Skip to contentSkip to conversation

bpf: Support private stack for bpf progs

RFC PATCH Superseded Yonghong Song 3 participants 6 messages Jul 11, 2024 – Jul 12, 2024

Subsystems (6)
x86 Kselftest BPF BPF selftests +2 more subsystems

v2 of 12 Superseded Latest is v12 →

lore
Share & tools
lore

Yonghong Song

author reviewer
#0 lore raw

RFC PATCH v2 1/2 bpf: Support private stack for bpf progs

The main motivation for private stack comes from nested scheduler in sched-ext from Tejun. The basic idea is that
- each cgroup will its own associated bpf program,
- bpf program with parent cgroup will call bpf programs in immediate child cgroups.

Let us say we have the following cgroup hierarchy:
root_cg (prog0):

    cg1 (prog1):
      cg11 (prog11):
        cg111 (prog111)
        cg112 (prog112)
      cg12 (prog12):
        cg121 (prog121)
        cg122 (prog122)
    cg2 (prog2):
      cg21 (prog21)
      cg22 (prog22)
      cg23 (prog23)

In the above example, prog0 will call a kfunc which will call prog1 and prog2 to get sched info for cg1 and cg2 and then the information is summarized and sent back to prog0. Similarly, prog11 and prog12 will be invoked in the kfunc and the result will be summarized and sent back to prog1, etc.

Currently, for each thread, the x86 kernel allocate 8KB stack. The each bpf program (including its subprograms) has maximum 512B stack size to avoid potential stack overflow. And nested bpf programs increase the risk of stack overflow. To avoid potential stack overflow caused by bpf programs, this patch implemented a private stack so bpf program stack space is allocated dynamically when the program is jited. Such private stack is applied to tracing programs like kprobe/uprobe, perf_event, tracepoint, raw tracepoint and tracing.

But more than one instance of the same bpf program may run in the system. To make things simple, percpu private stack is allocated for each program, so if the same program is running on different cpus concurrently, we won't have any issue. Note that the kernel already have logic to prevent the recursion for the same bpf program on the same cpu (kprobe, fentry, etc.).

The patch implemented a percpu private stack based approach for x86 arch.
- The stack size will be 0 and any stack access is from jit-time allocated percpu storage.
- In the beginning of jit, r9 is used to save percpu private stack pointer.
- Each rbp in the bpf asm insn is replaced by r9.
- For each call, push r9 before the call and pop r9 after the call to preserve r9 value.

Compared to previous RFC patch [1], this patch added some conditions to enable private stack, e.g., verifier calculated stack size, prog type, etc. The new patch also added a performance test to compare private stack vs. no private stack.

The following are some code example to illustrate the idea for selftest cgroup_skb_sk_lookup:

   the existing code                        the private-stack approach code
   endbr64                                  endbr64
   nop    DWORD PTR [rax+rax*1+0x0]         nop    DWORD PTR [rax+rax*1+0x0]
   xchg   ax,ax                             xchg   ax,ax
   push   rbp                               push   rbp
   mov    rbp,rsp                           mov    rbp,rsp
   endbr64                                  endbr64
   sub    rsp,0x68
   push   rbx                               push   rbx
   ...                                      ...
   ...                                      mov    r9d,0x8c1c860
   ...                                      add    r9,QWORD PTR gs:0x21a00
   ...                                      ...
   mov    rdx,rbp                           mov    rdx, r9
   add    rdx,0xffffffffffffffb4            rdx,0xffffffffffffffb4
   ...                                      ...
   mov    ecx,0x28                          mov    ecx,0x28
                                            push   r9
   call   0xffffffffe305e474                call   0xffffffffe305e524
                                            pop    r9
   mov    rdi,rax                           mov    rdi,rax
   ...                                      ...
   movzx  rdi,BYTE PTR [rbp-0x46]           movzx  rdi,BYTE PTR [r9-0x46]
   ...                                      ...

So the number of insns is increased by 1 + num_of_calls * 2. Here the number of calls are those calls in the final jited binary. Comparing function call itself, the push/pop overhead should be minimum in most common cases.

Question:
Note that I didn't add struct_ops bpf prgram in the list. We probably should for the use case I mentioned in the above for nested scheduler. But for tcp congestion control, there is no need to use private stack since there is no nesting there. I guess I can add struct_ops for sched-ext only once it is merged.

[1] https://lore.kernel.org/bpf/[email protected]/T/

---
 arch/x86/net/bpf_jit_comp.c | 63 ++++++++++++++++++++++++++++++++++---
 include/linux/bpf.h         |  2 ++
 kernel/bpf/core.c           | 20 ++++++++++++
 kernel/bpf/syscall.c        |  1 +
 4 files changed, 82 insertions(+), 4 deletions(-)
arch/x86/net/bpf_jit_comp.c +59 -4
index d25d81c8ecc0..60f5d86fb6aa 100644
@@ -1309,6 +1309,22 @@ static void emit_shiftx(u8 **pprog, u32 dst_reg, u8 src_reg, bool is64, u8 op)
 	*pprog = prog;
 }
 
+static void emit_private_frame_ptr(u8 **pprog, void *private_frame_ptr)
+{
+	u8 *prog = *pprog;
+
+	/* movabs r9, private_frame_ptr */
+	emit_mov_imm64(&prog, X86_REG_R9, (long) private_frame_ptr >> 32,
+		       (u32) (long) private_frame_ptr);
+
+	/* add <r9>, gs:[<off>] */
+	EMIT2(0x65, 0x4c);
+	EMIT3(0x03, 0x0c, 0x25);
+	EMIT((u32)(unsigned long)&this_cpu_off, 4);
+
+	*pprog = prog;
+}
+
 #define INSN_SZ_DIFF (((addrs[i] - addrs[i - 1]) - (prog - temp)))
 
 /* mov rax, qword ptr [rbp - rounded_stack_depth - 8] */
@@ -1324,18 +1340,25 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image, u8 *rw_image
 	int insn_cnt = bpf_prog->len;
 	bool seen_exit = false;
 	u8 temp[BPF_MAX_INSN_SIZE + BPF_INSN_SAFETY];
+	u32 stack_depth = bpf_prog->aux->stack_depth;
+	void __percpu *private_frame_ptr = NULL;
 	u64 arena_vm_start, user_vm_start;
 	int i, excnt = 0;
 	int ilen, proglen = 0;
 	u8 *prog = temp;
 	int err;
 
+	if (bpf_prog->private_stack_ptr) {
+		private_frame_ptr = bpf_prog->private_stack_ptr + round_up(stack_depth, 8);
+		stack_depth = 0;
+	}
+
 	arena_vm_start = bpf_arena_get_kern_vm_start(bpf_prog->aux->arena);
 	user_vm_start = bpf_arena_get_user_vm_start(bpf_prog->aux->arena);
 
 	detect_reg_usage(insn, insn_cnt, callee_regs_used);
 
-	emit_prologue(&prog, bpf_prog->aux->stack_depth,
+	emit_prologue(&prog, stack_depth,
 		      bpf_prog_was_classic(bpf_prog), tail_call_reachable,
 		      bpf_is_subprog(bpf_prog), bpf_prog->aux->exception_cb);
 	/* Exception callback will clobber callee regs for its own use, and
@@ -1357,6 +1380,9 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image, u8 *rw_image
 		emit_mov_imm64(&prog, X86_REG_R12,
 			       arena_vm_start >> 32, (u32) arena_vm_start);
 
+	if (private_frame_ptr)
+		emit_private_frame_ptr(&prog, private_frame_ptr);
+
 	ilen = prog - temp;
 	if (rw_image)
 		memcpy(rw_image + proglen, temp, ilen);
@@ -1376,6 +1402,14 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image, u8 *rw_image
 		u8 *func;
 		int nops;
 
+		if (private_frame_ptr) {
+			if (src_reg == BPF_REG_FP)
+				src_reg = X86_REG_R9;
+
+			if (dst_reg == BPF_REG_FP)
+				dst_reg = X86_REG_R9;
+		}
+
 		switch (insn->code) {
 			/* ALU */
 		case BPF_ALU | BPF_ADD | BPF_X:
@@ -2007,6 +2041,7 @@ st:			if (is_imm8(insn->off))
 				emit_mov_reg(&prog, is64, real_src_reg, BPF_REG_0);
 				/* Restore R0 after clobbering RAX */
 				emit_mov_reg(&prog, true, BPF_REG_0, BPF_REG_AX);
+
 				break;
 			}
 
@@ -2031,14 +2066,20 @@ st:			if (is_imm8(insn->off))
 
 			func = (u8 *) __bpf_call_base + imm32;
 			if (tail_call_reachable) {
-				RESTORE_TAIL_CALL_CNT(bpf_prog->aux->stack_depth);
+				RESTORE_TAIL_CALL_CNT(stack_depth);
 				ip += 7;
 			}
 			if (!imm32)
 				return -EINVAL;
+			if (private_frame_ptr) {
+				EMIT2(0x41, 0x51); /* push r9 */
+				ip += 2;
+			}
 			ip += x86_call_depth_emit_accounting(&prog, func, ip);
 			if (emit_call(&prog, func, ip))
 				return -EINVAL;
+			if (private_frame_ptr)
+				EMIT2(0x41, 0x59); /* pop r9 */
 			break;
 		}
 
@@ -2048,13 +2089,13 @@ st:			if (is_imm8(insn->off))
 							  &bpf_prog->aux->poke_tab[imm32 - 1],
 							  &prog, image + addrs[i - 1],
 							  callee_regs_used,
-							  bpf_prog->aux->stack_depth,
+							  stack_depth,
 							  ctx);
 			else
 				emit_bpf_tail_call_indirect(bpf_prog,
 							    &prog,
 							    callee_regs_used,
-							    bpf_prog->aux->stack_depth,
+							    stack_depth,
 							    image + addrs[i - 1],
 							    ctx);
 			break;
@@ -3218,6 +3259,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
 {
 	struct bpf_binary_header *rw_header = NULL;
 	struct bpf_binary_header *header = NULL;
+	void __percpu *private_stack_ptr = NULL;
 	struct bpf_prog *tmp, *orig_prog = prog;
 	struct x64_jit_data *jit_data;
 	int proglen, oldproglen = 0;
@@ -3284,6 +3326,15 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
 	ctx.cleanup_addr = proglen;
 skip_init_addrs:
 
+	if (bpf_enable_private_stack(prog) && !prog->private_stack_ptr) {
+		private_stack_ptr = __alloc_percpu_gfp(prog->aux->stack_depth, 8, GFP_KERNEL);
+		if (!private_stack_ptr) {
+			prog = orig_prog;
+			goto out_addrs;
+		}
+		prog->private_stack_ptr = private_stack_ptr;
+	}
+
 	/*
 	 * JITed image shrinks with every pass and the loop iterates
 	 * until the image stops shrinking. Very large BPF programs
@@ -3309,6 +3360,10 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
 				prog->jited = 0;
 				prog->jited_len = 0;
 			}
+			if (private_stack_ptr) {
+				free_percpu(private_stack_ptr);
+				prog->private_stack_ptr = NULL;
+			}
 			goto out_addrs;
 		}
 		if (image) {
include/linux/bpf.h +2 -0
index 4f1d4a97b9d1..19a3f5355363 100644
@@ -1563,6 +1563,7 @@ struct bpf_prog {
 					    const struct bpf_insn *insn);
 	struct bpf_prog_aux	*aux;		/* Auxiliary fields */
 	struct sock_fprog_kern	*orig_prog;	/* Original BPF program */
+	void __percpu		*private_stack_ptr;
 	/* Instructions for interpreter */
 	union {
 		DECLARE_FLEX_ARRAY(struct sock_filter, insns);
@@ -1819,6 +1820,7 @@ static inline void bpf_module_put(const void *data, struct module *owner)
 		module_put(owner);
 }
 int bpf_struct_ops_link_create(union bpf_attr *attr);
+bool bpf_enable_private_stack(struct bpf_prog *prog);
 
 #ifdef CONFIG_NET
 /* Define it here to avoid the use of forward declaration */
kernel/bpf/core.c +20 -0
index 7ee62e38faf0..f69eb0c5fe03 100644
@@ -2813,6 +2813,26 @@ void bpf_prog_free(struct bpf_prog *fp)
 }
 EXPORT_SYMBOL_GPL(bpf_prog_free);
 
+bool bpf_enable_private_stack(struct bpf_prog *prog)
+{
+	if (prog->aux->stack_depth <= 64)
+		return false;
+
+	switch (prog->aux->prog->type) {
+	case BPF_PROG_TYPE_KPROBE:
+	case BPF_PROG_TYPE_TRACEPOINT:
+	case BPF_PROG_TYPE_PERF_EVENT:
+	case BPF_PROG_TYPE_RAW_TRACEPOINT:
+		return true;
+	case BPF_PROG_TYPE_TRACING:
+		if (prog->expected_attach_type != BPF_TRACE_ITER)
+			return true;
+		fallthrough;
+	default:
+		return false;
+	}
+}
+
 /* RNG for unprivileged user space with separated state from prandom_u32(). */
 static DEFINE_PER_CPU(struct rnd_state, bpf_user_rnd_state);
 
kernel/bpf/syscall.c +1 -0
index 869265852d51..89162ddb4747 100644
@@ -2244,6 +2244,7 @@ static void __bpf_prog_put_rcu(struct rcu_head *rcu)
 
 	kvfree(aux->func_info);
 	kfree(aux->func_info_aux);
+	free_percpu(aux->prog->private_stack_ptr);
 	free_uid(aux->user);
 	security_bpf_prog_free(aux->prog);
 	bpf_prog_free(aux->prog);
Signature · 1 line
2.43.0
author reviewer
#1 lore raw

RFC PATCH v2 2/2 selftests/bpf: Benchmark runtime performance with private stack

This patch intends to show some benchmark results comparing a bpf program with vs. without private stack. The patch is not intended to land since it hacks existing kernel interface in order to do proper comparison.

The jited code without private stack:

0:  f3 0f 1e fa             endbr64
4:  0f 1f 44 00 00          nop    DWORD PTR [rax+rax*1+0x0]
9:  66 90                   xchg   ax,ax
b:  55                      push   rbp
c:  48 89 e5                mov    rbp,rsp
f:  f3 0f 1e fa             endbr64
13: 48 81 ec 60 00 00 00    sub    rsp,0x60
1a: 48 bf 00 20 1a 00 00    movabs rdi,0xffffc900001a2000
21: c9 ff ff
24: 48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
28: 48 83 c6 01             add    rsi,0x1
2c: 48 89 77 00             mov    QWORD PTR [rdi+0x0],rsi
30: 31 ff                   xor    edi,edi
32: 48 89 7d f8             mov    QWORD PTR [rbp-0x8],rdi
36: be 05 00 00 00          mov    esi,0x5
3b: 89 75 f8                mov    DWORD PTR [rbp-0x8],esi
3e: 48 89 7d f0             mov    QWORD PTR [rbp-0x10],rdi
42: 48 89 7d e8             mov    QWORD PTR [rbp-0x18],rdi
46: 48 89 7d e0             mov    QWORD PTR [rbp-0x20],rdi
4a: 48 89 7d d8             mov    QWORD PTR [rbp-0x28],rdi
4e: 48 89 7d d0             mov    QWORD PTR [rbp-0x30],rdi
52: 48 89 7d c0             mov    QWORD PTR [rbp-0x40],rdi
56: 48 89 7d c8             mov    QWORD PTR [rbp-0x38],rdi
5a: 48 89 7d b8             mov    QWORD PTR [rbp-0x48],rdi
5e: 48 89 7d b0             mov    QWORD PTR [rbp-0x50],rdi
62: 48 89 7d a8             mov    QWORD PTR [rbp-0x58],rdi
66: 48 89 7d a0             mov    QWORD PTR [rbp-0x60],rdi
6a: bf 0a 00 00 00          mov    edi,0xa
6f: 89 7d c0                mov    DWORD PTR [rbp-0x40],edi
72: 48 89 ee                mov    rsi,rbp
75: 48 83 c6 d0             add    rsi,0xffffffffffffffd0
79: 48 bf 00 f8 4b 0a 81    movabs rdi,0xffff88810a4bf800
80: 88 ff ff
83: e8 e0 1d 5f e1          call   0xffffffffe15f1e68
88: 48 85 c0                test   rax,rax
8b: 74 04                   je     0x91
8d: 48 83 c0 60             add    rax,0x60
91: 48 85 c0                test   rax,rax
94: 75 1f                   jne    0xb5
96: 48 89 ee                mov    rsi,rbp
99: 48 83 c6 d0             add    rsi,0xffffffffffffffd0
9d: 48 89 ea                mov    rdx,rbp
a0: 48 83 c2 a0             add    rdx,0xffffffffffffffa0
a4: 48 bf 00 f8 4b 0a 81    movabs rdi,0xffff88810a4bf800

ab: 88 ff ff
ae: 31 c9 xor ecx,ecx
b0: e8 73 d7 5e e1 call 0xffffffffe15ed828
b5: 48 89 ee mov rsi,rbp
b8: 48 83 c6 d0 add rsi,0xffffffffffffffd0
bc: 48 bf 00 f8 4b 0a 81 movabs rdi,0xffff88810a4bf800
c3: 88 ff ff
c6: e8 0d e0 5e e1 call 0xffffffffe15ee0d8
cb: 31 c0 xor eax,eax
cd: c9 leave
ce: c3 ret

The jited code with private stack:

0:  f3 0f 1e fa             endbr64
4:  0f 1f 44 00 00          nop    DWORD PTR [rax+rax*1+0x0]
9:  66 90                   xchg   ax,ax
b:  55                      push   rbp
c:  48 89 e5                mov    rbp,rsp
f:  f3 0f 1e fa             endbr64
13: 49 b9 40 af c1 08 7e    movabs r9,0x607e08c1af40

1a: 60 00 00
1d: 65 4c 03 0c 25 00 1a add r9,QWORD PTR gs:0x21a00
24: 02 00
26: 48 bf 00 60 68 00 00 movabs rdi,0xffffc90000686000
2d: c9 ff ff

30: 48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
34: 48 83 c6 01             add    rsi,0x1
38: 48 89 77 00             mov    QWORD PTR [rdi+0x0],rsi
3c: 31 ff                   xor    edi,edi
3e: 49 89 79 f8             mov    QWORD PTR [r9-0x8],rdi
42: be 05 00 00 00          mov    esi,0x5
47: 41 89 71 f8             mov    DWORD PTR [r9-0x8],esi
4b: 49 89 79 f0             mov    QWORD PTR [r9-0x10],rdi
4f: 49 89 79 e8             mov    QWORD PTR [r9-0x18],rdi
53: 49 89 79 e0             mov    QWORD PTR [r9-0x20],rdi
57: 49 89 79 d8             mov    QWORD PTR [r9-0x28],rdi
5b: 49 89 79 d0             mov    QWORD PTR [r9-0x30],rdi
5f: 49 89 79 c0             mov    QWORD PTR [r9-0x40],rdi
63: 49 89 79 c8             mov    QWORD PTR [r9-0x38],rdi
67: 49 89 79 b8             mov    QWORD PTR [r9-0x48],rdi
6b: 49 89 79 b0             mov    QWORD PTR [r9-0x50],rdi
6f: 49 89 79 a8             mov    QWORD PTR [r9-0x58],rdi
73: 49 89 79 a0             mov    QWORD PTR [r9-0x60],rdi
77: bf 0a 00 00 00          mov    edi,0xa
7c: 41 89 79 c0             mov    DWORD PTR [r9-0x40],edi
80: 4c 89 ce                mov    rsi,r9
83: 48 83 c6 d0             add    rsi,0xffffffffffffffd0
87: 48 bf 00 e0 22 0c 81    movabs rdi,0xffff88810c22e000
8e: 88 ff ff
91: 41 51                   push   r9
93: e8 10 1d 5f e1          call   0xffffffffe15f1da8
98: 41 59                   pop    r9
9a: 48 85 c0                test   rax,rax
9d: 74 04                   je     0xa3
9f: 48 83 c0 60             add    rax,0x60
a3: 48 85 c0                test   rax,rax
a6: 75 23                   jne    0xcb
a8: 4c 89 ce                mov    rsi,r9
ab: 48 83 c6 d0             add    rsi,0xffffffffffffffd0
af: 4c 89 ca                mov    rdx,r9
b2: 48 83 c2 a0             add    rdx,0xffffffffffffffa0
b6: 48 bf 00 e0 22 0c 81    movabs rdi,0xffff88810c22e000

bd: 88 ff ff
c0: 31 c9 xor ecx,ecx
c2: 41 51 push r9
c4: e8 9f d6 5e e1 call 0xffffffffe15ed768
c9: 41 59 pop r9
cb: 4c 89 ce mov rsi,r9
ce: 48 83 c6 d0 add rsi,0xffffffffffffffd0
d2: 48 bf 00 e0 22 0c 81 movabs rdi,0xffff88810c22e000
d9: 88 ff ff
dc: 41 51 push r9
de: e8 35 df 5e e1 call 0xffffffffe15ee018
e3: 41 59 pop r9
e5: 31 c0 xor eax,eax
e7: c9 leave
e8: c3 ret

It is clear that the main overhead is the push/pop r9 for three calls.

Five runs of the benchmarks:

[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh

no-private-stack:    0.662 ± 0.019M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.684 ± 0.005M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.676 ± 0.008M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.006M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.680 ± 0.011M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.626 ± 0.050M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.686 ± 0.007M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.003M/s (drops 0.000 ± 0.000M/s)

The performance is very similar between private-stack and no-private-stack.

---
 include/linux/bpf.h                           |   3 +-
 include/uapi/linux/bpf.h                      |   3 +
 kernel/bpf/core.c                             |   2 +-
 kernel/bpf/syscall.c                          |   4 +-
 tools/include/uapi/linux/bpf.h                |   3 +
 tools/testing/selftests/bpf/Makefile          |   2 +
 tools/testing/selftests/bpf/bench.c           |   6 +
 .../bpf/benchs/bench_private_stack.c          | 142 ++++++++++++++++++
 .../bpf/benchs/run_bench_private_stack.sh     |   9 ++
 .../selftests/bpf/progs/private_stack.c       |  40 +++++
 10 files changed, 211 insertions(+), 3 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/benchs/bench_private_stack.c
 create mode 100755 tools/testing/selftests/bpf/benchs/run_bench_private_stack.sh
 create mode 100644 tools/testing/selftests/bpf/progs/private_stack.c
include/linux/bpf.h +2 -1
index 19a3f5355363..2f8708465c19 100644
@@ -1551,7 +1551,8 @@ struct bpf_prog {
 				call_get_stack:1, /* Do we call bpf_get_stack() or bpf_get_stackid() */
 				call_get_func_ip:1, /* Do we call get_func_ip() */
 				tstamp_type_access:1, /* Accessed __sk_buff->tstamp_type */
-				sleepable:1;	/* BPF program is sleepable */
+				sleepable:1,	/* BPF program is sleepable */
+				disable_private_stack:1; /* Disable private stack */
 	enum bpf_prog_type	type;		/* Type of BPF program */
 	enum bpf_attach_type	expected_attach_type; /* For some prog types */
 	u32			len;		/* Number of filter blocks */
include/uapi/linux/bpf.h +3 -0
index 35bcf52dbc65..98af8ea8a4d6 100644
@@ -1409,6 +1409,9 @@ enum {
 
 /* Do not translate kernel bpf_arena pointers to user pointers */
 	BPF_F_NO_USER_CONV	= (1U << 18),
+
+/* Disable private stack */
+	BPF_F_DISABLE_PRIVATE_STACK	= (1U << 19),
 };
 
 /* Flags for BPF_PROG_QUERY. */
kernel/bpf/core.c +1 -1
index f69eb0c5fe03..297e76a8f463 100644
@@ -2815,7 +2815,7 @@ EXPORT_SYMBOL_GPL(bpf_prog_free);
 
 bool bpf_enable_private_stack(struct bpf_prog *prog)
 {
-	if (prog->aux->stack_depth <= 64)
+	if (prog->disable_private_stack || prog->aux->stack_depth <= 64)
 		return false;
 
 	switch (prog->aux->prog->type) {
kernel/bpf/syscall.c +3 -1
index 89162ddb4747..bb2b632c9c2c 100644
@@ -2715,7 +2715,8 @@ static int bpf_prog_load(union bpf_attr *attr, bpfptr_t uattr, u32 uattr_size)
 				 BPF_F_XDP_HAS_FRAGS |
 				 BPF_F_XDP_DEV_BOUND_ONLY |
 				 BPF_F_TEST_REG_INVARIANTS |
-				 BPF_F_TOKEN_FD))
+				 BPF_F_TOKEN_FD |
+				 BPF_F_DISABLE_PRIVATE_STACK))
 		return -EINVAL;
 
 	bpf_prog_load_fixup_attach_type(attr);
@@ -2828,6 +2829,7 @@ static int bpf_prog_load(union bpf_attr *attr, bpfptr_t uattr, u32 uattr_size)
 
 	prog->expected_attach_type = attr->expected_attach_type;
 	prog->sleepable = !!(attr->prog_flags & BPF_F_SLEEPABLE);
+	prog->disable_private_stack = !!(attr->prog_flags & BPF_F_DISABLE_PRIVATE_STACK);
 	prog->aux->attach_btf = attach_btf;
 	prog->aux->attach_btf_id = attr->attach_btf_id;
 	prog->aux->dst_prog = dst_prog;
tools/include/uapi/linux/bpf.h +3 -0
index 35bcf52dbc65..98af8ea8a4d6 100644
@@ -1409,6 +1409,9 @@ enum {
 
 /* Do not translate kernel bpf_arena pointers to user pointers */
 	BPF_F_NO_USER_CONV	= (1U << 18),
+
+/* Disable private stack */
+	BPF_F_DISABLE_PRIVATE_STACK	= (1U << 19),
 };
 
 /* Flags for BPF_PROG_QUERY. */
tools/testing/selftests/bpf/Makefile +2 -0
index dd49c1d23a60..44a6a43da71c 100644
@@ -733,6 +733,7 @@ $(OUTPUT)/bench_local_storage_create.o: $(OUTPUT)/bench_local_storage_create.ske
 $(OUTPUT)/bench_bpf_hashmap_lookup.o: $(OUTPUT)/bpf_hashmap_lookup.skel.h
 $(OUTPUT)/bench_htab_mem.o: $(OUTPUT)/htab_mem_bench.skel.h
 $(OUTPUT)/bench_bpf_crypto.o: $(OUTPUT)/crypto_bench.skel.h
+$(OUTPUT)/bench_private_stack.o: $(OUTPUT)/private_stack.skel.h
 $(OUTPUT)/bench.o: bench.h testing_helpers.h $(BPFOBJ)
 $(OUTPUT)/bench: LDLIBS += -lm
 $(OUTPUT)/bench: $(OUTPUT)/bench.o \
@@ -753,6 +754,7 @@ $(OUTPUT)/bench: $(OUTPUT)/bench.o \
 		 $(OUTPUT)/bench_local_storage_create.o \
 		 $(OUTPUT)/bench_htab_mem.o \
 		 $(OUTPUT)/bench_bpf_crypto.o \
+		 $(OUTPUT)/bench_private_stack.o \
 		 #
 	$(call msg,BINARY,,$@)
 	$(Q)$(CC) $(CFLAGS) $(LDFLAGS) $(filter %.a %.o,$^) $(LDLIBS) -o $@
tools/testing/selftests/bpf/bench.c +6 -0
index 627b74ae041b..4f4867cd80f9 100644
@@ -282,6 +282,7 @@ extern struct argp bench_local_storage_create_argp;
 extern struct argp bench_htab_mem_argp;
 extern struct argp bench_trigger_batch_argp;
 extern struct argp bench_crypto_argp;
+extern struct argp bench_private_stack_argp;
 
 static const struct argp_child bench_parsers[] = {
 	{ &bench_ringbufs_argp, 0, "Ring buffers benchmark", 0 },
@@ -296,6 +297,7 @@ static const struct argp_child bench_parsers[] = {
 	{ &bench_htab_mem_argp, 0, "hash map memory benchmark", 0 },
 	{ &bench_trigger_batch_argp, 0, "BPF triggering benchmark", 0 },
 	{ &bench_crypto_argp, 0, "bpf crypto benchmark", 0 },
+	{ &bench_private_stack_argp, 0, "bpf private stack benchmark", 0 },
 	{},
 };
 
@@ -542,6 +544,8 @@ extern const struct bench bench_local_storage_create;
 extern const struct bench bench_htab_mem;
 extern const struct bench bench_crypto_encrypt;
 extern const struct bench bench_crypto_decrypt;
+extern const struct bench bench_no_private_stack;
+extern const struct bench bench_private_stack;
 
 static const struct bench *benchs[] = {
 	&bench_count_global,
@@ -596,6 +600,8 @@ static const struct bench *benchs[] = {
 	&bench_htab_mem,
 	&bench_crypto_encrypt,
 	&bench_crypto_decrypt,
+	&bench_no_private_stack,
+	&bench_private_stack,
 };
 
 static void find_benchmark(void)
tools/testing/selftests/bpf/benchs/bench_private_stack.c new file +142 -0

142 lines not on this page · raw message

tools/testing/selftests/bpf/benchs/run_bench_private_stack.sh new file +9 -0
new file mode 100755
index 000000000000..e032353f7fa6
@@ -0,0 +1,9 @@
+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+source ./benchs/run_common.sh
+
+set -eufo pipefail
+
+summarize "no-private-stack: " "$($RUN_BENCH --disable-private-stack 1 no-private-stack)"
+summarize "private-stack: " "$($RUN_BENCH private-stack)"
tools/testing/selftests/bpf/progs/private_stack.c new file +40 -0
new file mode 100644
index 000000000000..3b062e5b27e9
@@ -0,0 +1,40 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2024 Meta Platforms, Inc. and affiliates. */
+#include <linux/types.h>
+#include <linux/bpf.h>
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+
+char _license[] SEC("license") = "GPL";
+
+struct data_t {
+        unsigned int d[12];
+};
+
+struct {
+        __uint(type, BPF_MAP_TYPE_HASH);
+	__uint(max_entries, 10);
+	__type(key, struct data_t);
+	__type(value, struct data_t);
+} htab SEC(".maps");
+
+unsigned long hits = 0;
+
+SEC("tp/syscalls/sys_enter_getpgid")
+int stack0(void *ctx)
+{
+	struct data_t key = {}, value = {};
+	struct data_t *pvalue;
+
+	hits++;
+	key.d[10] = 5;
+	value.d[8] = 10;
+
+	pvalue = bpf_map_lookup_elem(&htab, &key);
+	if (!pvalue)
+		bpf_map_update_elem(&htab, &key, &value, 0);
+	bpf_map_delete_elem(&htab, &key);
+
+	return 0;
+}
+
Signature · 1 line
2.43.0
#2 lore raw

Re: [RFC PATCH bpf-next v2 2/2] [no_merge] selftests/bpf: Benchmark runtime performance with private stack

Quoted 23 lines · Yonghong Song

It is clear that the main overhead is the push/pop r9 for three calls.

Five runs of the benchmarks:

[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh

no-private-stack:    0.662 ± 0.019M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.684 ± 0.005M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.676 ± 0.008M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.006M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.680 ± 0.011M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.626 ± 0.050M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.686 ± 0.007M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.003M/s (drops 0.000 ± 0.000M/s)
Quoted 1 line

The performance is very similar between private-stack and no-private-stack.

I'm not so sure.
What is the "perf report" before/after?
Are you sure that bench spends enough time inside the program itself? By the look of it it seems that most of the time will be in hashmap and syscall overhead.

You need that batch's one that uses for loop and attached to a helper. See commit 7df4e597ea2c ("selftests/bpf: add batched, mostly in-kernel BPF triggering benchmarks")

I think the next version doesn't need RFC tag. patch 1 lgtm.

author reviewer
#3 lore raw

Re: [RFC PATCH bpf-next v2 2/2] [no_merge] selftests/bpf: Benchmark runtime performance with private stack

Quoted 30 lines · Alexei Starovoitov
Quoted 22 lines · Yonghong Song

It is clear that the main overhead is the push/pop r9 for three calls.

Five runs of the benchmarks:

[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh

no-private-stack:    0.662 ± 0.019M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.684 ± 0.005M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.676 ± 0.008M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.006M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.680 ± 0.011M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.626 ± 0.050M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.686 ± 0.007M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.003M/s (drops 0.000 ± 0.000M/s)
Quoted 1 line

The performance is very similar between private-stack and no-private-stack.

I'm not so sure.
What is the "perf report" before/after?
Are you sure that bench spends enough time inside the program itself? By the look of it it seems that most of the time will be in hashmap and syscall overhead.

Quoted 3 lines

You need that batch's one that uses for loop and attached to a helper. See commit 7df4e597ea2c ("selftests/bpf: add batched, mostly in-kernel BPF triggering benchmarks")

Okay, I see. The current approach is one trigger, one prog run where each prog run exercise 3 syscalls. I should add a loop to the bpf program to make bpf program spends majority of time. Will do this in the next revision, plus running 'perf report'.

Quoted 2 lines

I think the next version doesn't need RFC tag. patch 1 lgtm.

#4 lore raw

Re: [RFC PATCH bpf-next v2 2/2] [no_merge] selftests/bpf: Benchmark runtime performance with private stack

Quoted 33 lines · Yonghong Song
Quoted 30 lines · Alexei Starovoitov
Quoted 22 lines · Yonghong Song

It is clear that the main overhead is the push/pop r9 for three calls.

Five runs of the benchmarks:

[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh

no-private-stack:    0.662 ± 0.019M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.684 ± 0.005M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.676 ± 0.008M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.006M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.680 ± 0.011M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.626 ± 0.050M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.686 ± 0.007M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.003M/s (drops 0.000 ± 0.000M/s)
Quoted 1 line

The performance is very similar between private-stack and no-private-stack.

I'm not so sure.
What is the "perf report" before/after?
Are you sure that bench spends enough time inside the program itself? By the look of it it seems that most of the time will be in hashmap and syscall overhead.

Quoted 8 lines
Quoted 3 lines

You need that batch's one that uses for loop and attached to a helper. See commit 7df4e597ea2c ("selftests/bpf: add batched, mostly in-kernel BPF triggering benchmarks")

Okay, I see. The current approach is one trigger, one prog run where each prog run exercise 3 syscalls. I should add a loop to the bpf program to make bpf program spends majority of time. Will do this in the next revision, plus running 'perf report'.

please also benchmark on real hardware, VM will not give reliable results

Quoted 4 lines
Quoted 3 lines

I think the next version doesn't need RFC tag. patch 1 lgtm.

author reviewer
#5 lore raw

Re: [RFC PATCH bpf-next v2 2/2] [no_merge] selftests/bpf: Benchmark runtime performance with private stack

Quoted 32 lines · Andrii Nakryiko
Quoted 31 lines · Yonghong Song
Quoted 29 lines · Alexei Starovoitov
Quoted 21 lines · Yonghong Song

It is clear that the main overhead is the push/pop r9 for three calls.

Five runs of the benchmarks:

[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh

no-private-stack:    0.662 ± 0.019M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.684 ± 0.005M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.676 ± 0.008M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.673 ± 0.017M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.006M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.680 ± 0.011M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.626 ± 0.050M/s (drops 0.000 ± 0.000M/s)
[root@arch-fb-vm1 bpf]# ./benchs/run_bench_private_stack.sh
no-private-stack:    0.686 ± 0.007M/s (drops 0.000 ± 0.000M/s)
private-stack:       0.683 ± 0.003M/s (drops 0.000 ± 0.000M/s)
Quoted 1 line

The performance is very similar between private-stack and no-private-stack.

I'm not so sure.
What is the "perf report" before/after?
Are you sure that bench spends enough time inside the program itself? By the look of it it seems that most of the time will be in hashmap and syscall overhead.

Quoted 8 lines
Quoted 7 lines
Quoted 3 lines

You need that batch's one that uses for loop and attached to a helper. See commit 7df4e597ea2c ("selftests/bpf: add batched, mostly in-kernel BPF triggering benchmarks")

Okay, I see. The current approach is one trigger, one prog run where each prog run exercise 3 syscalls. I should add a loop to the bpf program to make bpf program spends majority of time. Will do this in the next revision, plus running 'perf report'.

please also benchmark on real hardware, VM will not give reliable results

Sure. Will do.

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

I think the next version doesn't need RFC tag. patch 1 lgtm.