[PATCH stable 6.6 0/6] bpf: Fix bounds when ranges cross sign boundary

[PATCH stable 6.6 0/6] bpf: Fix bounds when ranges cross sign boundary

[Paul Chaignon]

As discussed in [1] yesterday, this series backports two sets of fixes
for BPF, with their selftests:
- 00bf8d0c6c9b ("bpf: Improve bounds when s64 crosses sign boundary")
- 26e5e346a52c ("selftests/bpf: Test cross-sign 64bits range
  refinement")
- f96841bbf4a1 ("selftests/bpf: Test invariants on JSLT crossing sign")
- 5dbb19b16ac4 ("bpf: Add third round of bounds deduction")
- fbc7aef517d8 ("bpf: Fix u32/s32 bounds when ranges cross min/max
  boundary")
- f81fdfd16771 ("selftests/bpf: test refining u32/s32 bounds when
  ranges cross min/max boundary")

Using Shung-Hsi's stable CI repo [2], I verified the BPF selftests pass
with these commits applied on top of v6.12.

1: https://lore.kernel.org/stable/2026040240-friday-gurgling-7088@gregkh/
2: https://github.com/pchaigno/stable-bpf-ci/actions/runs/23940850516/job/69826632354

Eduard Zingerman (2):
  bpf: Fix u32/s32 bounds when ranges cross min/max boundary
  selftests/bpf: test refining u32/s32 bounds when ranges cross min/max
    boundary

Paul Chaignon (4):
  bpf: Improve bounds when s64 crosses sign boundary
  selftests/bpf: Test cross-sign 64bits range refinement
  selftests/bpf: Test invariants on JSLT crossing sign
  bpf: Add third round of bounds deduction

 kernel/bpf/verifier.c                         |  77 +++++++++
 .../selftests/bpf/prog_tests/reg_bounds.c     |  62 ++++++-
 .../selftests/bpf/progs/verifier_bounds.c     | 159 +++++++++++++++++-
 3 files changed, 292 insertions(+), 6 deletions(-)

-- 
2.43.0

[PATCH stable 6.6 1/6] bpf: Improve bounds when s64 crosses sign boundary

[Paul Chaignon]

[ Upstream commit 00bf8d0c6c9be0c481fc45a3f7d87c7f8812f229 ]

__reg64_deduce_bounds currently improves the s64 range using the u64
range and vice versa, but only if it doesn't cross the sign boundary.

This patch improves __reg64_deduce_bounds to cover the case where the
s64 range crosses the sign boundary but overlaps with the u64 range on
only one end. In that case, we can improve both ranges. Consider the
following example, with the s64 range crossing the sign boundary:

    0                                                   U64_MAX
    |  [xxxxxxxxxxxxxx u64 range xxxxxxxxxxxxxx]              |
    |----------------------------|----------------------------|
    |xxxxx s64 range xxxxxxxxx]                       [xxxxxxx|
    0                     S64_MAX S64_MIN                    -1

The u64 range overlaps only with positive portion of the s64 range. We
can thus derive the following new s64 and u64 ranges.

    0                                                   U64_MAX
    |  [xxxxxx u64 range xxxxx]                               |
    |----------------------------|----------------------------|
    |  [xxxxxx s64 range xxxxx]                               |
    0                     S64_MAX S64_MIN                    -1

The same logic can probably apply to the s32/u32 ranges, but this patch
doesn't implement that change.

In addition to the selftests, the __reg64_deduce_bounds change was
also tested with Agni, the formal verification tool for the range
analysis [1].

Link: https://github.com/bpfverif/agni [1]
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
Link: https://lore.kernel.org/r/933bd9ce1f36ded5559f92fdc09e5dbc823fa245.1753695655.git.paul.chaignon@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
---
 kernel/bpf/verifier.c | 52 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 52 insertions(+)

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 68fa30852051..6448f9eeede0 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -2129,6 +2129,58 @@ static void __reg64_deduce_bounds(struct bpf_reg_state *reg)
 	if ((u64)reg->smin_value <= (u64)reg->smax_value) {
 		reg->umin_value = max_t(u64, reg->smin_value, reg->umin_value);
 		reg->umax_value = min_t(u64, reg->smax_value, reg->umax_value);
+	} else {
+		/* If the s64 range crosses the sign boundary, then it's split
+		 * between the beginning and end of the U64 domain. In that
+		 * case, we can derive new bounds if the u64 range overlaps
+		 * with only one end of the s64 range.
+		 *
+		 * In the following example, the u64 range overlaps only with
+		 * positive portion of the s64 range.
+		 *
+		 * 0                                                   U64_MAX
+		 * |  [xxxxxxxxxxxxxx u64 range xxxxxxxxxxxxxx]              |
+		 * |----------------------------|----------------------------|
+		 * |xxxxx s64 range xxxxxxxxx]                       [xxxxxxx|
+		 * 0                     S64_MAX S64_MIN                    -1
+		 *
+		 * We can thus derive the following new s64 and u64 ranges.
+		 *
+		 * 0                                                   U64_MAX
+		 * |  [xxxxxx u64 range xxxxx]                               |
+		 * |----------------------------|----------------------------|
+		 * |  [xxxxxx s64 range xxxxx]                               |
+		 * 0                     S64_MAX S64_MIN                    -1
+		 *
+		 * If they overlap in two places, we can't derive anything
+		 * because reg_state can't represent two ranges per numeric
+		 * domain.
+		 *
+		 * 0                                                   U64_MAX
+		 * |  [xxxxxxxxxxxxxxxxx u64 range xxxxxxxxxxxxxxxxx]        |
+		 * |----------------------------|----------------------------|
+		 * |xxxxx s64 range xxxxxxxxx]                    [xxxxxxxxxx|
+		 * 0                     S64_MAX S64_MIN                    -1
+		 *
+		 * The first condition below corresponds to the first diagram
+		 * above.
+		 */
+		if (reg->umax_value < (u64)reg->smin_value) {
+			reg->smin_value = (s64)reg->umin_value;
+			reg->umax_value = min_t(u64, reg->umax_value, reg->smax_value);
+		} else if ((u64)reg->smax_value < reg->umin_value) {
+			/* This second condition considers the case where the u64 range
+			 * overlaps with the negative portion of the s64 range:
+			 *
+			 * 0                                                   U64_MAX
+			 * |              [xxxxxxxxxxxxxx u64 range xxxxxxxxxxxxxx]  |
+			 * |----------------------------|----------------------------|
+			 * |xxxxxxxxx]                       [xxxxxxxxxxxx s64 range |
+			 * 0                     S64_MAX S64_MIN                    -1
+			 */
+			reg->smax_value = (s64)reg->umax_value;
+			reg->umin_value = max_t(u64, reg->umin_value, reg->smin_value);
+		}
 	}
 }
 
-- 
2.43.0

[PATCH stable 6.6 2/6] selftests/bpf: Test cross-sign 64bits range refinement

[Paul Chaignon]

[ Upstream commit 26e5e346a52c796190e63af1c2a80a417fda261a ]

This patch adds coverage for the new cross-sign 64bits range refinement
logic. The three tests cover the cases when the u64 and s64 ranges
overlap (1) in the negative portion of s64, (2) in the positive portion
of s64, and (3) in both portions.

The first test is a simplified version of a BPF program generated by
syzkaller that caused an invariant violation [1]. It looks like
syzkaller could not extract the reproducer itself (and therefore didn't
report it to the mailing list), but I was able to extract it from the
console logs of a crash.

The principle is similar to the invariant violation described in
commit 6279846b9b25 ("bpf: Forget ranges when refining tnum after
JSET"): the verifier walks a dead branch, uses the condition to refine
ranges, and ends up with inconsistent ranges. In this case, the dead
branch is when we fallthrough on both jumps. The new refinement logic
improves the bounds such that the second jump is properly detected as
always-taken and the verifier doesn't end up walking a dead branch.

The second and third tests are inspired by the first, but rely on
condition jumps to prepare the bounds instead of ALU instructions. An
R10 write is used to trigger a verifier error when the bounds can't be
refined.

Link: https://syzkaller.appspot.com/bug?extid=c711ce17dd78e5d4fdcf [1]
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
Link: https://lore.kernel.org/r/a0e17b00dab8dabcfa6f8384e7e151186efedfdd.1753695655.git.paul.chaignon@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
---
 .../selftests/bpf/progs/verifier_bounds.c     | 118 ++++++++++++++++++
 1 file changed, 118 insertions(+)

diff --git a/tools/testing/selftests/bpf/progs/verifier_bounds.c b/tools/testing/selftests/bpf/progs/verifier_bounds.c
index a0bb7fb40ea5..fe3e2b326c6b 100644
--- a/tools/testing/selftests/bpf/progs/verifier_bounds.c
+++ b/tools/testing/selftests/bpf/progs/verifier_bounds.c
@@ -1200,4 +1200,122 @@ l0_%=:	r0 = 0;						\
 	: __clobber_all);
 }
 
+/* This test covers the bounds deduction on 64bits when the s64 and u64 ranges
+ * overlap on the negative side. At instruction 7, the ranges look as follows:
+ *
+ * 0          umin=0xfffffcf1                 umax=0xff..ff6e  U64_MAX
+ * |                [xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx]        |
+ * |----------------------------|------------------------------|
+ * |xxxxxxxxxx]                                   [xxxxxxxxxxxx|
+ * 0    smax=0xeffffeee                       smin=-655        -1
+ *
+ * We should therefore deduce the following new bounds:
+ *
+ * 0                             u64=[0xff..ffd71;0xff..ff6e]  U64_MAX
+ * |                                              [xxx]        |
+ * |----------------------------|------------------------------|
+ * |                                              [xxx]        |
+ * 0                                        s64=[-655;-146]    -1
+ *
+ * Without the deduction cross sign boundary, we end up with an invariant
+ * violation error.
+ */
+SEC("socket")
+__description("bounds deduction cross sign boundary, negative overlap")
+__success __log_level(2) __flag(BPF_F_TEST_REG_INVARIANTS)
+__msg("7: (1f) r0 -= r6 {{.*}} R0=scalar(smin=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,smin32=-783,umin32=0xfffffcf1,umax32=0xffffff6e,var_off=(0xfffffffffffffc00; 0x3ff))")
+__retval(0)
+__naked void bounds_deduct_negative_overlap(void)
+{
+	asm volatile("			\
+	call %[bpf_get_prandom_u32];	\
+	w3 = w0;			\
+	w6 = (s8)w0;			\
+	r0 = (s8)r0;			\
+	if w6 >= 0xf0000000 goto l0_%=;	\
+	r0 += r6;			\
+	r6 += 400;			\
+	r0 -= r6;			\
+	if r3 < r0 goto l0_%=;		\
+l0_%=:	r0 = 0;				\
+	exit;				\
+"	:
+	: __imm(bpf_get_prandom_u32)
+	: __clobber_all);
+}
+
+/* This test covers the bounds deduction on 64bits when the s64 and u64 ranges
+ * overlap on the positive side. At instruction 3, the ranges look as follows:
+ *
+ * 0 umin=0                      umax=0xffffffffffffff00       U64_MAX
+ * [xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx]            |
+ * |----------------------------|------------------------------|
+ * |xxxxxxxx]                                         [xxxxxxxx|
+ * 0      smax=127                                smin=-128    -1
+ *
+ * We should therefore deduce the following new bounds:
+ *
+ * 0  u64=[0;127]                                              U64_MAX
+ * [xxxxxxxx]                                                  |
+ * |----------------------------|------------------------------|
+ * [xxxxxxxx]                                                  |
+ * 0  s64=[0;127]                                              -1
+ *
+ * Without the deduction cross sign boundary, the program is rejected due to
+ * the frame pointer write.
+ */
+SEC("socket")
+__description("bounds deduction cross sign boundary, positive overlap")
+__success __log_level(2) __flag(BPF_F_TEST_REG_INVARIANTS)
+__msg("3: (2d) if r0 > r1 {{.*}} R0_w=scalar(smin=smin32=0,smax=umax=smax32=umax32=127,var_off=(0x0; 0x7f))")
+__retval(0)
+__naked void bounds_deduct_positive_overlap(void)
+{
+	asm volatile("			\
+	call %[bpf_get_prandom_u32];	\
+	r0 = (s8)r0;			\
+	r1 = 0xffffffffffffff00;	\
+	if r0 > r1 goto l0_%=;		\
+	if r0 < 128 goto l0_%=;		\
+	r10 = 0;			\
+l0_%=:	r0 = 0;				\
+	exit;				\
+"	:
+	: __imm(bpf_get_prandom_u32)
+	: __clobber_all);
+}
+
+/* This test is the same as above, but the s64 and u64 ranges overlap in two
+ * places. At instruction 3, the ranges look as follows:
+ *
+ * 0 umin=0                           umax=0xffffffffffffff80  U64_MAX
+ * [xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx]        |
+ * |----------------------------|------------------------------|
+ * |xxxxxxxx]                                         [xxxxxxxx|
+ * 0      smax=127                                smin=-128    -1
+ *
+ * 0xffffffffffffff80 = (u64)-128. We therefore can't deduce anything new and
+ * the program should fail due to the frame pointer write.
+ */
+SEC("socket")
+__description("bounds deduction cross sign boundary, two overlaps")
+__failure __flag(BPF_F_TEST_REG_INVARIANTS)
+__msg("3: (2d) if r0 > r1 {{.*}} R0_w=scalar(smin=smin32=-128,smax=smax32=127,umax=0xffffffffffffff80)")
+__msg("frame pointer is read only")
+__naked void bounds_deduct_two_overlaps(void)
+{
+	asm volatile("			\
+	call %[bpf_get_prandom_u32];	\
+	r0 = (s8)r0;			\
+	r1 = 0xffffffffffffff80;	\
+	if r0 > r1 goto l0_%=;		\
+	if r0 < 128 goto l0_%=;		\
+	r10 = 0;			\
+l0_%=:	r0 = 0;				\
+	exit;				\
+"	:
+	: __imm(bpf_get_prandom_u32)
+	: __clobber_all);
+}
+
 char _license[] SEC("license") = "GPL";
-- 
2.43.0

[PATCH stable 6.6 3/6] selftests/bpf: Test invariants on JSLT crossing sign

[Paul Chaignon]

[ Upstream commit f96841bbf4a1ee4ed0336ba192a01278fdea6383 ]

The improvement of the u64/s64 range refinement fixed the invariant
violation that was happening on this test for BPF_JSLT when crossing the
sign boundary.

After this patch, we have one test remaining with a known invariant
violation. It's the same test as fixed here but for 32 bits ranges.

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
Link: https://lore.kernel.org/r/ad046fb0016428f1a33c3b81617aabf31b51183f.1753695655.git.paul.chaignon@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
---
 tools/testing/selftests/bpf/progs/verifier_bounds.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/bpf/progs/verifier_bounds.c b/tools/testing/selftests/bpf/progs/verifier_bounds.c
index fe3e2b326c6b..3924b1d1421b 100644
--- a/tools/testing/selftests/bpf/progs/verifier_bounds.c
+++ b/tools/testing/selftests/bpf/progs/verifier_bounds.c
@@ -1028,7 +1028,7 @@ l0_%=:	r0 = 0;						\
 SEC("xdp")
 __description("bound check with JMP_JSLT for crossing 64-bit signed boundary")
 __success __retval(0)
-__flag(!BPF_F_TEST_REG_INVARIANTS) /* known invariants violation */
+__flag(BPF_F_TEST_REG_INVARIANTS)
 __naked void crossing_64_bit_signed_boundary_2(void)
 {
 	asm volatile ("					\
-- 
2.43.0

[PATCH stable 6.6 4/6] bpf: Add third round of bounds deduction

[Paul Chaignon]

[ Upstream commit 5dbb19b16ac498b0b7f3a8a85f9d25d6d8af397d ]

Commit d7f008738171 ("bpf: try harder to deduce register bounds from
different numeric domains") added a second call to __reg_deduce_bounds
in reg_bounds_sync because a single call wasn't enough to converge to a
fixed point in terms of register bounds.

With patch "bpf: Improve bounds when s64 crosses sign boundary" from
this series, Eduard noticed that calling __reg_deduce_bounds twice isn't
enough anymore to converge. The first selftest added in "selftests/bpf:
Test cross-sign 64bits range refinement" highlights the need for a third
call to __reg_deduce_bounds. After instruction 7, reg_bounds_sync
performs the following bounds deduction:

  reg_bounds_sync entry:          scalar(smin=-655,smax=0xeffffeee,smin32=-783,smax32=-146)
  __update_reg_bounds:            scalar(smin=-655,smax=0xeffffeee,smin32=-783,smax32=-146)
  __reg_deduce_bounds:
      __reg32_deduce_bounds:      scalar(smin=-655,smax=0xeffffeee,smin32=-783,smax32=-146,umin32=0xfffffcf1,umax32=0xffffff6e)
      __reg64_deduce_bounds:      scalar(smin=-655,smax=0xeffffeee,smin32=-783,smax32=-146,umin32=0xfffffcf1,umax32=0xffffff6e)
      __reg_deduce_mixed_bounds:  scalar(smin=-655,smax=0xeffffeee,umin=umin32=0xfffffcf1,umax=0xffffffffffffff6e,smin32=-783,smax32=-146,umax32=0xffffff6e)
  __reg_deduce_bounds:
      __reg32_deduce_bounds:      scalar(smin=-655,smax=0xeffffeee,umin=umin32=0xfffffcf1,umax=0xffffffffffffff6e,smin32=-783,smax32=-146,umax32=0xffffff6e)
      __reg64_deduce_bounds:      scalar(smin=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,smin32=-783,umin32=0xfffffcf1,umax32=0xffffff6e)
      __reg_deduce_mixed_bounds:  scalar(smin=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,smin32=-783,umin32=0xfffffcf1,umax32=0xffffff6e)
  __reg_bound_offset:             scalar(smin=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,smin32=-783,umin32=0xfffffcf1,umax32=0xffffff6e,var_off=(0xfffffffffffffc00; 0x3ff))
  __update_reg_bounds:            scalar(smin=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,smin32=-783,umin32=0xfffffcf1,umax32=0xffffff6e,var_off=(0xfffffffffffffc00; 0x3ff))

In particular, notice how:
1. In the first call to __reg_deduce_bounds, __reg32_deduce_bounds
   learns new u32 bounds.
2. __reg64_deduce_bounds is unable to improve bounds at this point.
3. __reg_deduce_mixed_bounds derives new u64 bounds from the u32 bounds.
4. In the second call to __reg_deduce_bounds, __reg64_deduce_bounds
   improves the smax and umin bounds thanks to patch "bpf: Improve
   bounds when s64 crosses sign boundary" from this series.
5. Subsequent functions are unable to improve the ranges further (only
   tnums). Yet, a better smin32 bound could be learned from the smin
   bound.

__reg32_deduce_bounds is able to improve smin32 from smin, but for that
we need a third call to __reg_deduce_bounds.

As discussed in [1], there may be a better way to organize the deduction
rules to learn the same information with less calls to the same
functions. Such an optimization requires further analysis and is
orthogonal to the present patchset.

Link: https://lore.kernel.org/bpf/aIKtSK9LjQXB8FLY@mail.gmail.com/ [1]
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Co-developed-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
Link: https://lore.kernel.org/r/79619d3b42e5525e0e174ed534b75879a5ba15de.1753695655.git.paul.chaignon@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
---
 kernel/bpf/verifier.c                               | 1 +
 tools/testing/selftests/bpf/progs/verifier_bounds.c | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 6448f9eeede0..4ae3032f42e5 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -2292,6 +2292,7 @@ static void reg_bounds_sync(struct bpf_reg_state *reg)
 	/* We might have learned something about the sign bit. */
 	__reg_deduce_bounds(reg);
 	__reg_deduce_bounds(reg);
+	__reg_deduce_bounds(reg);
 	/* We might have learned some bits from the bounds. */
 	__reg_bound_offset(reg);
 	/* Intersecting with the old var_off might have improved our bounds
diff --git a/tools/testing/selftests/bpf/progs/verifier_bounds.c b/tools/testing/selftests/bpf/progs/verifier_bounds.c
index 3924b1d1421b..e6297e9dd2ed 100644
--- a/tools/testing/selftests/bpf/progs/verifier_bounds.c
+++ b/tools/testing/selftests/bpf/progs/verifier_bounds.c
@@ -1223,7 +1223,7 @@ l0_%=:	r0 = 0;						\
 SEC("socket")
 __description("bounds deduction cross sign boundary, negative overlap")
 __success __log_level(2) __flag(BPF_F_TEST_REG_INVARIANTS)
-__msg("7: (1f) r0 -= r6 {{.*}} R0=scalar(smin=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,smin32=-783,umin32=0xfffffcf1,umax32=0xffffff6e,var_off=(0xfffffffffffffc00; 0x3ff))")
+__msg("7: (1f) r0 -= r6 {{.*}} R0=scalar(smin=smin32=-655,smax=smax32=-146,umin=0xfffffffffffffd71,umax=0xffffffffffffff6e,umin32=0xfffffd71,umax32=0xffffff6e,var_off=(0xfffffffffffffc00; 0x3ff))")
 __retval(0)
 __naked void bounds_deduct_negative_overlap(void)
 {
-- 
2.43.0

[PATCH stable 6.6 5/6] bpf: Fix u32/s32 bounds when ranges cross min/max boundary

[Paul Chaignon]

From: Eduard Zingerman <eddyz87@gmail.com>

[ Upstream commit fbc7aef517d8765e4c425d2792409bb9bf2e1f13 ]

Same as in __reg64_deduce_bounds(), refine s32/u32 ranges
in __reg32_deduce_bounds() in the following situations:

- s32 range crosses U32_MAX/0 boundary, positive part of the s32 range
  overlaps with u32 range:

  0                                                   U32_MAX
  |  [xxxxxxxxxxxxxx u32 range xxxxxxxxxxxxxx]              |
  |----------------------------|----------------------------|
  |xxxxx s32 range xxxxxxxxx]                       [xxxxxxx|
  0                     S32_MAX S32_MIN                    -1

- s32 range crosses U32_MAX/0 boundary, negative part of the s32 range
  overlaps with u32 range:

  0                                                   U32_MAX
  |              [xxxxxxxxxxxxxx u32 range xxxxxxxxxxxxxx]  |
  |----------------------------|----------------------------|
  |xxxxxxxxx]                       [xxxxxxxxxxxx s32 range |
  0                     S32_MAX S32_MIN                    -1

- No refinement if ranges overlap in two intervals.

This helps for e.g. consider the following program:

   call %[bpf_get_prandom_u32];
   w0 &= 0xffffffff;
   if w0 < 0x3 goto 1f;    // on fall-through u32 range [3..U32_MAX]
   if w0 s> 0x1 goto 1f;   // on fall-through s32 range [S32_MIN..1]
   if w0 s< 0x0 goto 1f;   // range can be narrowed to  [S32_MIN..-1]
   r10 = 0;
1: ...;

The reg_bounds.c selftest is updated to incorporate identical logic,
refinement based on non-overflowing range halves:

  ((x ∩ [0, smax]) ∩ (y ∩ [0, smax])) ∪
  ((x ∩ [smin,-1]) ∩ (y ∩ [smin,-1]))

Reported-by: Andrea Righi <arighi@nvidia.com>
Reported-by: Emil Tsalapatis <emil@etsalapatis.com>
Closes: https://lore.kernel.org/bpf/aakqucg4vcujVwif@gpd4/T/
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20260306-bpf-32-bit-range-overflow-v3-1-f7f67e060a6b@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
---
 kernel/bpf/verifier.c                         | 24 +++++++
 .../selftests/bpf/prog_tests/reg_bounds.c     | 62 +++++++++++++++++--
 2 files changed, 82 insertions(+), 4 deletions(-)

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 4ae3032f42e5..7769a2a47bcb 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -2046,6 +2046,30 @@ static void __reg32_deduce_bounds(struct bpf_reg_state *reg)
 	if ((u32)reg->s32_min_value <= (u32)reg->s32_max_value) {
 		reg->u32_min_value = max_t(u32, reg->s32_min_value, reg->u32_min_value);
 		reg->u32_max_value = min_t(u32, reg->s32_max_value, reg->u32_max_value);
+	} else {
+		if (reg->u32_max_value < (u32)reg->s32_min_value) {
+			/* See __reg64_deduce_bounds() for detailed explanation.
+			 * Refine ranges in the following situation:
+			 *
+			 * 0                                                   U32_MAX
+			 * |  [xxxxxxxxxxxxxx u32 range xxxxxxxxxxxxxx]              |
+			 * |----------------------------|----------------------------|
+			 * |xxxxx s32 range xxxxxxxxx]                       [xxxxxxx|
+			 * 0                     S32_MAX S32_MIN                    -1
+			 */
+			reg->s32_min_value = (s32)reg->u32_min_value;
+			reg->u32_max_value = min_t(u32, reg->u32_max_value, reg->s32_max_value);
+		} else if ((u32)reg->s32_max_value < reg->u32_min_value) {
+			/*
+			 * 0                                                   U32_MAX
+			 * |              [xxxxxxxxxxxxxx u32 range xxxxxxxxxxxxxx]  |
+			 * |----------------------------|----------------------------|
+			 * |xxxxxxxxx]                       [xxxxxxxxxxxx s32 range |
+			 * 0                     S32_MAX S32_MIN                    -1
+			 */
+			reg->s32_max_value = (s32)reg->u32_max_value;
+			reg->u32_min_value = max_t(u32, reg->u32_min_value, reg->s32_min_value);
+		}
 	}
 }
 
diff --git a/tools/testing/selftests/bpf/prog_tests/reg_bounds.c b/tools/testing/selftests/bpf/prog_tests/reg_bounds.c
index 39d42271cc46..ac4e2557f739 100644
--- a/tools/testing/selftests/bpf/prog_tests/reg_bounds.c
+++ b/tools/testing/selftests/bpf/prog_tests/reg_bounds.c
@@ -422,15 +422,69 @@ static bool is_valid_range(enum num_t t, struct range x)
 	}
 }
 
-static struct range range_improve(enum num_t t, struct range old, struct range new)
+static struct range range_intersection(enum num_t t, struct range old, struct range new)
 {
 	return range(t, max_t(t, old.a, new.a), min_t(t, old.b, new.b));
 }
 
+/*
+ * Result is precise when 'x' and 'y' overlap or form a continuous range,
+ * result is an over-approximation if 'x' and 'y' do not overlap.
+ */
+static struct range range_union(enum num_t t, struct range x, struct range y)
+{
+	if (!is_valid_range(t, x))
+		return y;
+	if (!is_valid_range(t, y))
+		return x;
+	return range(t, min_t(t, x.a, y.a), max_t(t, x.b, y.b));
+}
+
+/*
+ * This function attempts to improve x range intersecting it with y.
+ * range_cast(... to_t ...) looses precision for ranges that pass to_t
+ * min/max boundaries. To avoid such precision loses this function
+ * splits both x and y into halves corresponding to non-overflowing
+ * sub-ranges: [0, smin] and [smax, -1].
+ * Final result is computed as follows:
+ *
+ *   ((x ∩ [0, smax]) ∩ (y ∩ [0, smax])) ∪
+ *   ((x ∩ [smin,-1]) ∩ (y ∩ [smin,-1]))
+ *
+ * Precision might still be lost if final union is not a continuous range.
+ */
+static struct range range_refine_in_halves(enum num_t x_t, struct range x,
+					   enum num_t y_t, struct range y)
+{
+	struct range x_pos, x_neg, y_pos, y_neg, r_pos, r_neg;
+	u64 smax, smin, neg_one;
+
+	if (t_is_32(x_t)) {
+		smax = (u64)(u32)S32_MAX;
+		smin = (u64)(u32)S32_MIN;
+		neg_one = (u64)(u32)(s32)(-1);
+	} else {
+		smax = (u64)S64_MAX;
+		smin = (u64)S64_MIN;
+		neg_one = U64_MAX;
+	}
+	x_pos = range_intersection(x_t, x, range(x_t, 0, smax));
+	x_neg = range_intersection(x_t, x, range(x_t, smin, neg_one));
+	y_pos = range_intersection(y_t, y, range(x_t, 0, smax));
+	y_neg = range_intersection(y_t, y, range(y_t, smin, neg_one));
+	r_pos = range_intersection(x_t, x_pos, range_cast(y_t, x_t, y_pos));
+	r_neg = range_intersection(x_t, x_neg, range_cast(y_t, x_t, y_neg));
+	return range_union(x_t, r_pos, r_neg);
+
+}
+
 static struct range range_refine(enum num_t x_t, struct range x, enum num_t y_t, struct range y)
 {
 	struct range y_cast;
 
+	if (t_is_32(x_t) == t_is_32(y_t))
+		x = range_refine_in_halves(x_t, x, y_t, y);
+
 	y_cast = range_cast(y_t, x_t, y);
 
 	/* If we know that
@@ -444,7 +498,7 @@ static struct range range_refine(enum num_t x_t, struct range x, enum num_t y_t,
 	 */
 	if (x_t == S64 && y_t == S32 && y_cast.a <= S32_MAX  && y_cast.b <= S32_MAX &&
 	    (s64)x.a >= S32_MIN && (s64)x.b <= S32_MAX)
-		return range_improve(x_t, x, y_cast);
+		return range_intersection(x_t, x, y_cast);
 
 	/* the case when new range knowledge, *y*, is a 32-bit subregister
 	 * range, while previous range knowledge, *x*, is a full register
@@ -462,11 +516,11 @@ static struct range range_refine(enum num_t x_t, struct range x, enum num_t y_t,
 		x_swap = range(x_t, swap_low32(x.a, y_cast.a), swap_low32(x.b, y_cast.b));
 		if (!is_valid_range(x_t, x_swap))
 			return x;
-		return range_improve(x_t, x, x_swap);
+		return range_intersection(x_t, x, x_swap);
 	}
 
 	/* otherwise, plain range cast and intersection works */
-	return range_improve(x_t, x, y_cast);
+	return range_intersection(x_t, x, y_cast);
 }
 
 /* =======================
-- 
2.43.0

[PATCH stable 6.6 6/6] selftests/bpf: test refining u32/s32 bounds when ranges cross min/max boundary

[Paul Chaignon]

From: Eduard Zingerman <eddyz87@gmail.com>

[ Upstream commit f81fdfd16771e266753146bd83f6dd23515ebee9 ]

Two test cases for signed/unsigned 32-bit bounds refinement
when s32 range crosses the sign boundary:
- s32 range [S32_MIN..1] overlapping with u32 range [3..U32_MAX],
  s32 range tail before sign boundary overlaps with u32 range.
- s32 range [-3..5] overlapping with u32 range [0..S32_MIN+3],
  s32 range head after the sign boundary overlaps with u32 range.

This covers both branches added in the __reg32_deduce_bounds().

Also, crossing_32_bit_signed_boundary_2() no longer triggers invariant
violations.

Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Reviewed-by: Paul Chaignon <paul.chaignon@gmail.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20260306-bpf-32-bit-range-overflow-v3-2-f7f67e060a6b@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
---
 .../selftests/bpf/progs/verifier_bounds.c     | 39 ++++++++++++++++++-
 1 file changed, 38 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/bpf/progs/verifier_bounds.c b/tools/testing/selftests/bpf/progs/verifier_bounds.c
index e6297e9dd2ed..5feb07584084 100644
--- a/tools/testing/selftests/bpf/progs/verifier_bounds.c
+++ b/tools/testing/selftests/bpf/progs/verifier_bounds.c
@@ -1110,7 +1110,7 @@ l0_%=:	r0 = 0;						\
 SEC("xdp")
 __description("bound check with JMP32_JSLT for crossing 32-bit signed boundary")
 __success __retval(0)
-__flag(!BPF_F_TEST_REG_INVARIANTS) /* known invariants violation */
+__flag(BPF_F_TEST_REG_INVARIANTS)
 __naked void crossing_32_bit_signed_boundary_2(void)
 {
 	asm volatile ("					\
@@ -1318,4 +1318,41 @@ l0_%=:	r0 = 0;				\
 	: __clobber_all);
 }
 
+SEC("socket")
+__success
+__flag(BPF_F_TEST_REG_INVARIANTS)
+__naked void signed_unsigned_intersection32_case1(void *ctx)
+{
+	asm volatile("									\
+	call %[bpf_get_prandom_u32];							\
+	w0 &= 0xffffffff;								\
+	if w0 < 0x3 goto 1f;		/* on fall-through u32 range [3..U32_MAX]  */	\
+	if w0 s> 0x1 goto 1f;		/* on fall-through s32 range [S32_MIN..1]  */	\
+	if w0 s< 0x0 goto 1f;		/* range can be narrowed to  [S32_MIN..-1] */	\
+	r10 = 0;			/* thus predicting the jump. */			\
+1:	exit;										\
+"	:
+	: __imm(bpf_get_prandom_u32)
+	: __clobber_all);
+}
+
+SEC("socket")
+__success
+__flag(BPF_F_TEST_REG_INVARIANTS)
+__naked void signed_unsigned_intersection32_case2(void *ctx)
+{
+	asm volatile("									\
+	call %[bpf_get_prandom_u32];							\
+	w0 &= 0xffffffff;								\
+	if w0 > 0x80000003 goto 1f;	/* on fall-through u32 range [0..S32_MIN+3] */	\
+	if w0 s< -3 goto 1f;		/* on fall-through s32 range [-3..S32_MAX] */	\
+	if w0 s> 5 goto 1f;		/* on fall-through s32 range [-3..5] */		\
+	if w0 <= 5 goto 1f;		/* range can be narrowed to  [0..5] */		\
+	r10 = 0;			/* thus predicting the jump */			\
+1:	exit;										\
+"	:
+	: __imm(bpf_get_prandom_u32)
+	: __clobber_all);
+}
+
 char _license[] SEC("license") = "GPL";
-- 
2.43.0

Re: [PATCH stable 6.6 0/6] bpf: Fix bounds when ranges cross sign boundary

[Paul Chaignon]

On Fri, Apr 03, 2026 at 05:33:59PM +0200, Paul Chaignon wrote:
> As discussed in [1] yesterday, this series backports two sets of fixes
> for BPF, with their selftests:
> - 00bf8d0c6c9b ("bpf: Improve bounds when s64 crosses sign boundary")
> - 26e5e346a52c ("selftests/bpf: Test cross-sign 64bits range
>   refinement")
> - f96841bbf4a1 ("selftests/bpf: Test invariants on JSLT crossing sign")
> - 5dbb19b16ac4 ("bpf: Add third round of bounds deduction")
> - fbc7aef517d8 ("bpf: Fix u32/s32 bounds when ranges cross min/max
>   boundary")
> - f81fdfd16771 ("selftests/bpf: test refining u32/s32 bounds when
>   ranges cross min/max boundary")
> 
> Using Shung-Hsi's stable CI repo [2], I verified the BPF selftests pass
> with these commits applied on top of v6.12.

As hinted here, the subject prefix is incorrect. This series is meant
for v6.12, not v6.6. Should I resend?

> 
> 1: https://lore.kernel.org/stable/2026040240-friday-gurgling-7088@gregkh/
> 2: https://github.com/pchaigno/stable-bpf-ci/actions/runs/23940850516/job/69826632354
> 
> Eduard Zingerman (2):
>   bpf: Fix u32/s32 bounds when ranges cross min/max boundary
>   selftests/bpf: test refining u32/s32 bounds when ranges cross min/max
>     boundary
> 
> Paul Chaignon (4):
>   bpf: Improve bounds when s64 crosses sign boundary
>   selftests/bpf: Test cross-sign 64bits range refinement
>   selftests/bpf: Test invariants on JSLT crossing sign
>   bpf: Add third round of bounds deduction
> 
>  kernel/bpf/verifier.c                         |  77 +++++++++
>  .../selftests/bpf/prog_tests/reg_bounds.c     |  62 ++++++-
>  .../selftests/bpf/progs/verifier_bounds.c     | 159 +++++++++++++++++-
>  3 files changed, 292 insertions(+), 6 deletions(-)
> 
> -- 
> 2.43.0
> 

Re: [PATCH stable 6.6 0/6] bpf: Fix bounds when ranges cross sign boundary

[Greg Kroah-Hartman]

On Fri, Apr 03, 2026 at 11:05:47PM +0200, Paul Chaignon wrote:
> On Fri, Apr 03, 2026 at 05:33:59PM +0200, Paul Chaignon wrote:
> > As discussed in [1] yesterday, this series backports two sets of fixes
> > for BPF, with their selftests:
> > - 00bf8d0c6c9b ("bpf: Improve bounds when s64 crosses sign boundary")
> > - 26e5e346a52c ("selftests/bpf: Test cross-sign 64bits range
> >   refinement")
> > - f96841bbf4a1 ("selftests/bpf: Test invariants on JSLT crossing sign")
> > - 5dbb19b16ac4 ("bpf: Add third round of bounds deduction")
> > - fbc7aef517d8 ("bpf: Fix u32/s32 bounds when ranges cross min/max
> >   boundary")
> > - f81fdfd16771 ("selftests/bpf: test refining u32/s32 bounds when
> >   ranges cross min/max boundary")
> > 
> > Using Shung-Hsi's stable CI repo [2], I verified the BPF selftests pass
> > with these commits applied on top of v6.12.
> 
> As hinted here, the subject prefix is incorrect. This series is meant
> for v6.12, not v6.6. Should I resend?

Yes please!

[PATCH stable 6.12 0/6] bpf: Fix bounds when ranges cross sign boundary

[Paul Chaignon]

As discussed in [1], this series backports two sets of fixes for BPF,
with their selftests:
- 00bf8d0c6c9b ("bpf: Improve bounds when s64 crosses sign boundary")
- 26e5e346a52c ("selftests/bpf: Test cross-sign 64bits range
  refinement")
- f96841bbf4a1 ("selftests/bpf: Test invariants on JSLT crossing sign")
- 5dbb19b16ac4 ("bpf: Add third round of bounds deduction")
- fbc7aef517d8 ("bpf: Fix u32/s32 bounds when ranges cross min/max
  boundary")
- f81fdfd16771 ("selftests/bpf: test refining u32/s32 bounds when
  ranges cross min/max boundary")

Using Shung-Hsi's stable CI repo [2], I verified the BPF selftests pass
with these commits applied on top of v6.12.

1: https://lore.kernel.org/stable/2026040240-friday-gurgling-7088@gregkh/
2: https://github.com/pchaigno/stable-bpf-ci/actions/runs/23940850516/job/69826632354

Eduard Zingerman (2):
  bpf: Fix u32/s32 bounds when ranges cross min/max boundary
  selftests/bpf: test refining u32/s32 bounds when ranges cross min/max
    boundary

Paul Chaignon (4):
  bpf: Improve bounds when s64 crosses sign boundary
  selftests/bpf: Test cross-sign 64bits range refinement
  selftests/bpf: Test invariants on JSLT crossing sign
  bpf: Add third round of bounds deduction

 kernel/bpf/verifier.c                         |  77 +++++++++
 .../selftests/bpf/prog_tests/reg_bounds.c     |  62 ++++++-
 .../selftests/bpf/progs/verifier_bounds.c     | 159 +++++++++++++++++-
 3 files changed, 292 insertions(+), 6 deletions(-)

-- 
2.43.0