linux-kernel - [for-linus][PATCH] tracing: Fix WARN_ON in tracing_buffers_mmap

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <20251125194151.3ca9b873@gandalf.local.home>
Date: Tue, 25 Nov 2025 19:41:51 -0500
From: Steven Rostedt <rostedt@...dmis.org>
To: LKML <linux-kernel@...r.kernel.org>
Cc: Masami Hiramatsu <mhiramat@...nel.org>, Mathieu Desnoyers
 <mathieu.desnoyers@...icios.com>, Deepanshu Kartikey
 <kartikey406@...il.com>
Subject: [for-linus][PATCH] tracing: Fix WARN_ON in
 tracing_buffers_mmap_close for split VMAs

ring-buffer fixes for v6.18:

- Do not allow mmapped ring buffer to be split

  When the ring buffer VMA is split by a partial munmap or a MAP_FIXED, the
  kernel calls vm_ops->close() on each portion. This causes the
  ring_buffer_unmap() to be called multiple times. This causes subsequent
  calls to return -ENODEV and triggers a warning.

  There's no reason to allow user space to split up memory mapping of the
  ring buffer. Have it return -EINVAL when that happens.

  git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace.git
ring-buffer/fixes

Head SHA1: b042fdf18e89a347177a49e795d8e5184778b5b6


Deepanshu Kartikey (1):
      tracing: Fix WARN_ON in tracing_buffers_mmap_close for split VMAs

----
 kernel/trace/trace.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
---------------------------
commit b042fdf18e89a347177a49e795d8e5184778b5b6
Author: Deepanshu Kartikey <kartikey406@...il.com>
Date:   Wed Nov 19 12:10:19 2025 +0530

    tracing: Fix WARN_ON in tracing_buffers_mmap_close for split VMAs
    
    When a VMA is split (e.g., by partial munmap or MAP_FIXED), the kernel
    calls vm_ops->close on each portion. For trace buffer mappings, this
    results in ring_buffer_unmap() being called multiple times while
    ring_buffer_map() was only called once.
    
    This causes ring_buffer_unmap() to return -ENODEV on subsequent calls
    because user_mapped is already 0, triggering a WARN_ON.
    
    Trace buffer mappings cannot support partial mappings because the ring
    buffer structure requires the complete buffer including the meta page.
    
    Fix this by adding a may_split callback that returns -EINVAL to prevent
    VMA splits entirely.
    
    Cc: stable@...r.kernel.org
    Fixes: cf9f0f7c4c5bb ("tracing: Allow user-space mapping of the ring-buffer")
    Link: https://patch.msgid.link/20251119064019.25904-1-kartikey406@gmail.com
    Closes: https://syzkaller.appspot.com/bug?extid=a72c325b042aae6403c7
    Tested-by: syzbot+a72c325b042aae6403c7@...kaller.appspotmail.com
    Reported-by: syzbot+a72c325b042aae6403c7@...kaller.appspotmail.com
    Signed-off-by: Deepanshu Kartikey <kartikey406@...il.com>
    Signed-off-by: Steven Rostedt (Google) <rostedt@...dmis.org>

diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index d1e527cf2aae..304e93597126 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -8781,8 +8781,18 @@ static void tracing_buffers_mmap_close(struct vm_area_struct *vma)
 	put_snapshot_map(iter->tr);
 }
 
+static int tracing_buffers_may_split(struct vm_area_struct *vma, unsigned long addr)
+{
+	/*
+	 * Trace buffer mappings require the complete buffer including
+	 * the meta page. Partial mappings are not supported.
+	 */
+	return -EINVAL;
+}
+
 static const struct vm_operations_struct tracing_buffers_vmops = {
 	.close		= tracing_buffers_mmap_close,
+	.may_split      = tracing_buffers_may_split,
 };
 
 static int tracing_buffers_mmap(struct file *filp, struct vm_area_struct *vma)