is_per_cpu() check and checking
the PF_NO_SETAFFINITY flag instead.

Fixes: ec5fbdfb99d1 ("cgroup/cpuset: Enable update_tasks_cpumask() on top_cpuset")
Signed-off-by: Waiman Long 
Acked-by: Frederic Weisbecker 
Signed-off-by: Tejun Heo 
Signed-off-by: Sasha Levin

module: ensure that kobject_put() is safe for module type kobjects

2025-05-18T06:24:08+00:00

commit a6aeb739974ec73e5217c75a7c008a688d3d5cf1 upstream.

In 'lookup_or_create_module_kobject()', an internal kobject is created
using 'module_ktype'. So call to 'kobject_put()' on error handling
path causes an attempt to use an uninitialized completion pointer in
'module_kobject_release()'. In this scenario, we just want to release
kobject without an extra synchronization required for a regular module
unloading process, so adding an extra check whether 'complete()' is
actually required makes 'kobject_put()' safe.

Reported-by: syzbot+7fb8a372e1f6add936dd@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=7fb8a372e1f6add936dd
Fixes: 942e443127e9 ("module: Fix mod->mkobj.kobj potentially freed too early")
Cc: stable@vger.kernel.org
Suggested-by: Petr Pavlu 
Signed-off-by: Dmitry Antipov 
Link: https://lore.kernel.org/r/20250507065044.86529-1-dmantipov@yandex.ru
Signed-off-by: Petr Pavlu 
Signed-off-by: Greg Kroah-Hartman

kernel: globalize lookup_or_create_module_kobject()

2025-05-09T07:44:07+00:00

[ Upstream commit 7c76c813cfc42a7376378a0c4b7250db2eebab81 ]

lookup_or_create_module_kobject() is marked as static and __init,
to make it global drop static keyword.
Since this function can be called from non-init code, use __modinit
instead of __init, __modinit marker will make it __init if
CONFIG_MODULES is not defined.

Suggested-by: Rasmus Villemoes 
Signed-off-by: Shyam Saini 
Link: https://lore.kernel.org/r/20250227184930.34163-4-shyamsaini@linux.microsoft.com
Signed-off-by: Petr Pavlu 
Stable-dep-of: f95bbfe18512 ("drivers: base: handle module_kobject creation")
Signed-off-by: Sasha Levin

kernel: param: rename locate_module_kobject

2025-05-09T07:44:06+00:00

[ Upstream commit bbc9462f0cb0c8917a4908e856731708f0cee910 ]

The locate_module_kobject() function looks up an existing
module_kobject for a given module name. If it cannot find the
corresponding module_kobject, it creates one for the given name.

This commit renames locate_module_kobject() to
lookup_or_create_module_kobject() to better describe its operations.

This doesn't change anything functionality wise.

Suggested-by: Rasmus Villemoes 
Signed-off-by: Shyam Saini 
Link: https://lore.kernel.org/r/20250227184930.34163-2-shyamsaini@linux.microsoft.com
Signed-off-by: Petr Pavlu 
Stable-dep-of: f95bbfe18512 ("drivers: base: handle module_kobject creation")
Signed-off-by: Sasha Levin

bpf: fix null dereference when computing changes_pkt_data of prog w/o subprogs

2025-05-09T07:43:55+00:00

commit ac6542ad92759cda383ad62b4e4cbfc28136abc1 upstream.

bpf_prog_aux->func field might be NULL if program does not have
subprograms except for main sub-program. The fixed commit does
bpf_prog_aux->func access unconditionally, which might lead to null
pointer dereference.

The bug could be triggered by replacing the following BPF program:

    SEC("tc")
    int main_changes(struct __sk_buff *sk)
    {
        bpf_skb_pull_data(sk, 0);
        return 0;
    }

With the following BPF program:

    SEC("freplace")
    long changes_pkt_data(struct __sk_buff *sk)
    {
        return bpf_skb_pull_data(sk, 0);
    }

bpf_prog_aux instance itself represents the main sub-program,
use this property to fix the bug.

Fixes: 81f6d0530ba0 ("bpf: check changes_pkt_data property for extension programs")
Reported-by: kernel test robot 
Reported-by: Dan Carpenter 
Closes: https://lore.kernel.org/r/202412111822.qGw6tOyB-lkp@intel.com/
Signed-off-by: Eduard Zingerman 
Link: https://lore.kernel.org/r/20241212070711.427443-1-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov 
Signed-off-by: Shung-Hsi Yu 
Signed-off-by: Greg Kroah-Hartman

bpf: check changes_pkt_data property for extension programs

2025-05-09T07:43:54+00:00

commit 81f6d0530ba031b5f038a091619bf2ff29568852 upstream.

When processing calls to global sub-programs, verifier decides whether
to invalidate all packet pointers in current state depending on the
changes_pkt_data property of the global sub-program.

Because of this, an extension program replacing a global sub-program
must be compatible with changes_pkt_data property of the sub-program
being replaced.

This commit:
- adds changes_pkt_data flag to struct bpf_prog_aux:
  - this flag is set in check_cfg() for main sub-program;
  - in jit_subprogs() for other sub-programs;
- modifies bpf_check_attach_btf_id() to check changes_pkt_data flag;
- moves call to check_attach_btf_id() after the call to check_cfg(),
  because it needs changes_pkt_data flag to be set:

    bpf_check:
      ...                             ...
    - check_attach_btf_id             resolve_pseudo_ldimm64
      resolve_pseudo_ldimm64   -->    bpf_prog_is_offloaded
      bpf_prog_is_offloaded           check_cfg
      check_cfg                     + check_attach_btf_id
      ...                             ...

The following fields are set by check_attach_btf_id():
- env->ops
- prog->aux->attach_btf_trace
- prog->aux->attach_func_name
- prog->aux->attach_func_proto
- prog->aux->dst_trampoline
- prog->aux->mod
- prog->aux->saved_dst_attach_type
- prog->aux->saved_dst_prog_type
- prog->expected_attach_type

Neither of these fields are used by resolve_pseudo_ldimm64() or
bpf_prog_offload_verifier_prep() (for netronome and netdevsim
drivers), so the reordering is safe.

Suggested-by: Alexei Starovoitov 
Signed-off-by: Eduard Zingerman 
Link: https://lore.kernel.org/r/20241210041100.1898468-6-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov 
[ shung-hsi.yu: adapt to missing fields in "struct bpf_prog_aux". Context
difference in jit_subprogs() because BPF Exception is not supported. Context
difference in bpf_check() because commit 5b5f51bff1b6 "bpf:
no_caller_saved_registers attribute for helper calls" is not present. ]
Signed-off-by: Shung-Hsi Yu 
Signed-off-by: Greg Kroah-Hartman

bpf: track changes_pkt_data property for global functions

2025-05-09T07:43:54+00:00

commit 51081a3f25c742da5a659d7fc6fd77ebfdd555be upstream.

When processing calls to certain helpers, verifier invalidates all
packet pointers in a current state. For example, consider the
following program:

    __attribute__((__noinline__))
    long skb_pull_data(struct __sk_buff *sk, __u32 len)
    {
        return bpf_skb_pull_data(sk, len);
    }

    SEC("tc")
    int test_invalidate_checks(struct __sk_buff *sk)
    {
        int *p = (void *)(long)sk->data;
        if ((void *)(p + 1) > (void *)(long)sk->data_end) return TCX_DROP;
        skb_pull_data(sk, 0);
        *p = 42;
        return TCX_PASS;
    }

After a call to bpf_skb_pull_data() the pointer 'p' can't be used
safely. See function filter.c:bpf_helper_changes_pkt_data() for a list
of such helpers.

At the moment verifier invalidates packet pointers when processing
helper function calls, and does not traverse global sub-programs when
processing calls to global sub-programs. This means that calls to
helpers done from global sub-programs do not invalidate pointers in
the caller state. E.g. the program above is unsafe, but is not
rejected by verifier.

This commit fixes the omission by computing field
bpf_subprog_info->changes_pkt_data for each sub-program before main
verification pass.
changes_pkt_data should be set if:
- subprogram calls helper for which bpf_helper_changes_pkt_data
  returns true;
- subprogram calls a global function,
  for which bpf_subprog_info->changes_pkt_data should be set.

The verifier.c:check_cfg() pass is modified to compute this
information. The commit relies on depth first instruction traversal
done by check_cfg() and absence of recursive function calls:
- check_cfg() would eventually visit every call to subprogram S in a
  state when S is fully explored;
- when S is fully explored:
  - every direct helper call within S is explored
    (and thus changes_pkt_data is set if needed);
  - every call to subprogram S1 called by S was visited with S1 fully
    explored (and thus S inherits changes_pkt_data from S1).

The downside of such approach is that dead code elimination is not
taken into account: if a helper call inside global function is dead
because of current configuration, verifier would conservatively assume
that the call occurs for the purpose of the changes_pkt_data
computation.

Reported-by: Nick Zavaritsky 
Closes: https://lore.kernel.org/bpf/0498CA22-5779-4767-9C0C-A9515CEA711F@gmail.com/
Signed-off-by: Eduard Zingerman 
Link: https://lore.kernel.org/r/20241210041100.1898468-4-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov 
[shung-hsi.yu: do not use bitfield in "struct bpf_subprog_info" because commit
406a6fa44bfb ("bpf: use bitfields for simple per-subprog bool flags") is not
present and minor context difference in check_func_call() because commit
491dd8edecbc ("bpf: Emit global subprog name in verifier logs") is not present. ]
Signed-off-by: Shung-Hsi Yu 
Signed-off-by: Greg Kroah-Hartman