bpf: Fix incorrect immediate spill
Subsystems (5)
v2 cover bpf: Fix incorrect immediate spill
Immediate is incorrectly cast to u32 before being spilled, losing sign information. The range information is incorrect after load again. Fix immediate spill by remove the cast. The second patch add a test case for this.
---
Changes in v2:
- Add fix and cc tags.
- Link to v1: https://lore.kernel.org/r/[email protected]
---
Hao Sun (2):
bpf: Fix check_stack_write_fixed_off() to correctly spill imm selftests/bpf: Add test for immediate spilled to stack
kernel/bpf/verifier.c | 2 +- tools/testing/selftests/bpf/verifier/bpf_st_mem.c | 32 +++++++++++++++++++++++ 2 files changed, 33 insertions(+), 1 deletion(-)
---
base-commit: f1c73396133cb3d913e2075298005644ee8dfade
change-id: 20231026-fix-check-stack-write-c40996694dfa
Best regards,
Signature · 1 line
Hao Sun <[email protected]>
PATCH v2 1/2 bpf: Fix check_stack_write_fixed_off() to correctly spill imm
In check_stack_write_fixed_off(), imm value is cast to u32 before being spilled to the stack. Therefore, the sign information is lost, and the range information is incorrect when load from the stack again.
For the following prog:
0: r2 = r10
1: *(u64*)(r2 -40) = -44
2: r0 = *(u64*)(r2 - 40)
3: if r0 s<= 0xa goto +2
4: r0 = 1
5: exit
6: r0 = 0
7: exit
The verifier gives:
func#0 @0
0: R1=ctx(off=0,imm=0) R10=fp0
0: (bf) r2 = r10 ; R2_w=fp0 R10=fp0 1: (7a) *(u64 *)(r2 -40) = -44 ; R2_w=fp0 fp-40_w=4294967252 2: (79) r0 = *(u64 *)(r2 -40) ; R0_w=4294967252 R2_w=fp0
fp-40_w=4294967252
3: (c5) if r0 s< 0xa goto pc+2mark_precise: frame0: last_idx 3 first_idx 0 subseq_idx -1mark_precise: frame0: regs=r0 stack= before 2: (79) r0 = *(u64 *)(r2 -40) 3: R0_w=4294967252
4: (b7) r0 = 1 ; R0_w=1
5: (95) exit
verification time 7971 usec
stack depth 40
processed 6 insns (limit 1000000) max_states_per_insn 0 total_states 0 peak_states 0 mark_read 0
So remove the incorrect cast, since imm field is declared as s32, and __mark_reg_known() takes u64, so imm would be correctly sign extended by compiler.
--- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/bpf/verifier.c +1 -1
@@ -4674,7 +4674,7 @@ static int check_stack_write_fixed_off(struct bpf_verifier_env *env, insn->imm != 0 && env->bpf_capable) {struct bpf_reg_state fake_reg = {}; -__mark_reg_known (&fake_reg, (u32 )insn->imm); +__mark_reg_known (&fake_reg, insn->imm); fake_reg.type =SCALAR_VALUE ;save_register_state (state, spi, &fake_reg, size); } else if (reg && is_spillable_regtype(reg->type)) {
Signature · 1 line
2.34.1
Re: [PATCH bpf v2 1/2] bpf: Fix check_stack_write_fixed_off() to correctly spill imm
Quoted 32 lines · Hao Sun
In
check_stack_write_fixed_off(), imm value is cast to u32 before being spilled to the stack. Therefore, the sign information is lost, and the range information is incorrect when load from the stack again.For the following prog:
0: r2 = r10
1: *(u64*)(r2 -40) = -44
2: r0 = *(u64*)(r2 - 40)
3: if r0 s<=0xagoto +2
4: r0 = 1
5: exit
6: r0 = 0
7: exitThe verifier gives:
func#0 @0
0: R1=ctx(off=0,imm=0) R10=fp00: (bf) r2 = r10 ; R2_w=fp0 R10=fp0 1: (7a) *(u64 *)(r2 -40) = -44 ; R2_w=fp0 fp-40_w=4294967252 2: (79) r0 = *(u64 *)(r2 -40) ; R0_w=4294967252 R2_w=fp0fp-40_w=4294967252
3: (c5) if r0 s<0xagoto pc+2mark_precise: frame0:last_idx3first_idx0subseq_idx-1mark_precise: frame0: regs=r0 stack= before 2: (79) r0 = *(u64 *)(r2 -40) 3: R0_w=4294967252
4: (b7) r0 = 1 ; R0_w=1
5: (95) exit
verification time 7971 usec
stack depth 40
processed 6 insns (limit 1000000)max_states_per_insn0total_states0peak_states0mark_read0
Quoted 7 lines
So remove the incorrect cast, since imm field is declared as s32, and
__mark_reg_known()takes u64, so imm would be correctly sign extended by compiler.Fixes ecdf985d7615 ("bpf: track immediate values written to stack by BPF_ST instruction")Signed-off-by Hao Sun <[email protected]>
Re: [PATCH bpf v2 1/2] bpf: Fix check_stack_write_fixed_off() to correctly spill imm
In reply to Hao SunQuoted 32 lines · Hao Sun
In
check_stack_write_fixed_off(), imm value is cast to u32 before being spilled to the stack. Therefore, the sign information is lost, and the range information is incorrect when load from the stack again.For the following prog:
0: r2 = r10
1: *(u64*)(r2 -40) = -44
2: r0 = *(u64*)(r2 - 40)
3: if r0 s<=0xagoto +2
4: r0 = 1
5: exit
6: r0 = 0
7: exitThe verifier gives:
func#0 @0
0: R1=ctx(off=0,imm=0) R10=fp00: (bf) r2 = r10 ; R2_w=fp0 R10=fp0 1: (7a) *(u64 *)(r2 -40) = -44 ; R2_w=fp0 fp-40_w=4294967252 2: (79) r0 = *(u64 *)(r2 -40) ; R0_w=4294967252 R2_w=fp0fp-40_w=4294967252
3: (c5) if r0 s<0xagoto pc+2mark_precise: frame0:last_idx3first_idx0subseq_idx-1mark_precise: frame0: regs=r0 stack= before 2: (79) r0 = *(u64 *)(r2 -40) 3: R0_w=4294967252
4: (b7) r0 = 1 ; R0_w=1
5: (95) exit
verification time 7971 usec
stack depth 40
processed 6 insns (limit 1000000)max_states_per_insn0total_states0peak_states0mark_read0
Quoted 7 lines
So remove the incorrect cast, since imm field is declared as s32, and
__mark_reg_known()takes u64, so imm would be correctly sign extended by compiler.Fixes ecdf985d7615 ("bpf: track immediate values written to stack by BPF_ST instruction")Signed-off-by Hao Sun <[email protected]>
Quoted 18 lines
--- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)kernel/bpf/verifier.c +1
-1@@ -4674,7 +4674,7 @@ static int check_stack_write_fixed_off(struct bpf_verifier_env *env, insn->imm != 0 && env->bpf_capable) {struct bpf_reg_state fake_reg = {}; -__mark_reg_known (&fake_reg, (u32 )insn->imm); +__mark_reg_known (&fake_reg, insn->imm); fake_reg.type =SCALAR_VALUE ;save_register_state (state, spi, &fake_reg, size); } else if (reg && is_spillable_regtype(reg->type)) {
PATCH v2 2/2 selftests/bpf: Add test for immediate spilled to stack
In reply to Hao SunAdd a test to check if the verifier correctly reason about the sign of an immediate spilled to stack by BPF_ST instruction.
--- tools/testing/selftests/bpf/verifier/bpf_st_mem.c | 32 +++++++++++++++++++++++ 1 file changed, 32 insertions(+)
tools/testing/selftests/bpf/verifier/bpf_st_mem.c +32 -0
@@ -65,3 +65,35 @@ .expected_attach_type = BPF_SK_LOOKUP, .runs = -1, }, +{ + "BPF_ST_MEM stack imm sign", +/* Check if verifier correctly reasons about sign of an +* immediate spilled to stack by BPF_ST instruction. +* +* fp[-8] = -44; +* r0 = fp[-8]; +* if r0 s< 0 goto ret0; +* r0 = -1; +* exit; +* ret0: +* r0 = 0; +* exit; +*/ + .insns = { +BPF_ST_MEM (BPF_DW ,BPF_REG_10 ,-8 ,-44 ), +BPF_LDX_MEM (BPF_DW ,BPF_REG_0 ,BPF_REG_10 ,-8 ), +BPF_JMP_IMM (BPF_JSLT ,BPF_REG_0 ,0 ,2 ), +BPF_MOV64_IMM (BPF_REG_0 ,-1 ), +BPF_EXIT_INSN (), +BPF_MOV64_IMM (BPF_REG_0 ,0 ), + BPF_EXIT_INSN(), + }, +/* Use prog type that requires return value in range [0, 1] */ + .prog_type =BPF_PROG_TYPE_SK_LOOKUP , + .expected_attach_type = BPF_SK_LOOKUP, + .result = VERBOSE_ACCEPT, + .runs = -1, + .errstr = "0: (7a) *(u64 *)(r10 -8) = -44 ; R10=fp0 fp-8_w=-44\ +2: (c5) if r0 s< 0x0 goto pc+2\ + 2: R0_w=-44", +},
Signature · 1 line
2.34.1
Eduard Zingerman
maintainerRe: [PATCH bpf v2 2/2] selftests/bpf: Add test for immediate spilled to stack
Quoted 39 lines · Hao Sun
Add a test to check if the verifier correctly reason about the sign of an immediate spilled to stack by
BPF_STinstruction.Signed-off-by Hao Sun <[email protected]>--- tools/testing/selftests/bpf/verifier/bpf_st_mem.c | 32 +++++++++++++++++++++++ 1 file changed, 32 insertions(+)tools/testing/selftests/bpf/verifier/bpf_st_mem.c +22
-0@@ -65,3 +65,35 @@ .expected_attach_type = BPF_SK_LOOKUP, .runs = -1, }, +{ + "BPF_ST_MEM stack imm sign", +/* Check if verifier correctly reasons about sign of an +* immediate spilled to stack by BPF_ST instruction. +* +* fp[-8] = -44; +* r0 = fp[-8]; +* if r0 s< 0 goto ret0; +* r0 = -1; +* exit; +* ret0: +* r0 = 0; +* exit; +*/ + .insns = { +BPF_ST_MEM (BPF_DW ,BPF_REG_10 ,-8 ,-44 ), +BPF_LDX_MEM (BPF_DW ,BPF_REG_0 ,BPF_REG_10 ,-8 ), +BPF_JMP_IMM (BPF_JSLT ,BPF_REG_0 ,0 ,2 ), +BPF_MOV64_IMM (BPF_REG_0 ,-1 ), +BPF_EXIT_INSN (), +BPF_MOV64_IMM (BPF_REG_0 ,0 ), + BPF_EXIT_INSN(),
tools/testing/selftests/bpf/verifier/bpf_st_mem.c line 99
Quoted 10 lines
tools/testing/selftests/bpf/verifier/bpf_st_mem.c +10
-0+ }, + /* Use prog type that requires return value in range [0, 1] */ + .prog_type = BPF_PROG_TYPE_SK_LOOKUP, + .expected_attach_type = BPF_SK_LOOKUP, + .result = VERBOSE_ACCEPT, + .runs = -1, + .errstr ="0: (7a) *(u64 *)(r10 -8) = -44 ; R10=fp0 fp-8_w=-44\ +2: (c5) if r0 s< 0x0 goto pc+2\ + 2: R0_w=-44", +},
Please note that this test case fails on CI [0], full log below:
2023-11-01T07:49:51.2841702Z #116/p BPF_ST_MEM stack imm sign FAIL 2023-11-01T07:49:51.2843456Z Unexpected verifier log! 2023-11-01T07:49:51.2844968Z EXP: 2: R0_w=-44 2023-11-01T07:49:51.2845583Z RES: 2023-11-01T07:49:51.2846693Z func#0 @0 2023-11-01T07:49:51.2848932Z 0: R1=ctx(off=0,imm=0) R10=fp0 2023-11-01T07:49:51.2853045Z 0: (7a) *(u64 *)(r10 -8) = -44 ; R10=fp0 fp-8_w=-44 2023-11-01T07:49:51.2857391Z 1: (79) r0 = *(u64 *)(r10 -8) ; R0_w=-44 R10=fp0 fp-8_w=-44 2023-11-01T07:49:51.2859127Z 2: (c5) if r0 s< 0x0 goto pc+2 2023-11-01T07:49:51.2862943Z mark_precise: frame0: last_idx 2 first_idx 0 subseq_idx -1 2023-11-01T07:49:51.2867511Z mark_precise: frame0: regs=r0 stack= before 1: (79) r0 = *(u64 *)(r10 -8) 2023-11-01T07:49:51.2872217Z mark_precise: frame0: regs= stack=-8 before 0: (7a) *(u64 *)(r10 -8) = -44 2023-11-01T07:49:51.2872816Z 5: R0_w=-44 2023-11-01T07:49:51.2875653Z 5: (b7) r0 = 0 ; R0_w=0 2023-11-01T07:49:51.2876493Z 6: (95) exit
I suspect that after recent logging fixes instruction number printed after jump changed and that's why test case no longer passes.
Note: you can check CI status for submitted patch-sets using link [1].
[0] https://github.com/kernel-patches/bpf/actions/runs/6717053909/job/18254330860
[1] https://patchwork.kernel.org/project/netdevbpf/list/
Re: [PATCH bpf v2 2/2] selftests/bpf: Add test for immediate spilled to stack
Quoted 70 lines · Eduard Zingerman
Quoted 39 lines · Hao Sun
Add a test to check if the verifier correctly reason about the sign of an immediate spilled to stack by
BPF_STinstruction.Signed-off-by Hao Sun <[email protected]>--- tools/testing/selftests/bpf/verifier/bpf_st_mem.c | 32 +++++++++++++++++++++++ 1 file changed, 32 insertions(+)tools/testing/selftests/bpf/verifier/bpf_st_mem.c +22
-0@@ -65,3 +65,35 @@ .expected_attach_type = BPF_SK_LOOKUP, .runs = -1, }, +{ + "BPF_ST_MEM stack imm sign", +/* Check if verifier correctly reasons about sign of an +* immediate spilled to stack by BPF_ST instruction. +* +* fp[-8] = -44; +* r0 = fp[-8]; +* if r0 s< 0 goto ret0; +* r0 = -1; +* exit; +* ret0: +* r0 = 0; +* exit; +*/ + .insns = { +BPF_ST_MEM (BPF_DW ,BPF_REG_10 ,-8 ,-44 ), +BPF_LDX_MEM (BPF_DW ,BPF_REG_0 ,BPF_REG_10 ,-8 ), +BPF_JMP_IMM (BPF_JSLT ,BPF_REG_0 ,0 ,2 ), +BPF_MOV64_IMM (BPF_REG_0 ,-1 ), +BPF_EXIT_INSN (), +BPF_MOV64_IMM (BPF_REG_0 ,0 ), + BPF_EXIT_INSN(),tools/testing/selftests/bpf/verifier/bpf_st_mem.c line 99
Quoted 10 lines
tools/testing/selftests/bpf/verifier/bpf_st_mem.c +10
-0+ }, + /* Use prog type that requires return value in range [0, 1] */ + .prog_type = BPF_PROG_TYPE_SK_LOOKUP, + .expected_attach_type = BPF_SK_LOOKUP, + .result = VERBOSE_ACCEPT, + .runs = -1, + .errstr ="0: (7a) *(u64 *)(r10 -8) = -44 ; R10=fp0 fp-8_w=-44\ +2: (c5) if r0 s< 0x0 goto pc+2\ + 2: R0_w=-44", +},Please note that this test case fails on CI [0], full log below:
2023-11-01T07:49:51.2841702Z #116/p BPF_ST_MEM stack imm sign FAIL 2023-11-01T07:49:51.2843456Z Unexpected verifier log! 2023-11-01T07:49:51.2844968Z EXP: 2: R0_w=-44 2023-11-01T07:49:51.2845583Z RES: 2023-11-01T07:49:51.2846693Z func#0 @0 2023-11-01T07:49:51.2848932Z 0: R1=ctx(off=0,imm=0) R10=fp0 2023-11-01T07:49:51.2853045Z 0: (7a) *(u64 *)(r10 -8) = -44 ; R10=fp0 fp-8_w=-44 2023-11-01T07:49:51.2857391Z 1: (79) r0 = *(u64 *)(r10 -8) ; R0_w=-44 R10=fp0 fp-8_w=-44 2023-11-01T07:49:51.2859127Z 2: (c5) if r0 s< 0x0 goto pc+2 2023-11-01T07:49:51.2862943Z mark_precise: frame0: last_idx 2 first_idx 0 subseq_idx -1 2023-11-01T07:49:51.2867511Z mark_precise: frame0: regs=r0 stack= before 1: (79) r0 = *(u64 *)(r10 -8) 2023-11-01T07:49:51.2872217Z mark_precise: frame0: regs= stack=-8 before 0: (7a) *(u64 *)(r10 -8) = -44 2023-11-01T07:49:51.2872816Z 5: R0_w=-44 2023-11-01T07:49:51.2875653Z 5: (b7) r0 = 0 ; R0_w=0 2023-11-01T07:49:51.2876493Z 6: (95) exit
Quoted 3 lines
I suspect that after recent logging fixes instruction number printed after jump changed and that's why test case no longer passes.
Yes, so I guess we can just drop the line number there, will send patch v3.
Quoted 4 lines
Note: you can check CI status for submitted patch-sets using link [1].
[0] https://github.com/kernel-patches/bpf/actions/runs/6717053909/job/18254330860
[1] https://patchwork.kernel.org/project/netdevbpf/list/
Thanks.