Netdev List
 help / color / mirror / Atom feed
From: Rongguang Wei <clementwei90@163.com>
To: netdev@vger.kernel.org
Cc: willemdebruijn.kernel@gmail.com, jasowangio@gmail.com,
	andrew+netdev@lunn.ch, davem@davemloft.net, kuba@kernel.org,
	Rongguang Wei <weirongguang@kylinos.cn>
Subject: [PATCH net v3 0/3] tun: fix re-attaching the socket filter
Date: Wed, 30 Sep 2026 16:07:43 +0800	[thread overview]
Message-ID: <20260930080746.135017-1-clementwei90@163.com> (raw)

From: Rongguang Wei <weirongguang@kylinos.cn>

This is v3 of the series that fixes attaching a queue to a TAP device which
re-installs the socket filter of a persistent device.

Patch 1 keeps a copy of the program in the kernel when it is configured,
instead of reading tun->fprog from user space again on every later attach.
It also adds sk_attach_filter_kern(), the kernel memory counterpart of
sk_attach_filter(), which the patch uses. That is the second issue reported
in the review of v1; IFF_NOFILTER keeps working and is no longer needed as
a workaround.

Patch 2 fixes the inverted error check after the filter attach in
tun_attach(): a successful re-attach returned early, so the queue was never
published in tun->tfiles[] while TUNSETIFF still reported success. A failed
re-attach now aborts the attach, and the filter is detached again if a
later step of the attach fails.

It has to follow patch 1: with the check fixed but the program still read
from the caller's address space, an attach that used to succeed without a
filter would fail with -EFAULT or -EINVAL.

Patch 3 adds selftests for the re-attach path and for attaching a queue
without a filter. The two tests that cover the kernel copy fail without
patch 1; the rollback in patch 2 needs a failure inside tun_attach() after
the filter attach, which userspace cannot trigger reliably, so it has no
test.

Thanks to Willem de Bruijn for the review.

Rongguang Wei (3):
  tun: keep a kernel copy of the socket filter program
  tun: fix inverted error check when re-attaching the filter
  selftests: net: add TAP socket filter attach tests

---
v3:
  - reorder: the kernel copy of the program comes before the corrected
    error check, so that no step of the series turns an attach that
    succeeded without a filter into a failure
  - add sk_attach_filter_kern() in the patch that first uses it
  - address the review of the selftests
v2:
  - roll back the filter attach when a later step of tun_attach() fails
  - keep a copy of the program in the kernel, so that a later attach does
    not depend on the address space of the process that set the filter
  - add selftests for the re-attach and for a rejected TUNATTACHFILTER
v1:
  - https://lore.kernel.org/netdev/20260923025653.59348-1-clementwei90@163.com/
---
 drivers/net/tun.c                 |  50 +++++++++-
 include/linux/filter.h            |   1 +
 net/core/filter.c                 |  22 +++++
 tools/testing/selftests/net/tun.c | 155 ++++++++++++++++++++++++++++++
 4 files changed, 223 insertions(+), 5 deletions(-)


base-commit: 72d3fcf802c45d00b300f25b848a93c3a2bd7c7e
-- 
2.43.0

No virus found
		Checked by Hillstone Network AntiVirus


             reply	other threads:[~2026-09-30  8:08 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-30  8:07 Rongguang Wei [this message]
2026-09-30  8:07 ` [PATCH net v3 1/3] tun: keep a kernel copy of the socket filter program Rongguang Wei
2026-09-30 18:33   ` Willem de Bruijn
2026-10-02  3:25     ` Rongguang Wei
2026-10-01 14:10   ` netdev-bot+sashiko
2026-10-02  3:13     ` Rongguang Wei
2026-09-30  8:07 ` [PATCH net v3 2/3] tun: fix inverted error check when re-attaching the filter Rongguang Wei
2026-09-30 18:34   ` Willem de Bruijn
2026-10-01 14:10   ` netdev-bot+sashiko
2026-10-02  3:17     ` Rongguang Wei
2026-09-30  8:07 ` [PATCH net v3 3/3] selftests: net: add TAP socket filter attach tests Rongguang Wei
2026-09-30 19:20   ` Willem de Bruijn
2026-10-02  3:20     ` Rongguang Wei
2026-10-01 14:10   ` netdev-bot+sashiko
2026-09-30  8:13 ` [PATCH net v3 0/3] tun: fix re-attaching the socket filter netdev-bot+sinfo

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260930080746.135017-1-clementwei90@163.com \
    --to=clementwei90@163.com \
    --cc=andrew+netdev@lunn.ch \
    --cc=davem@davemloft.net \
    --cc=jasowangio@gmail.com \
    --cc=kuba@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=weirongguang@kylinos.cn \
    --cc=willemdebruijn.kernel@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox