All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	patches@lists.linux.dev, Lonial Con <kongln9170@gmail.com>,
	Eduard Zingerman <eddyz87@gmail.com>,
	Andrii Nakryiko <andrii@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Shung-Hsi Yu <shung-hsi.yu@suse.com>
Subject: [PATCH 5.10 30/43] bpf: sync_linked_regs() must preserve subreg_def
Date: Tue, 17 Dec 2024 18:07:21 +0100	[thread overview]
Message-ID: <20241217170521.666678246@linuxfoundation.org> (raw)
In-Reply-To: <20241217170520.459491270@linuxfoundation.org>

5.10-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Eduard Zingerman <eddyz87@gmail.com>

commit e9bd9c498cb0f5843996dbe5cbce7a1836a83c70 upstream.

Range propagation must not affect subreg_def marks, otherwise the
following example is rewritten by verifier incorrectly when
BPF_F_TEST_RND_HI32 flag is set:

  0: call bpf_ktime_get_ns                   call bpf_ktime_get_ns
  1: r0 &= 0x7fffffff       after verifier   r0 &= 0x7fffffff
  2: w1 = w0                rewrites         w1 = w0
  3: if w0 < 10 goto +0     -------------->  r11 = 0x2f5674a6     (r)
  4: r1 >>= 32                               r11 <<= 32           (r)
  5: r0 = r1                                 r1 |= r11            (r)
  6: exit;                                   if w0 < 0xa goto pc+0
                                             r1 >>= 32
                                             r0 = r1
                                             exit

(or zero extension of w1 at (2) is missing for architectures that
 require zero extension for upper register half).

The following happens w/o this patch:
- r0 is marked as not a subreg at (0);
- w1 is marked as subreg at (2);
- w1 subreg_def is overridden at (3) by copy_register_state();
- w1 is read at (5) but mark_insn_zext() does not mark (2)
  for zero extension, because w1 subreg_def is not set;
- because of BPF_F_TEST_RND_HI32 flag verifier inserts random
  value for hi32 bits of (2) (marked (r));
- this random value is read at (5).

Fixes: 75748837b7e5 ("bpf: Propagate scalar ranges through register assignments.")
Reported-by: Lonial Con <kongln9170@gmail.com>
Signed-off-by: Lonial Con <kongln9170@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Closes: https://lore.kernel.org/bpf/7e2aa30a62d740db182c170fdd8f81c596df280d.camel@gmail.com
Link: https://lore.kernel.org/bpf/20240924210844.1758441-1-eddyz87@gmail.com
[ shung-hsi.yu: sync_linked_regs() was called find_equal_scalars() before commit
  4bf79f9be434 ("bpf: Track equal scalars history on per-instruction level"), and
  modification is done because there is only a single call to
  copy_register_state() before commit 98d7ca374ba4 ("bpf: Track delta between
  "linked" registers."). ]
Signed-off-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 kernel/bpf/verifier.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -8168,8 +8168,11 @@ static void find_equal_scalars(struct bp
 	struct bpf_reg_state *reg;
 
 	bpf_for_each_reg_in_vstate(vstate, state, reg, ({
-		if (reg->type == SCALAR_VALUE && reg->id == known_reg->id)
+		if (reg->type == SCALAR_VALUE && reg->id == known_reg->id) {
+			s32 saved_subreg_def = reg->subreg_def;
 			copy_register_state(reg, known_reg);
+			reg->subreg_def = saved_subreg_def;
+		}
 	}));
 }
 



  parent reply	other threads:[~2024-12-17 17:12 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-12-17 17:06 [PATCH 5.10 00/43] 5.10.232-rc1 review Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 01/43] tcp: check space before adding MPTCP SYN options Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 02/43] usb: host: max3421-hcd: Correctly abort a USB request Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 03/43] ata: sata_highbank: fix OF node reference leak in highbank_initialize_phys() Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 04/43] usb: dwc2: hcd: Fix GetPortStatus & SetPortFeature Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 05/43] usb: ehci-hcd: fix call balance of clocks handling routines Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 06/43] usb: gadget: u_serial: Fix the issue that gs_start_io crashed due to accessing null pointer Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 07/43] xfs: dont drop errno values when we fail to ficlone the entire range Greg Kroah-Hartman
2024-12-17 17:06 ` [PATCH 5.10 08/43] xfs: fix scrub tracepoints when inode-rooted btrees are involved Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 09/43] bpf, sockmap: Fix update element with same Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 10/43] virtio/vsock: Fix accept_queue memory leak Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 11/43] exfat: fix potential deadlock on __exfat_get_dentry_set Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 12/43] acpi: nfit: vmalloc-out-of-bounds Read in acpi_nfit_ctl Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 13/43] batman-adv: Do not send uninitialized TT changes Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 14/43] batman-adv: Remove uninitialized data in full table TT response Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 15/43] batman-adv: Do not let TT changes list grows indefinitely Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 16/43] tipc: fix NULL deref in cleanup_bearer() Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 17/43] selftests: mlxsw: sharedbuffer: Remove h1 ingress test case Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 18/43] selftests: mlxsw: sharedbuffer: Remove duplicate test cases Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 19/43] net: lapb: increase LAPB_HEADER_LEN Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 20/43] ACPI: resource: Fix memory resource type union access Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 21/43] cxgb4: use port number to set mac addr Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 22/43] qca_spi: Fix clock speed for multiple QCA7000 Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 23/43] qca_spi: Make driver probing reliable Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 24/43] net/sched: netem: account for backlog updates from child qdisc Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 25/43] net: bonding, dummy, ifb, team: advertise NETIF_F_GSO_SOFTWARE Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 26/43] bonding: Fix feature propagation of NETIF_F_GSO_ENCAP_ALL Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 27/43] team: " Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 28/43] ACPICA: events/evxfregn: dont release the ContextMutex that was never acquired Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 29/43] blk-iocost: Avoid using clamp() on inuse in __propagate_weights() Greg Kroah-Hartman
2024-12-17 17:07 ` Greg Kroah-Hartman [this message]
2024-12-17 17:07 ` [PATCH 5.10 31/43] tracing/kprobes: Skip symbol counting logic for module symbols in create_local_trace_kprobe() Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 32/43] drm/i915: Fix memory leak by correcting cache object name in error handler Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 33/43] Revert "clocksource/drivers:sp804: Make user selectable" Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 34/43] Revert "clkdev: remove CONFIG_CLKDEV_LOOKUP" Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 35/43] xen/netfront: fix crash when removing device Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 36/43] x86: make get_cpu_vendor() accessible from Xen code Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 37/43] objtool/x86: allow syscall instruction Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 38/43] x86/static-call: provide a way to do very early static-call updates Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 39/43] x86/xen: dont do PV iret hypercall through hypercall page Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 40/43] x86/xen: add central hypercall functions Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 41/43] x86/xen: use new hypercall functions instead of hypercall page Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 42/43] x86/xen: remove " Greg Kroah-Hartman
2024-12-17 17:07 ` [PATCH 5.10 43/43] ALSA: usb-audio: Fix a DMA to stack memory bug Greg Kroah-Hartman
2024-12-17 18:51 ` [PATCH 5.10 00/43] 5.10.232-rc1 review Florian Fainelli
2024-12-17 20:37 ` Pavel Machek
2024-12-18  9:17 ` Dominique Martinet
2024-12-18 12:38 ` Mark Brown
2024-12-18 17:20 ` Jon Hunter
2024-12-18 17:55 ` Naresh Kamboju

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=20241217170521.666678246@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=andrii@kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=eddyz87@gmail.com \
    --cc=kongln9170@gmail.com \
    --cc=patches@lists.linux.dev \
    --cc=shung-hsi.yu@suse.com \
    --cc=stable@vger.kernel.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.