All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Jann Horn <jannh@google.com>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Alexei Starovoitov <ast@kernel.org>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.6 11/29] bpf: Fix tnum constraints for 32-bit comparisons
Date: Tue,  7 Apr 2020 12:22:08 +0200	[thread overview]
Message-ID: <20200407101453.370611188@linuxfoundation.org> (raw)
In-Reply-To: <20200407101452.046058399@linuxfoundation.org>

From: Jann Horn <jannh@google.com>

[ Upstream commit 604dca5e3af1db98bd123b7bfc02b017af99e3a0 ]

The BPF verifier tried to track values based on 32-bit comparisons by
(ab)using the tnum state via 581738a681b6 ("bpf: Provide better register
bounds after jmp32 instructions"). The idea is that after a check like
this:

    if ((u32)r0 > 3)
      exit

We can't meaningfully constrain the arithmetic-range-based tracking, but
we can update the tnum state to (value=0,mask=0xffff'ffff'0000'0003).
However, the implementation from 581738a681b6 didn't compute the tnum
constraint based on the fixed operand, but instead derives it from the
arithmetic-range-based tracking. This means that after the following
sequence of operations:

    if (r0 >= 0x1'0000'0001)
      exit
    if ((u32)r0 > 7)
      exit

The verifier assumed that the lower half of r0 is in the range (0, 0)
and apply the tnum constraint (value=0,mask=0xffff'ffff'0000'0000) thus
causing the overall tnum to be (value=0,mask=0x1'0000'0000), which was
incorrect. Provide a fixed implementation.

Fixes: 581738a681b6 ("bpf: Provide better register bounds after jmp32 instructions")
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200330160324.15259-3-daniel@iogearbox.net
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 kernel/bpf/verifier.c | 108 ++++++++++++++++++++++++++++--------------
 1 file changed, 72 insertions(+), 36 deletions(-)

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 5080469094afe..595b39eee6422 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -5590,6 +5590,70 @@ static bool cmp_val_with_extended_s64(s64 sval, struct bpf_reg_state *reg)
 		reg->smax_value <= 0 && reg->smin_value >= S32_MIN);
 }
 
+/* Constrain the possible values of @reg with unsigned upper bound @bound.
+ * If @is_exclusive, @bound is an exclusive limit, otherwise it is inclusive.
+ * If @is_jmp32, @bound is a 32-bit value that only constrains the low 32 bits
+ * of @reg.
+ */
+static void set_upper_bound(struct bpf_reg_state *reg, u64 bound, bool is_jmp32,
+			    bool is_exclusive)
+{
+	if (is_exclusive) {
+		/* There are no values for `reg` that make `reg<0` true. */
+		if (bound == 0)
+			return;
+		bound--;
+	}
+	if (is_jmp32) {
+		/* Constrain the register's value in the tnum representation.
+		 * For 64-bit comparisons this happens later in
+		 * __reg_bound_offset(), but for 32-bit comparisons, we can be
+		 * more precise than what can be derived from the updated
+		 * numeric bounds.
+		 */
+		struct tnum t = tnum_range(0, bound);
+
+		t.mask |= ~0xffffffffULL; /* upper half is unknown */
+		reg->var_off = tnum_intersect(reg->var_off, t);
+
+		/* Compute the 64-bit bound from the 32-bit bound. */
+		bound += gen_hi_max(reg->var_off);
+	}
+	reg->umax_value = min(reg->umax_value, bound);
+}
+
+/* Constrain the possible values of @reg with unsigned lower bound @bound.
+ * If @is_exclusive, @bound is an exclusive limit, otherwise it is inclusive.
+ * If @is_jmp32, @bound is a 32-bit value that only constrains the low 32 bits
+ * of @reg.
+ */
+static void set_lower_bound(struct bpf_reg_state *reg, u64 bound, bool is_jmp32,
+			    bool is_exclusive)
+{
+	if (is_exclusive) {
+		/* There are no values for `reg` that make `reg>MAX` true. */
+		if (bound == (is_jmp32 ? U32_MAX : U64_MAX))
+			return;
+		bound++;
+	}
+	if (is_jmp32) {
+		/* Constrain the register's value in the tnum representation.
+		 * For 64-bit comparisons this happens later in
+		 * __reg_bound_offset(), but for 32-bit comparisons, we can be
+		 * more precise than what can be derived from the updated
+		 * numeric bounds.
+		 */
+		struct tnum t = tnum_range(bound, U32_MAX);
+
+		t.mask |= ~0xffffffffULL; /* upper half is unknown */
+		reg->var_off = tnum_intersect(reg->var_off, t);
+
+		/* Compute the 64-bit bound from the 32-bit bound. */
+		bound += gen_hi_min(reg->var_off);
+	}
+	reg->umin_value = max(reg->umin_value, bound);
+}
+
 /* Adjusts the register min/max values in the case that the dst_reg is the
  * variable register that we are working on, and src_reg is a constant or we're
  * simply doing a BPF_K check.
@@ -5645,15 +5709,8 @@ static void reg_set_min_max(struct bpf_reg_state *true_reg,
 	case BPF_JGE:
 	case BPF_JGT:
 	{
-		u64 false_umax = opcode == BPF_JGT ? val    : val - 1;
-		u64 true_umin = opcode == BPF_JGT ? val + 1 : val;
-
-		if (is_jmp32) {
-			false_umax += gen_hi_max(false_reg->var_off);
-			true_umin += gen_hi_min(true_reg->var_off);
-		}
-		false_reg->umax_value = min(false_reg->umax_value, false_umax);
-		true_reg->umin_value = max(true_reg->umin_value, true_umin);
+		set_upper_bound(false_reg, val, is_jmp32, opcode == BPF_JGE);
+		set_lower_bound(true_reg, val, is_jmp32, opcode == BPF_JGT);
 		break;
 	}
 	case BPF_JSGE:
@@ -5674,15 +5731,8 @@ static void reg_set_min_max(struct bpf_reg_state *true_reg,
 	case BPF_JLE:
 	case BPF_JLT:
 	{
-		u64 false_umin = opcode == BPF_JLT ? val    : val + 1;
-		u64 true_umax = opcode == BPF_JLT ? val - 1 : val;
-
-		if (is_jmp32) {
-			false_umin += gen_hi_min(false_reg->var_off);
-			true_umax += gen_hi_max(true_reg->var_off);
-		}
-		false_reg->umin_value = max(false_reg->umin_value, false_umin);
-		true_reg->umax_value = min(true_reg->umax_value, true_umax);
+		set_lower_bound(false_reg, val, is_jmp32, opcode == BPF_JLE);
+		set_upper_bound(true_reg, val, is_jmp32, opcode == BPF_JLT);
 		break;
 	}
 	case BPF_JSLE:
@@ -5757,15 +5807,8 @@ static void reg_set_min_max_inv(struct bpf_reg_state *true_reg,
 	case BPF_JGE:
 	case BPF_JGT:
 	{
-		u64 false_umin = opcode == BPF_JGT ? val    : val + 1;
-		u64 true_umax = opcode == BPF_JGT ? val - 1 : val;
-
-		if (is_jmp32) {
-			false_umin += gen_hi_min(false_reg->var_off);
-			true_umax += gen_hi_max(true_reg->var_off);
-		}
-		false_reg->umin_value = max(false_reg->umin_value, false_umin);
-		true_reg->umax_value = min(true_reg->umax_value, true_umax);
+		set_lower_bound(false_reg, val, is_jmp32, opcode == BPF_JGE);
+		set_upper_bound(true_reg, val, is_jmp32, opcode == BPF_JGT);
 		break;
 	}
 	case BPF_JSGE:
@@ -5783,15 +5826,8 @@ static void reg_set_min_max_inv(struct bpf_reg_state *true_reg,
 	case BPF_JLE:
 	case BPF_JLT:
 	{
-		u64 false_umax = opcode == BPF_JLT ? val    : val - 1;
-		u64 true_umin = opcode == BPF_JLT ? val + 1 : val;
-
-		if (is_jmp32) {
-			false_umax += gen_hi_max(false_reg->var_off);
-			true_umin += gen_hi_min(true_reg->var_off);
-		}
-		false_reg->umax_value = min(false_reg->umax_value, false_umax);
-		true_reg->umin_value = max(true_reg->umin_value, true_umin);
+		set_upper_bound(false_reg, val, is_jmp32, opcode == BPF_JLE);
+		set_lower_bound(true_reg, val, is_jmp32, opcode == BPF_JLT);
 		break;
 	}
 	case BPF_JSLE:
-- 
2.20.1




  parent reply	other threads:[~2020-04-07 10:28 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-07 10:21 [PATCH 5.6 00/29] 5.6.3-rc1 review Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.6 01/29] ipv4: fix a RCU-list lock in fib_triestat_seq_show Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.6 02/29] net: dsa: ksz: Select KSZ protocol tag Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 03/29] net, ip_tunnel: fix interface lookup with no key Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 04/29] sctp: fix possibly using a bad saddr with a given dst Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 05/29] sctp: fix refcount bug in sctp_wfree Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 06/29] net: macb: Fix handling of fixed-link node Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 07/29] net: fix fraglist segmentation reference count leak Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 08/29] udp: initialize is_flist with 0 in udp_gro_receive Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 09/29] padata: fix uninitialized return value in padata_replace() Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 10/29] brcmfmac: abort and release host after error Greg Kroah-Hartman
2020-04-07 10:22 ` Greg Kroah-Hartman [this message]
2020-04-07 10:22 ` [PATCH 5.6 12/29] XArray: Fix xa_find_next for large multi-index entries Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 13/29] drm/bridge: analogix-anx6345: Avoid duplicate -supply suffix Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 14/29] drm/i915/display: Fix mode private_flags comparison at atomic_check Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 15/29] misc: rtsx: set correct pcr_ops for rts522A Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 16/29] misc: pci_endpoint_test: Fix to support > 10 pci-endpoint-test devices Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 17/29] misc: pci_endpoint_test: Avoid using module parameter to determine irqtype Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 18/29] PCI: sysfs: Revert "rescan" file renames Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 19/29] coresight: do not use the BIT() macro in the UAPI header Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 20/29] mei: me: add cedar fork device ids Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 21/29] nvmem: release the write-protect pin Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 22/29] nvmem: check for NULL reg_read and reg_write before dereferencing Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 23/29] nvmem: sprd: Fix the block lock operation Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 24/29] extcon: axp288: Add wakeup support Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 25/29] power: supply: axp288_charger: Add special handling for HP Pavilion x2 10 Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 26/29] Revert "ALSA: uapi: Drop asound.h inclusion from asoc.h" Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 27/29] Revert "dm: always call blk_queue_split() in dm_process_bio()" Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 28/29] ALSA: hda/ca0132 - Add Recon3Di quirk to handle integrated sound on EVGA X99 Classified motherboard Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.6 29/29] soc: mediatek: knows_txdone needs to be set in Mediatek CMDQ helper Greg Kroah-Hartman
2020-04-07 12:37 ` [PATCH 5.6 00/29] 5.6.3-rc1 review Jon Hunter
2020-04-07 12:37   ` Jon Hunter
     [not found]   ` <dd65ddb4-478c-c022-542c-5e0b44ab8962-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
2020-04-07 14:49     ` Greg Kroah-Hartman
2020-04-07 14:49       ` Greg Kroah-Hartman
2020-04-07 13:51 ` Daniel Díaz
2020-04-07 14:43   ` Greg Kroah-Hartman

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=20200407101453.370611188@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=ast@kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=jannh@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sashal@kernel.org \
    --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.