diff mbox series

[bpf-next,1/2] bpf: Fix updating attached freplace prog to PROG_ARRAY map

Message ID 20240725003251.37855-2-leon.hwang@linux.dev (mailing list archive)
State Superseded
Delegated to: BPF
Headers show
Series bpf: Fix updating attached freplace prog to PROG_ARRAY map | expand

Checks

Context Check Description
netdev/series_format success Posting correctly formatted
netdev/tree_selection success Clearly marked for bpf-next, async
netdev/ynl success Generated files up to date; no warnings/errors; no diff in generated;
netdev/fixes_present success Fixes tag not required for -next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 661 this patch: 661
netdev/build_tools success Errors and warnings before: 0 this patch: 0
netdev/cc_maintainers warning 7 maintainers not CCed: kpsingh@kernel.org martin.lau@linux.dev haoluo@google.com jolsa@kernel.org song@kernel.org sdf@fomichev.me john.fastabend@gmail.com
netdev/build_clang success Errors and warnings before: 662 this patch: 662
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success Fixes tag looks correct
netdev/build_allmodconfig_warn success Errors and warnings before: 725 this patch: 725
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 10 lines checked
netdev/build_clang_rust success No Rust files in patch. Skipping build
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/source_inline success Was 0 now: 0
bpf/vmtest-bpf-next-PR success PR summary
bpf/vmtest-bpf-next-VM_Test-0 success Logs for Lint
bpf/vmtest-bpf-next-VM_Test-1 success Logs for ShellCheck
bpf/vmtest-bpf-next-VM_Test-2 success Logs for Unittests
bpf/vmtest-bpf-next-VM_Test-3 success Logs for Validate matrix.py
bpf/vmtest-bpf-next-VM_Test-5 success Logs for aarch64-gcc / build-release
bpf/vmtest-bpf-next-VM_Test-4 success Logs for aarch64-gcc / build / build for aarch64 with gcc
bpf/vmtest-bpf-next-VM_Test-10 success Logs for aarch64-gcc / veristat
bpf/vmtest-bpf-next-VM_Test-12 success Logs for s390x-gcc / build-release
bpf/vmtest-bpf-next-VM_Test-9 success Logs for aarch64-gcc / test (test_verifier, false, 360) / test_verifier on aarch64 with gcc
bpf/vmtest-bpf-next-VM_Test-11 success Logs for s390x-gcc / build / build for s390x with gcc
bpf/vmtest-bpf-next-VM_Test-17 success Logs for s390x-gcc / veristat
bpf/vmtest-bpf-next-VM_Test-18 success Logs for set-matrix
bpf/vmtest-bpf-next-VM_Test-19 success Logs for x86_64-gcc / build / build for x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-20 success Logs for x86_64-gcc / build-release
bpf/vmtest-bpf-next-VM_Test-28 success Logs for x86_64-llvm-17 / build / build for x86_64 with llvm-17
bpf/vmtest-bpf-next-VM_Test-29 success Logs for x86_64-llvm-17 / build-release / build for x86_64 with llvm-17-O2
bpf/vmtest-bpf-next-VM_Test-34 success Logs for x86_64-llvm-17 / veristat
bpf/vmtest-bpf-next-VM_Test-35 success Logs for x86_64-llvm-18 / build / build for x86_64 with llvm-18
bpf/vmtest-bpf-next-VM_Test-36 success Logs for x86_64-llvm-18 / build-release / build for x86_64 with llvm-18-O2
bpf/vmtest-bpf-next-VM_Test-42 success Logs for x86_64-llvm-18 / veristat
bpf/vmtest-bpf-next-VM_Test-6 success Logs for aarch64-gcc / test (test_maps, false, 360) / test_maps on aarch64 with gcc
bpf/vmtest-bpf-next-VM_Test-7 success Logs for aarch64-gcc / test (test_progs, false, 360) / test_progs on aarch64 with gcc
bpf/vmtest-bpf-next-VM_Test-8 success Logs for aarch64-gcc / test (test_progs_no_alu32, false, 360) / test_progs_no_alu32 on aarch64 with gcc
bpf/vmtest-bpf-next-VM_Test-26 success Logs for x86_64-gcc / test (test_verifier, false, 360) / test_verifier on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-30 success Logs for x86_64-llvm-17 / test (test_maps, false, 360) / test_maps on x86_64 with llvm-17
bpf/vmtest-bpf-next-VM_Test-33 success Logs for x86_64-llvm-17 / test (test_verifier, false, 360) / test_verifier on x86_64 with llvm-17
bpf/vmtest-bpf-next-VM_Test-37 success Logs for x86_64-llvm-18 / test (test_maps, false, 360) / test_maps on x86_64 with llvm-18
bpf/vmtest-bpf-next-VM_Test-41 success Logs for x86_64-llvm-18 / test (test_verifier, false, 360) / test_verifier on x86_64 with llvm-18
bpf/vmtest-bpf-next-VM_Test-21 success Logs for x86_64-gcc / test (test_maps, false, 360) / test_maps on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-22 success Logs for x86_64-gcc / test (test_progs, false, 360) / test_progs on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-25 success Logs for x86_64-gcc / test (test_progs_parallel, true, 30) / test_progs_parallel on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-24 success Logs for x86_64-gcc / test (test_progs_no_alu32_parallel, true, 30) / test_progs_no_alu32_parallel on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-23 success Logs for x86_64-gcc / test (test_progs_no_alu32, false, 360) / test_progs_no_alu32 on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-27 success Logs for x86_64-gcc / veristat / veristat on x86_64 with gcc
bpf/vmtest-bpf-next-VM_Test-31 success Logs for x86_64-llvm-17 / test (test_progs, false, 360) / test_progs on x86_64 with llvm-17
bpf/vmtest-bpf-next-VM_Test-32 success Logs for x86_64-llvm-17 / test (test_progs_no_alu32, false, 360) / test_progs_no_alu32 on x86_64 with llvm-17
bpf/vmtest-bpf-next-VM_Test-38 success Logs for x86_64-llvm-18 / test (test_progs, false, 360) / test_progs on x86_64 with llvm-18
bpf/vmtest-bpf-next-VM_Test-39 success Logs for x86_64-llvm-18 / test (test_progs_cpuv4, false, 360) / test_progs_cpuv4 on x86_64 with llvm-18
bpf/vmtest-bpf-next-VM_Test-40 success Logs for x86_64-llvm-18 / test (test_progs_no_alu32, false, 360) / test_progs_no_alu32 on x86_64 with llvm-18
bpf/vmtest-bpf-next-VM_Test-16 success Logs for s390x-gcc / test (test_verifier, false, 360) / test_verifier on s390x with gcc
bpf/vmtest-bpf-next-VM_Test-14 success Logs for s390x-gcc / test (test_progs, false, 360) / test_progs on s390x with gcc
bpf/vmtest-bpf-next-VM_Test-13 success Logs for s390x-gcc / test (test_maps, false, 360) / test_maps on s390x with gcc
bpf/vmtest-bpf-next-VM_Test-15 success Logs for s390x-gcc / test (test_progs_no_alu32, false, 360) / test_progs_no_alu32 on s390x with gcc

Commit Message

Leon Hwang July 25, 2024, 12:32 a.m. UTC 
The commit f7866c3587337731 ("bpf: Fix null pointer dereference in
resolve_prog_type() for BPF_PROG_TYPE_EXT") fixed the following panic,
which was caused by updating attached freplace prog to PROG_ARRAY map.

But, it does not support updating attached freplace prog to PROG_ARRAY
map.

[309049.036402] BUG: kernel NULL pointer dereference, address: 0000000000000004
[309049.036419] #PF: supervisor read access in kernel mode
[309049.036426] #PF: error_code(0x0000) - not-present page
[309049.036432] PGD 0 P4D 0
[309049.036437] Oops: 0000 [#1] PREEMPT SMP NOPTI
[309049.036444] CPU: 2 PID: 788148 Comm: test_progs Not tainted 6.8.0-31-generic #31-Ubuntu
[309049.036465] Hardware name: VMware, Inc. VMware20,1/440BX Desktop Reference Platform, BIOS VMW201.00V.21805430.B64.2305221830 05/22/2023
[309049.036477] RIP: 0010:bpf_prog_map_compatible+0x2a/0x140
[309049.036488] Code: 0f 1f 44 00 00 55 48 89 e5 41 57 41 56 49 89 fe 41 55 41 54 53 44 8b 6e 04 48 89 f3 41 83 fd 1c 75 0c 48 8b 46 38 48 8b 40 70 <44> 8b 68 04 f6 43 03 01 75 1c 48 8b 43 38 44 0f b6 a0 89 00 00 00
[309049.036505] RSP: 0018:ffffb2e080fd7ce0 EFLAGS: 00010246
[309049.036513] RAX: 0000000000000000 RBX: ffffb2e0807c1000 RCX: 0000000000000000
[309049.036521] RDX: 0000000000000000 RSI: ffffb2e0807c1000 RDI: ffff990290259e00
[309049.036528] RBP: ffffb2e080fd7d08 R08: 0000000000000000 R09: 0000000000000000
[309049.036536] R10: 0000000000000000 R11: 0000000000000000 R12: ffff990290259e00
[309049.036543] R13: 000000000000001c R14: ffff990290259e00 R15: ffff99028e29c400
[309049.036551] FS:  00007b82cbc28140(0000) GS:ffff9903b3f00000(0000) knlGS:0000000000000000
[309049.036559] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[309049.036566] CR2: 0000000000000004 CR3: 0000000101286002 CR4: 00000000003706f0
[309049.036573] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[309049.036581] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[309049.036588] Call Trace:
[309049.036592]  <TASK>
[309049.036597]  ? show_regs+0x6d/0x80
[309049.036604]  ? __die+0x24/0x80
[309049.036619]  ? page_fault_oops+0x99/0x1b0
[309049.036628]  ? do_user_addr_fault+0x2ee/0x6b0
[309049.036634]  ? exc_page_fault+0x83/0x1b0
[309049.036641]  ? asm_exc_page_fault+0x27/0x30
[309049.036649]  ? bpf_prog_map_compatible+0x2a/0x140
[309049.036656]  prog_fd_array_get_ptr+0x2c/0x70
[309049.036664]  bpf_fd_array_map_update_elem+0x37/0x130
[309049.036671]  bpf_map_update_value+0x1d3/0x260
[309049.036677]  map_update_elem+0x1fa/0x360
[309049.036683]  __sys_bpf+0x54c/0xa10
[309049.036689]  __x64_sys_bpf+0x1a/0x30
[309049.036694]  x64_sys_call+0x1936/0x25c0
[309049.036700]  do_syscall_64+0x7f/0x180
[309049.036706]  ? do_syscall_64+0x8c/0x180
[309049.036712]  ? do_syscall_64+0x8c/0x180
[309049.036717]  ? irqentry_exit+0x43/0x50
[309049.036723]  ? common_interrupt+0x54/0xb0
[309049.036729]  entry_SYSCALL_64_after_hwframe+0x73/0x7b

Since commit 1c123c567fb138eb ("bpf: Resolve fext program type when
checking map compatibility"), freplace prog can be used as tail-callee
of its target prog.
And the commit 3aac1ead5eb6b76f ("bpf: Move prog->aux->linked_prog and
trampoline into bpf_link on attach") sets prog->aux->dst_prog as NULL
when attach freplace prog to its target.

Then, as for following example:

tailcall_freplace.c:

// SPDX-License-Identifier: GPL-2.0

\#include <linux/bpf.h>
\#include <bpf/bpf_helpers.h>
\#include "bpf_legacy.h"

struct {
	__uint(type, BPF_MAP_TYPE_PROG_ARRAY);
	__uint(max_entries, 1);
	__uint(key_size, sizeof(__u32));
	__uint(value_size, sizeof(__u32));
} jmp_table SEC(".maps");

int count = 0;

__noinline int
subprog(struct __sk_buff *skb)
{
	volatile int ret = 1;

	count++;

	bpf_tail_call_static(skb, &jmp_table, 0);

	return ret;
}

SEC("freplace")
int entry(struct __sk_buff *skb)
{
	return subprog(skb);
}

char __license[] SEC("license") = "GPL";

tc_bpf2bpf.c:

// SPDX-License-Identifier: GPL-2.0

\#include <linux/bpf.h>
\#include <bpf/bpf_helpers.h>
\#include "bpf_legacy.h"

__noinline int
subprog(struct __sk_buff *skb)
{
	volatile int ret = 1;

	return ret;
}

SEC("tc")
int entry(struct __sk_buff *skb)
{
	return subprog(skb);
}

char __license[] SEC("license") = "GPL";

And freplace entry prog's target is the tc subprog.

After loading, the freplace jmp_table's owner type is
BPF_PROG_TYPE_SCHED_CLS.

Next, after attaching freplace prog to tc subprog, its prog->aux->
dst_prog is NULL.

Next, when update freplace prog to jmp_table, bpf_prog_map_compatible()
returns false because resolve_prog_type() returns BPF_PROG_TYPE_EXT instead
of BPF_PROG_TYPE_SCHED_CLS.

With this patch, resolve_prog_type() returns BPF_PROG_TYPE_SCHED_CLS to
support updating attached freplace prog to PROG_ARRY map for this
example.

Fixes: f7866c358733 ("bpf: Fix null pointer dereference in resolve_prog_type() for BPF_PROG_TYPE_EXT")
Cc: Toke Høiland-Jørgensen <toke@redhat.com>
Cc: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
---
 include/linux/bpf_verifier.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Yonghong Song July 25, 2024, 8:58 p.m. UTC | #1 
On 7/24/24 5:32 PM, Leon Hwang wrote:
> The commit f7866c3587337731 ("bpf: Fix null pointer dereference in
> resolve_prog_type() for BPF_PROG_TYPE_EXT") fixed the following panic,
> which was caused by updating attached freplace prog to PROG_ARRAY map.

I am confused here. You mentioned that commit f7866c3587337731
fixed the panic below. But looking at commit message:
   https://lore.kernel.org/bpf/20240711145819.254178-2-wutengda@huaweicloud.com
it does not seem the case.

>
> But, it does not support updating attached freplace prog to PROG_ARRAY
> map.

This seems true since this patch itself intends fixing this issue.

>
> [309049.036402] BUG: kernel NULL pointer dereference, address: 0000000000000004
> [309049.036419] #PF: supervisor read access in kernel mode
> [309049.036426] #PF: error_code(0x0000) - not-present page
> [309049.036432] PGD 0 P4D 0
> [309049.036437] Oops: 0000 [#1] PREEMPT SMP NOPTI
> [309049.036444] CPU: 2 PID: 788148 Comm: test_progs Not tainted 6.8.0-31-generic #31-Ubuntu
> [309049.036465] Hardware name: VMware, Inc. VMware20,1/440BX Desktop Reference Platform, BIOS VMW201.00V.21805430.B64.2305221830 05/22/2023
> [309049.036477] RIP: 0010:bpf_prog_map_compatible+0x2a/0x140
> [309049.036488] Code: 0f 1f 44 00 00 55 48 89 e5 41 57 41 56 49 89 fe 41 55 41 54 53 44 8b 6e 04 48 89 f3 41 83 fd 1c 75 0c 48 8b 46 38 48 8b 40 70 <44> 8b 68 04 f6 43 03 01 75 1c 48 8b 43 38 44 0f b6 a0 89 00 00 00
> [309049.036505] RSP: 0018:ffffb2e080fd7ce0 EFLAGS: 00010246
> [309049.036513] RAX: 0000000000000000 RBX: ffffb2e0807c1000 RCX: 0000000000000000
> [309049.036521] RDX: 0000000000000000 RSI: ffffb2e0807c1000 RDI: ffff990290259e00
> [309049.036528] RBP: ffffb2e080fd7d08 R08: 0000000000000000 R09: 0000000000000000
> [309049.036536] R10: 0000000000000000 R11: 0000000000000000 R12: ffff990290259e00
> [309049.036543] R13: 000000000000001c R14: ffff990290259e00 R15: ffff99028e29c400
> [309049.036551] FS:  00007b82cbc28140(0000) GS:ffff9903b3f00000(0000) knlGS:0000000000000000
> [309049.036559] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [309049.036566] CR2: 0000000000000004 CR3: 0000000101286002 CR4: 00000000003706f0
> [309049.036573] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [309049.036581] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> [309049.036588] Call Trace:
> [309049.036592]  <TASK>
> [309049.036597]  ? show_regs+0x6d/0x80
> [309049.036604]  ? __die+0x24/0x80
> [309049.036619]  ? page_fault_oops+0x99/0x1b0
> [309049.036628]  ? do_user_addr_fault+0x2ee/0x6b0
> [309049.036634]  ? exc_page_fault+0x83/0x1b0
> [309049.036641]  ? asm_exc_page_fault+0x27/0x30
> [309049.036649]  ? bpf_prog_map_compatible+0x2a/0x140
> [309049.036656]  prog_fd_array_get_ptr+0x2c/0x70
> [309049.036664]  bpf_fd_array_map_update_elem+0x37/0x130
> [309049.036671]  bpf_map_update_value+0x1d3/0x260
> [309049.036677]  map_update_elem+0x1fa/0x360
> [309049.036683]  __sys_bpf+0x54c/0xa10
> [309049.036689]  __x64_sys_bpf+0x1a/0x30
> [309049.036694]  x64_sys_call+0x1936/0x25c0
> [309049.036700]  do_syscall_64+0x7f/0x180
> [309049.036706]  ? do_syscall_64+0x8c/0x180
> [309049.036712]  ? do_syscall_64+0x8c/0x180
> [309049.036717]  ? irqentry_exit+0x43/0x50
> [309049.036723]  ? common_interrupt+0x54/0xb0
> [309049.036729]  entry_SYSCALL_64_after_hwframe+0x73/0x7b

I actually tried your selftest (patch 2/2) without patch 1/1, I got the
following error:

All error logs:
tester_init:PASS:tester_log_buf 0 nsec
process_subtest:PASS:obj_open_mem 0 nsec
process_subtest:PASS:specs_alloc 0 nsec
test_tailcall_freplace:PASS:open fr_skel 0 nsec
test_tailcall_freplace:PASS:open tc_skel 0 nsec
test_tailcall_freplace:PASS:tc_skel entry prog_id 0 nsec
test_tailcall_freplace:PASS:set_attach_target 0 nsec
test_tailcall_freplace:PASS:load fr_skel 0 nsec
test_tailcall_freplace:PASS:attach_freplace 0 nsec
test_tailcall_freplace:PASS:fr_skel entry prog_fd 0 nsec
test_tailcall_freplace:PASS:fr_skel jmp_table map_fd 0 nsec
test_tailcall_freplace:FAIL:update jmp_table unexpected error: -22 (errno 22)
#328/25  tailcalls/tailcall_freplace:FAIL
#328     tailcalls:FAIL

I didn't see kernel panic.

>
> Since commit 1c123c567fb138eb ("bpf: Resolve fext program type when
> checking map compatibility"), freplace prog can be used as tail-callee
> of its target prog.

the tailcall target can be a freplace prog.

> And the commit 3aac1ead5eb6b76f ("bpf: Move prog->aux->linked_prog and
> trampoline into bpf_link on attach") sets prog->aux->dst_prog as NULL
> when attach freplace prog to its target.

when attach -> after attaching

>
> Then, as for following example:
>
> tailcall_freplace.c:
>
> // SPDX-License-Identifier: GPL-2.0
>
> \#include <linux/bpf.h>
> \#include <bpf/bpf_helpers.h>
> \#include "bpf_legacy.h"
>
> struct {
> 	__uint(type, BPF_MAP_TYPE_PROG_ARRAY);
> 	__uint(max_entries, 1);
> 	__uint(key_size, sizeof(__u32));
> 	__uint(value_size, sizeof(__u32));
> } jmp_table SEC(".maps");
>
> int count = 0;
>
> __noinline int
> subprog(struct __sk_buff *skb)
> {
> 	volatile int ret = 1;
>
> 	count++;
>
> 	bpf_tail_call_static(skb, &jmp_table, 0);
>
> 	return ret;
> }

This subprog is not needed and could be misleading,
just inline subprog into entry prog, it should be okay.

>
> SEC("freplace")
> int entry(struct __sk_buff *skb)
> {
> 	return subprog(skb);
> }
>
> char __license[] SEC("license") = "GPL";
>
> tc_bpf2bpf.c:
>
> // SPDX-License-Identifier: GPL-2.0
>
> \#include <linux/bpf.h>
> \#include <bpf/bpf_helpers.h>
> \#include "bpf_legacy.h"
>
> __noinline int
> subprog(struct __sk_buff *skb)
> {
> 	volatile int ret = 1;
>
> 	return ret;
> }
>
> SEC("tc")
> int entry(struct __sk_buff *skb)
> {
> 	return subprog(skb);
> }
>
> char __license[] SEC("license") = "GPL";
>
> And freplace entry prog's target is the tc subprog.
>
> After loading, the freplace jmp_table's owner type is
> BPF_PROG_TYPE_SCHED_CLS.
>
> Next, after attaching freplace prog to tc subprog, its prog->aux->
> dst_prog is NULL.
>
> Next, when update freplace prog to jmp_table, bpf_prog_map_compatible()
> returns false because resolve_prog_type() returns BPF_PROG_TYPE_EXT instead
> of BPF_PROG_TYPE_SCHED_CLS.
>
> With this patch, resolve_prog_type() returns BPF_PROG_TYPE_SCHED_CLS to
> support updating attached freplace prog to PROG_ARRY map for this
> example.
>
> Fixes: f7866c358733 ("bpf: Fix null pointer dereference in resolve_prog_type() for BPF_PROG_TYPE_EXT")
> Cc: Toke Høiland-Jørgensen <toke@redhat.com>
> Cc: Martin KaFai Lau <martin.lau@kernel.org>
> Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
> ---
>   include/linux/bpf_verifier.h | 4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
> index 5cea15c81b8a8..387e034e73d0e 100644
> --- a/include/linux/bpf_verifier.h
> +++ b/include/linux/bpf_verifier.h
> @@ -874,8 +874,8 @@ static inline u32 type_flag(u32 type)
>   /* only use after check_attach_btf_id() */
>   static inline enum bpf_prog_type resolve_prog_type(const struct bpf_prog *prog)
>   {
> -	return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->dst_prog) ?
> -		prog->aux->dst_prog->type : prog->type;
> +	return prog->type == BPF_PROG_TYPE_EXT ?
> +		prog->aux->saved_dst_prog_type : prog->type;

If prog->aux->dst_prog is NULL, is it possible that prog->aux->saved_dst_prog_type
(0, corresponding to BPF_PROG_TYPE_UNSPEC) could be returned? Do we need to do
	return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->saved_dst_prog_type) ?
		prog->aux->saved_dst_prog_type : prog->type;

Maybe I missed something here?

>   }
>   
>   static inline bool bpf_prog_check_recur(const struct bpf_prog *prog)
Leon Hwang July 26, 2024, 3:27 a.m. UTC | #2 
26 July 2024 at 04:58, "Yonghong Song" <yonghong.song@linux.dev> wrote:



> 
> On 7/24/24 5:32 PM, Leon Hwang wrote:
> 
> > 
> > The commit f7866c3587337731 ("bpf: Fix null pointer dereference in
> > 
> >  resolve_prog_type() for BPF_PROG_TYPE_EXT") fixed the following panic,
> > 
> >  which was caused by updating attached freplace prog to PROG_ARRAY map.
> > 
> 
> I am confused here. You mentioned that commit f7866c3587337731
> 
> fixed the panic below. But looking at commit message:
> 
>  https://lore.kernel.org/bpf/20240711145819.254178-2-wutengda@huaweicloud.com
> 
> it does not seem the case.

The commit fixed this panic meanwhile.

This panic seems confusing. I'll remove it in patch v2.

> 
> > 
> > But, it does not support updating attached freplace prog to PROG_ARRAY
> > 
> >  map.
> > 
> 
> This seems true since this patch itself intends fixing this issue.

Yes, it is to fix this issue.

> 
> > 
> > [309049.036402] BUG: kernel NULL pointer dereference, address: 0000000000000004
> > 
> >  [309049.036419] #PF: supervisor read access in kernel mode
> > 
> >  [309049.036426] #PF: error_code(0x0000) - not-present page
> > 
> >  [309049.036432] PGD 0 P4D 0
> > 
> >  [309049.036437] Oops: 0000 [#1] PREEMPT SMP NOPTI
> > 
> >  [309049.036444] CPU: 2 PID: 788148 Comm: test_progs Not tainted 6.8.0-31-generic #31-Ubuntu
> > 
> >  [309049.036465] Hardware name: VMware, Inc. VMware20,1/440BX Desktop Reference Platform, BIOS VMW201.00V.21805430.B64.2305221830 05/22/2023
> > 
> >  [309049.036477] RIP: 0010:bpf_prog_map_compatible+0x2a/0x140
> > 
> >  [309049.036488] Code: 0f 1f 44 00 00 55 48 89 e5 41 57 41 56 49 89 fe 41 55 41 54 53 44 8b 6e 04 48 89 f3 41 83 fd 1c 75 0c 48 8b 46 38 48 8b 40 70 <44> 8b 68 04 f6 43 03 01 75 1c 48 8b 43 38 44 0f b6 a0 89 00 00 00
> > 
> >  [309049.036505] RSP: 0018:ffffb2e080fd7ce0 EFLAGS: 00010246
> > 
> >  [309049.036513] RAX: 0000000000000000 RBX: ffffb2e0807c1000 RCX: 0000000000000000
> > 
> >  [309049.036521] RDX: 0000000000000000 RSI: ffffb2e0807c1000 RDI: ffff990290259e00
> > 
> >  [309049.036528] RBP: ffffb2e080fd7d08 R08: 0000000000000000 R09: 0000000000000000
> > 
> >  [309049.036536] R10: 0000000000000000 R11: 0000000000000000 R12: ffff990290259e00
> > 
> >  [309049.036543] R13: 000000000000001c R14: ffff990290259e00 R15: ffff99028e29c400
> > 
> >  [309049.036551] FS: 00007b82cbc28140(0000) GS:ffff9903b3f00000(0000) knlGS:0000000000000000
> > 
> >  [309049.036559] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > 
> >  [309049.036566] CR2: 0000000000000004 CR3: 0000000101286002 CR4: 00000000003706f0
> > 
> >  [309049.036573] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> > 
> >  [309049.036581] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> > 
> >  [309049.036588] Call Trace:
> > 
> >  [309049.036592] <TASK>
> > 
> >  [309049.036597] ? show_regs+0x6d/0x80
> > 
> >  [309049.036604] ? __die+0x24/0x80
> > 
> >  [309049.036619] ? page_fault_oops+0x99/0x1b0
> > 
> >  [309049.036628] ? do_user_addr_fault+0x2ee/0x6b0
> > 
> >  [309049.036634] ? exc_page_fault+0x83/0x1b0
> > 
> >  [309049.036641] ? asm_exc_page_fault+0x27/0x30
> > 
> >  [309049.036649] ? bpf_prog_map_compatible+0x2a/0x140
> > 
> >  [309049.036656] prog_fd_array_get_ptr+0x2c/0x70
> > 
> >  [309049.036664] bpf_fd_array_map_update_elem+0x37/0x130
> > 
> >  [309049.036671] bpf_map_update_value+0x1d3/0x260
> > 
> >  [309049.036677] map_update_elem+0x1fa/0x360
> > 
> >  [309049.036683] __sys_bpf+0x54c/0xa10
> > 
> >  [309049.036689] __x64_sys_bpf+0x1a/0x30
> > 
> >  [309049.036694] x64_sys_call+0x1936/0x25c0
> > 
> >  [309049.036700] do_syscall_64+0x7f/0x180
> > 
> >  [309049.036706] ? do_syscall_64+0x8c/0x180
> > 
> >  [309049.036712] ? do_syscall_64+0x8c/0x180
> > 
> >  [309049.036717] ? irqentry_exit+0x43/0x50
> > 
> >  [309049.036723] ? common_interrupt+0x54/0xb0
> > 
> >  [309049.036729] entry_SYSCALL_64_after_hwframe+0x73/0x7b
> > 
> 
> I actually tried your selftest (patch 2/2) without patch 1/1, I got the
> 
> following error:
> 
> All error logs:
> 
> tester_init:PASS:tester_log_buf 0 nsec
> 
> process_subtest:PASS:obj_open_mem 0 nsec
> 
> process_subtest:PASS:specs_alloc 0 nsec
> 
> test_tailcall_freplace:PASS:open fr_skel 0 nsec
> 
> test_tailcall_freplace:PASS:open tc_skel 0 nsec
> 
> test_tailcall_freplace:PASS:tc_skel entry prog_id 0 nsec
> 
> test_tailcall_freplace:PASS:set_attach_target 0 nsec
> 
> test_tailcall_freplace:PASS:load fr_skel 0 nsec
> 
> test_tailcall_freplace:PASS:attach_freplace 0 nsec
> 
> test_tailcall_freplace:PASS:fr_skel entry prog_fd 0 nsec
> 
> test_tailcall_freplace:PASS:fr_skel jmp_table map_fd 0 nsec
> 
> test_tailcall_freplace:FAIL:update jmp_table unexpected error: -22 (errno 22)
> 
> #328/25 tailcalls/tailcall_freplace:FAIL
> 
> #328 tailcalls:FAIL
> 
> I didn't see kernel panic.

Indeed.

> 
> > 
> > Since commit 1c123c567fb138eb ("bpf: Resolve fext program type when
> > 
> >  checking map compatibility"), freplace prog can be used as tail-callee
> > 
> >  of its target prog.
> > 
> 
> the tailcall target can be a freplace prog.

Ack.

> 
> > 
> > And the commit 3aac1ead5eb6b76f ("bpf: Move prog->aux->linked_prog and
> > 
> >  trampoline into bpf_link on attach") sets prog->aux->dst_prog as NULL
> > 
> >  when attach freplace prog to its target.
> > 
> 
> when attach -> after attaching

Ack.

> 
> > 
> > Then, as for following example:
> > 
> >  tailcall_freplace.c:
> > 
> >  // SPDX-License-Identifier: GPL-2.0
> > 
> >  \#include <linux/bpf.h>
> > 
> >  \#include <bpf/bpf_helpers.h>
> > 
> >  \#include "bpf_legacy.h"
> > 
> >  struct {
> > 
> >  __uint(type, BPF_MAP_TYPE_PROG_ARRAY);
> > 
> >  __uint(max_entries, 1);
> > 
> >  __uint(key_size, sizeof(__u32));
> > 
> >  __uint(value_size, sizeof(__u32));
> > 
> >  } jmp_table SEC(".maps");
> > 
> >  int count = 0;
> > 
> >  __noinline int
> > 
> >  subprog(struct __sk_buff *skb)
> > 
> >  {
> > 
> >  volatile int ret = 1;
> > 
> >  count++;
> > 
> >  bpf_tail_call_static(skb, &jmp_table, 0);
> > 
> >  return ret;
> > 
> >  }
> > 
> 
> This subprog is not needed and could be misleading,
> 
> just inline subprog into entry prog, it should be okay.

Ack.

> 
> > 
> > SEC("freplace")
> > 
> >  int entry(struct __sk_buff *skb)
> > 
> >  {
> > 
> >  return subprog(skb);
> > 
> >  }
> > 
> >  char __license[] SEC("license") = "GPL";
> > 
> >  tc_bpf2bpf.c:
> > 
> >  // SPDX-License-Identifier: GPL-2.0
> > 
> >  \#include <linux/bpf.h>
> > 
> >  \#include <bpf/bpf_helpers.h>
> > 
> >  \#include "bpf_legacy.h"
> > 
> >  __noinline int
> > 
> >  subprog(struct __sk_buff *skb)
> > 
> >  {
> > 
> >  volatile int ret = 1;
> > 
> >  return ret;
> > 
> >  }
> > 
> >  SEC("tc")
> > 
> >  int entry(struct __sk_buff *skb)
> > 
> >  {
> > 
> >  return subprog(skb);
> > 
> >  }
> > 
> >  char __license[] SEC("license") = "GPL";
> > 
> >  And freplace entry prog's target is the tc subprog.
> > 
> >  After loading, the freplace jmp_table's owner type is
> > 
> >  BPF_PROG_TYPE_SCHED_CLS.
> > 
> >  Next, after attaching freplace prog to tc subprog, its prog->aux->
> > 
> >  dst_prog is NULL.
> > 
> >  Next, when update freplace prog to jmp_table, bpf_prog_map_compatible()
> > 
> >  returns false because resolve_prog_type() returns BPF_PROG_TYPE_EXT instead
> > 
> >  of BPF_PROG_TYPE_SCHED_CLS.
> > 
> >  With this patch, resolve_prog_type() returns BPF_PROG_TYPE_SCHED_CLS to
> > 
> >  support updating attached freplace prog to PROG_ARRY map for this
> > 
> >  example.
> > 
> >  Fixes: f7866c358733 ("bpf: Fix null pointer dereference in resolve_prog_type() for BPF_PROG_TYPE_EXT")
> > 
> >  Cc: Toke Høiland-Jørgensen <toke@redhat.com>
> > 
> >  Cc: Martin KaFai Lau <martin.lau@kernel.org>
> > 
> >  Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
> > 
> >  ---
> > 
> >  include/linux/bpf_verifier.h | 4 ++--
> > 
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> >  diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
> > 
> >  index 5cea15c81b8a8..387e034e73d0e 100644
> > 
> >  --- a/include/linux/bpf_verifier.h
> > 
> >  +++ b/include/linux/bpf_verifier.h
> > 
> >  @@ -874,8 +874,8 @@ static inline u32 type_flag(u32 type)
> > 
> >  /* only use after check_attach_btf_id() */
> > 
> >  static inline enum bpf_prog_type resolve_prog_type(const struct bpf_prog *prog)
> > 
> >  {
> > 
> >  - return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->dst_prog) ?
> > 
> >  - prog->aux->dst_prog->type : prog->type;
> > 
> >  + return prog->type == BPF_PROG_TYPE_EXT ?
> > 
> >  + prog->aux->saved_dst_prog_type : prog->type;
> > 
> 
> If prog->aux->dst_prog is NULL, is it possible that prog->aux->saved_dst_prog_type
> 
> (0, corresponding to BPF_PROG_TYPE_UNSPEC) could be returned? Do we need to do
> 
>  return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->saved_dst_prog_type) ?
> 
>  prog->aux->saved_dst_prog_type : prog->type;
> 
> Maybe I missed something here?

It seems better to check prog->aux->saved_dst_prog_type. But I don't think so.

prog->aux->saved_dst_prog_type is set in check_attach_btf_id(). And there is no
resolve_prog_type() before check_attach_btf_id() in bpf_check().

Therefore, resolve_prog_type() must be called after check_attach_btf_id().

Thanks,
Leon

> 
> > 
> > }
> > 
> >  > static inline bool bpf_prog_check_recur(const struct bpf_prog *prog)
> >
>
Yonghong Song July 26, 2024, 6:15 a.m. UTC | #3 
On 7/25/24 8:27 PM, leon.hwang@linux.dev wrote:
> 26 July 2024 at 04:58, "Yonghong Song" <yonghong.song@linux.dev> wrote:
>
>
>
>> On 7/24/24 5:32 PM, Leon Hwang wrote:
>>
>>> The commit f7866c3587337731 ("bpf: Fix null pointer dereference in
>>>
>>>   resolve_prog_type() for BPF_PROG_TYPE_EXT") fixed the following panic,
>>>
>>>   which was caused by updating attached freplace prog to PROG_ARRAY map.
>>>
>> I am confused here. You mentioned that commit f7866c3587337731
>>
>> fixed the panic below. But looking at commit message:
>>
>>   https://lore.kernel.org/bpf/20240711145819.254178-2-wutengda@huaweicloud.com
>>
>> it does not seem the case.
> The commit fixed this panic meanwhile.
>
> This panic seems confusing. I'll remove it in patch v2.
>
[...]

>>>   ---
>>>
>>>   include/linux/bpf_verifier.h | 4 ++--
>>>
>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>>   diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
>>>
>>>   index 5cea15c81b8a8..387e034e73d0e 100644
>>>
>>>   --- a/include/linux/bpf_verifier.h
>>>
>>>   +++ b/include/linux/bpf_verifier.h
>>>
>>>   @@ -874,8 +874,8 @@ static inline u32 type_flag(u32 type)
>>>
>>>   /* only use after check_attach_btf_id() */
>>>
>>>   static inline enum bpf_prog_type resolve_prog_type(const struct bpf_prog *prog)
>>>
>>>   {
>>>
>>>   - return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->dst_prog) ?
>>>
>>>   - prog->aux->dst_prog->type : prog->type;
>>>
>>>   + return prog->type == BPF_PROG_TYPE_EXT ?
>>>
>>>   + prog->aux->saved_dst_prog_type : prog->type;
>>>
>> If prog->aux->dst_prog is NULL, is it possible that prog->aux->saved_dst_prog_type
>>
>> (0, corresponding to BPF_PROG_TYPE_UNSPEC) could be returned? Do we need to do
>>
>>   return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->saved_dst_prog_type) ?
>>
>>   prog->aux->saved_dst_prog_type : prog->type;
>>
>> Maybe I missed something here?
> It seems better to check prog->aux->saved_dst_prog_type. But I don't think so.
>
> prog->aux->saved_dst_prog_type is set in check_attach_btf_id(). And there is no
> resolve_prog_type() before check_attach_btf_id() in bpf_check().
>
> Therefore, resolve_prog_type() must be called after check_attach_btf_id().

In check_attach_btf_id(), I see
         if (tgt_prog) {
                 prog->aux->saved_dst_prog_type = tgt_prog->type;
                 prog->aux->saved_dst_attach_type = tgt_prog->expected_attach_type;
         }

So it is possible prog->aux->saved_dst_prog_type is 0 (default value).
I don't know that if tgt_prog is NULL, whether later resolve_prog_type()
will be called or not. Need more checking here.
Leon Hwang July 26, 2024, 7:31 a.m. UTC | #4 
26 July 2024 at 14:15, "Yonghong Song" <yonghong.song@linux.dev> wrote:

> 
> On 7/25/24 8:27 PM, leon.hwang@linux.dev wrote:
> 
> > 
> > 26 July 2024 at 04:58, "Yonghong Song" <yonghong.song@linux.dev> wrote:
> > 
> > > 
> > > On 7/24/24 5:32 PM, Leon Hwang wrote:
> > > 
> > 
> >  The commit f7866c3587337731 ("bpf: Fix null pointer dereference in
> > 
> >  resolve_prog_type() for BPF_PROG_TYPE_EXT") fixed the following panic,
> > 
> >  which was caused by updating attached freplace prog to PROG_ARRAY map.
> > 
> > > 
> > > I am confused here. You mentioned that commit f7866c3587337731
> > > 
> > >  fixed the panic below. But looking at commit message:
> > > 
> > >  https://lore.kernel.org/bpf/20240711145819.254178-2-wutengda@huaweicloud.com
> > > 
> > >  it does not seem the case.
> > > 
> > 
> >  The commit fixed this panic meanwhile.
> > 
> >  This panic seems confusing. I'll remove it in patch v2.
> > 
> 
> [...]
> 
> > 
> > ---
> > 
> >  include/linux/bpf_verifier.h | 4 ++--
> > 
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> >  diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
> > 
> >  index 5cea15c81b8a8..387e034e73d0e 100644
> > 
> >  --- a/include/linux/bpf_verifier.h
> > 
> >  +++ b/include/linux/bpf_verifier.h
> > 
> >  @@ -874,8 +874,8 @@ static inline u32 type_flag(u32 type)
> > 
> >  /* only use after check_attach_btf_id() */
> > 
> >  static inline enum bpf_prog_type resolve_prog_type(const struct bpf_prog *prog)
> > 
> >  {
> > 
> >  - return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->dst_prog) ?
> > 
> >  - prog->aux->dst_prog->type : prog->type;
> > 
> >  + return prog->type == BPF_PROG_TYPE_EXT ?
> > 
> >  + prog->aux->saved_dst_prog_type : prog->type;
> > 
> > > 
> > > If prog->aux->dst_prog is NULL, is it possible that prog->aux->saved_dst_prog_type
> > > 
> > >  (0, corresponding to BPF_PROG_TYPE_UNSPEC) could be returned? Do we need to do
> > > 
> > >  return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->saved_dst_prog_type) ?
> > > 
> > >  prog->aux->saved_dst_prog_type : prog->type;
> > > 
> > >  Maybe I missed something here?
> > > 
> > 
> >  It seems better to check prog->aux->saved_dst_prog_type. But I don't think so.
> > 
> >  prog->aux->saved_dst_prog_type is set in check_attach_btf_id(). And there is no
> > 
> >  resolve_prog_type() before check_attach_btf_id() in bpf_check().
> > 
> >  Therefore, resolve_prog_type() must be called after check_attach_btf_id().
> > 
> 
> In check_attach_btf_id(), I see
> 
>  if (tgt_prog) {
> 
>  prog->aux->saved_dst_prog_type = tgt_prog->type;
> 
>  prog->aux->saved_dst_attach_type = tgt_prog->expected_attach_type;
> 
>  }
> 
> So it is possible prog->aux->saved_dst_prog_type is 0 (default value).
> 
> I don't know that if tgt_prog is NULL, whether later resolve_prog_type()
> 
> will be called or not. Need more checking here.
>

This is the case that commit f7866c3587337731 ("bpf: Fix null pointer dereference
in resolve_prog_type() for BPF_PROG_TYPE_EXT") fixed, which is loading freplace
prog without tgt_prog.

With this patch, while loading freplace prog without tgt_prog, resolve_prog_type()
returns 0 instead of BPF_PROG_TYPE_EXT.

It's better to return a meaningful prog type in resolve_prog_type() anyway.

I accept your suggestion.

Thanks,
Leon
diff mbox series

Patch

diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
index 5cea15c81b8a8..387e034e73d0e 100644
--- a/include/linux/bpf_verifier.h
+++ b/include/linux/bpf_verifier.h
@@ -874,8 +874,8 @@  static inline u32 type_flag(u32 type)
 /* only use after check_attach_btf_id() */
 static inline enum bpf_prog_type resolve_prog_type(const struct bpf_prog *prog)
 {
-	return (prog->type == BPF_PROG_TYPE_EXT && prog->aux->dst_prog) ?
-		prog->aux->dst_prog->type : prog->type;
+	return prog->type == BPF_PROG_TYPE_EXT ?
+		prog->aux->saved_dst_prog_type : prog->type;
 }
 
 static inline bool bpf_prog_check_recur(const struct bpf_prog *prog)