Eduard Zingerman says:
====================
no_caller_saved_registers attribute for helper calls
This patch-set seeks to allow using no_caller_saved_registers gcc/clang
attribute with some BPF helper functions (and kfuncs in the future).
As documented in [1], this attribute means that function scratches
only some of the caller saved registers defined by ABI.
For BPF the set of such registers could be defined as follows:
- R0 is scratched only if function is non-void;
- R1-R5 are scratched only if corresponding parameter type is defined
in the function prototype.
The goal of the patch-set is to implement no_caller_saved_registers
(nocsr for short) in a backwards compatible manner:
- for kernels that support the feature, gain some performance boost
from better register allocation;
- for kernels that don't support the feature, allow programs execution
with minor performance losses.
To achieve this, use a scheme suggested by Alexei Starovoitov:
- for nocsr calls clang allocates registers as-if relevant r0-r5
registers are not scratched by the call;
- as a post-processing step, clang visits each nocsr call and adds
spill/fill for every live r0-r5;
- stack offsets used for spills/fills are allocated as lowest
stack offsets in whole function and are not used for any other
purpose;
- when kernel loads a program, it looks for such patterns
(nocsr function surrounded by spills/fills) and checks if
spill/fill stack offsets are used exclusively in nocsr patterns;
- if so, and if current JIT inlines the call to the nocsr function
(e.g. a helper call), kernel removes unnecessary spill/fill pairs;
- when old kernel loads a program, presence of spill/fill pairs
keeps BPF program valid, albeit slightly less efficient.
Corresponding clang/llvm changes are available in [2].
The patch-set uses bpf_get_smp_processor_id() function as a canary,
making it the first helper with nocsr attribute.
For example, consider the following program:
#define __no_csr __attribute__((no_caller_saved_registers))
#define SEC(name) __attribute__((section(name), used))
#define bpf_printk(fmt, ...) bpf_trace_printk((fmt), sizeof(fmt), __VA_ARGS__)
typedef unsigned int __u32;
static long (* const bpf_trace_printk)(const char *fmt, __u32 fmt_size, ...) = (void *) 6;
static __u32 (*const bpf_get_smp_processor_id)(void) __no_csr = (void *)8;
SEC("raw_tp")
int test(void *ctx)
{
__u32 task = bpf_get_smp_processor_id();
bpf_printk("ctx=%p, smp=%d", ctx, task);
return 0;
}
char _license[] SEC("license") = "GPL";
Compiled (using [2]) as follows:
$ clang --target=bpf -O2 -g -c -o nocsr.bpf.o nocsr.bpf.c
$ llvm-objdump --no-show-raw-insn -Sd nocsr.bpf.o
...
3rd parameter for printk call removable spill/fill pair
.--- 0: r3 = r1 |
; | __u32 task = bpf_get_smp_processor_id(); |
| 1: *(u64 *)(r10 - 0x8) = r3 <----------|
| 2: call 0x8 |
| 3: r3 = *(u64 *)(r10 - 0x8) <----------'
; | bpf_printk("ctx=%p, smp=%d", ctx, task);
| 4: r1 = 0x0 ll
| 6: r2 = 0xf
| 7: r4 = r0
'--> 8: call 0x6
; return 0;
9: r0 = 0x0
10: exit
Here is how the program looks after verifier processing:
# bpftool prog load ./nocsr.bpf.o /sys/fs/bpf/nocsr-test
# bpftool prog dump xlated pinned /sys/fs/bpf/nocsr-test
int test(void * ctx):
0: (bf) r3 = r1 <--- 3rd printk parameter
; __u32 task = bpf_get_smp_processor_id();
1: (b4) w0 = 197324 <--. inlined helper call,
2: (bf) r0 = &(void __percpu *)(r0) <--- spill/fill
3: (61) r0 = *(u32 *)(r0 +0) <--' pair removed
; bpf_printk("ctx=%p, smp=%d", ctx, task);
4: (18) r1 = map[id:5][0]+0
6: (b7) r2 = 15
7: (bf) r4 = r0
8: (85) call bpf_trace_printk#-125920
; return 0;
9: (b7) r0 = 0
10: (95) exit
[1] https://clang.llvm.org/docs/AttributeReference.html#no-caller-saved-registers
[2] https://github.com/eddyz87/llvm-project/tree/bpf-no-caller-saved-registers
Change list:
- v3 -> v4:
- When nocsr spills/fills are removed in the subprogram, allow these
spills/fills to reside in [-MAX_BPF_STACK-48..MAX_BPF_STACK) range
(suggested by Alexei);
- Dropped patches with special handling for bpf_probe_read_kernel()
(requested by Alexei);
- Reset aux .nocsr_pattern and .nocsr_spills_num fields in
check_nocsr_stack_contract() (requested by Andrii).
Andrii, I have not added an additional flag to
struct bpf_subprog_info, it currently does not have holes
and I really don't like adding a bool field there just as an
alternative indicator that nocsr is disabled.
Indicator at the moment:
- nocsr_stack_off >= S16_MIN means that nocsr rewrite is enabled;
- nocsr_stack_off == S16_MIN means that nocsr rewrite is disabled.
- v2 -> v3:
- As suggested by Andrii, 'nocsr_stack_off' is no longer checked at
rewrite time, instead mark_nocsr_patterns() now does two passes
over BPF program:
- on a first pass it computes the lowest stack spill offset for
the subprogram;
- on a second pass this offset is used to recognize nocsr pattern.
- As suggested by Alexei, a new mechanic is added to work around a
situation mentioned by Andrii, when more helper functions are
marked as nocsr at compile time than current kernel supports:
- all {spill*,helper call,fill*} patterns are now marked as
insn_aux_data[*].nocsr_pattern, thus relaxing failure condition
for check_nocsr_stack_contract();
- spill/fill pairs are not removed for patterns where helper can't
be inlined;
- see mark_nocsr_pattern_for_call() for details an example.
- As suggested by Alexei, subprogram stack depth is now adjusted
if all spill/fill pairs could be removed. This adjustment has
to take place before optimize_bpf_loop(), hence the rewrite
is moved from do_misc_fixups() to remove_nocsr_spills_fills()
(again).
- As suggested by Andrii, special measures are taken to work around
bpf_probe_read_kernel() access to BPF stack, see patches 11, 12.
Patch #11 is very simplistic, a more comprehensive solution would
be to change the type of the third parameter of the
bpf_probe_read_kernel() from ARG_ANYTHING to something else and
not only check nocsr contract, but also propagate stack slot
liveness information. However, such change would require update in
struct bpf_call_arg_meta processing, which currently implies that
every memory parameter is followed by a size parameter.
I can work on these changes, please comment.
- Stylistic changes suggested by Andrii.
- Added acks from Andrii.
- Dropped RFC tag.
- v1 -> v2:
- assume that functions inlined by either jit or verifier
conform to no_caller_saved_registers contract (Andrii, Puranjay);
- allow nocsr rewrite for bpf_get_smp_processor_id()
on arm64 and riscv64 architectures (Puranjay);
- __arch_{x86_64,arm64,riscv64} macro for test_loader;
- moved remove_nocsr_spills_fills() inside do_misc_fixups() (Andrii);
- moved nocsr pattern detection from check_cfg() to a separate pass
(Andrii);
- various stylistic/correctness changes according to Andrii's
comments.
Revisions:
- v1 https://lore.kernel.org/bpf/20240629094733.3863850-1-eddyz87@gmail.com/
- v2 https://lore.kernel.org/bpf/20240704102402.1644916-1-eddyz87@gmail.com/
- v3 https://lore.kernel.org/bpf/20240715230201.3901423-1-eddyz87@gmail.com/
====================
Link: https://lore.kernel.org/r/20240722233844.1406874-1-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Tests for no_caller_saved_registers processing logic
(see verifier.c:match_and_mark_nocsr_pattern()):
- a canary positive test case;
- a canary test case for arm64 and riscv64;
- various tests with broken patterns;
- tests with read/write fixed/varying stack access that violate nocsr
stack access contract;
- tests with multiple subprograms;
- tests using nocsr in combination with may_goto/bpf_loop,
as all of these features affect stack depth;
- tests for nocsr stack spills below max stack depth.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-11-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Add annotations __arch_x86_64, __arch_arm64, __arch_riscv64
to specify on which architecture the test case should be tested.
Several __arch_* annotations could be specified at once.
When test case is not run on current arch it is marked as skipped.
For example, the following would be tested only on arm64 and riscv64:
SEC("raw_tp")
__arch_arm64
__arch_riscv64
__xlated("1: *(u64 *)(r10 - 16) = r1")
__xlated("2: call")
__xlated("3: r1 = *(u64 *)(r10 - 16);")
__success
__naked void canary_arm64_riscv64(void)
{
asm volatile (
"r1 = 1;"
"*(u64 *)(r10 - 16) = r1;"
"call %[bpf_get_smp_processor_id];"
"r1 = *(u64 *)(r10 - 16);"
"exit;"
:
: __imm(bpf_get_smp_processor_id)
: __clobber_all);
}
On x86 it would be skipped:
#467/2 verifier_nocsr/canary_arm64_riscv64:SKIP
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-10-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Add a macro __xlated("...") for use with test_loader tests.
When such annotations are present for the test case:
- bpf_prog_get_info_by_fd() is used to get BPF program after all
rewrites are applied by verifier.
- the program is disassembled and patterns specified in __xlated are
searched for in the disassembly text.
__xlated matching follows the same mechanics as __msg:
each subsequent pattern is matched from the point where
previous pattern ended.
This allows to write tests like below, where the goal is to verify the
behavior of one of the of the transformations applied by verifier:
SEC("raw_tp")
__xlated("1: w0 = ")
__xlated("2: r0 = &(void __percpu *)(r0)")
__xlated("3: r0 = *(u32 *)(r0 +0)")
__xlated("4: exit")
__success __naked void simple(void)
{
asm volatile (
"call %[bpf_get_smp_processor_id];"
"exit;"
:
: __imm(bpf_get_smp_processor_id)
: __clobber_all);
}
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-9-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
The call stack for validate_case() function looks as follows:
- test_loader__run_subtests()
- process_subtest()
- run_subtest()
- prepare_case(), which does 'tester->next_match_pos = 0';
- validate_case(), which increments tester->next_match_pos.
Hence, each subtest is run with next_match_pos freshly set to zero.
Meaning that there is no need to persist this variable in the
struct test_loader, use local variable instead.
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-7-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Adjust disasm_helpers.c:disasm_insn() to account for the following
part of the verifier.c:jit_subprogs:
for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) {
/* ... */
if (!bpf_pseudo_call(insn))
continue;
insn->off = env->insn_aux_data[i].call_imm;
subprog = find_subprog(env, i + insn->off + 1);
insn->imm = subprog;
}
Where verifier moves offset of the subprogram to the insn->off field.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-6-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
struct bpf_insn *disasm_insn(struct bpf_insn *insn, char *buf, size_t buf_sz);
Disassembles instruction 'insn' to a text buffer 'buf'.
Removes insn->code hex prefix added by kernel disassembly routine.
Returns a pointer to the next instruction
(increments insn by either 1 or 2).
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-5-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
With latest llvm19, the selftest iters/iter_arr_with_actual_elem_count
failed with -mcpu=v4.
The following are the details:
0: R1=ctx() R10=fp0
; int iter_arr_with_actual_elem_count(const void *ctx) @ iters.c:1420
0: (b4) w7 = 0 ; R7_w=0
; int i, n = loop_data.n, sum = 0; @ iters.c:1422
1: (18) r1 = 0xffffc90000191478 ; R1_w=map_value(map=iters.bss,ks=4,vs=1280,off=1144)
3: (61) r6 = *(u32 *)(r1 +128) ; R1_w=map_value(map=iters.bss,ks=4,vs=1280,off=1144) R6_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff))
; if (n > ARRAY_SIZE(loop_data.data)) @ iters.c:1424
4: (26) if w6 > 0x20 goto pc+27 ; R6_w=scalar(smin=smin32=0,smax=umax=smax32=umax32=32,var_off=(0x0; 0x3f))
5: (bf) r8 = r10 ; R8_w=fp0 R10=fp0
6: (07) r8 += -8 ; R8_w=fp-8
; bpf_for(i, 0, n) { @ iters.c:1427
7: (bf) r1 = r8 ; R1_w=fp-8 R8_w=fp-8
8: (b4) w2 = 0 ; R2_w=0
9: (bc) w3 = w6 ; R3_w=scalar(id=1,smin=smin32=0,smax=umax=smax32=umax32=32,var_off=(0x0; 0x3f)) R6_w=scalar(id=1,smin=smin32=0,smax=umax=smax32=umax32=32,var_off=(0x0; 0x3f))
10: (85) call bpf_iter_num_new#45179 ; R0=scalar() fp-8=iter_num(ref_id=2,state=active,depth=0) refs=2
11: (bf) r1 = r8 ; R1=fp-8 R8=fp-8 refs=2
12: (85) call bpf_iter_num_next#45181 13: R0=rdonly_mem(id=3,ref_obj_id=2,sz=4) R6=scalar(id=1,smin=smin32=0,smax=umax=smax32=umax32=32,var_off=(0x0; 0x3f)) R7=0 R8=fp-8 R10=fp0 fp-8=iter_num(ref_id=2,state=active,depth=1) refs=2
; bpf_for(i, 0, n) { @ iters.c:1427
13: (15) if r0 == 0x0 goto pc+2 ; R0=rdonly_mem(id=3,ref_obj_id=2,sz=4) refs=2
14: (81) r1 = *(s32 *)(r0 +0) ; R0=rdonly_mem(id=3,ref_obj_id=2,sz=4) R1_w=scalar(smin=0xffffffff80000000,smax=0x7fffffff) refs=2
15: (ae) if w1 < w6 goto pc+4 20: R0=rdonly_mem(id=3,ref_obj_id=2,sz=4) R1=scalar(smin=0xffffffff80000000,smax=smax32=umax32=31,umax=0xffffffff0000001f,smin32=0,var_off=(0x0; 0xffffffff0000001f)) R6=scalar(id=1,smin=umin=smin32=umin32=1,smax=umax=smax32=umax32=32,var_off=(0x0; 0x3f)) R7=0 R8=fp-8 R10=fp0 fp-8=iter_num(ref_id=2,state=active,depth=1) refs=2
; sum += loop_data.data[i]; @ iters.c:1429
20: (67) r1 <<= 2 ; R1_w=scalar(smax=0x7ffffffc0000007c,umax=0xfffffffc0000007c,smin32=0,smax32=umax32=124,var_off=(0x0; 0xfffffffc0000007c)) refs=2
21: (18) r2 = 0xffffc90000191478 ; R2_w=map_value(map=iters.bss,ks=4,vs=1280,off=1144) refs=2
23: (0f) r2 += r1
math between map_value pointer and register with unbounded min value is not allowed
The source code:
int iter_arr_with_actual_elem_count(const void *ctx)
{
int i, n = loop_data.n, sum = 0;
if (n > ARRAY_SIZE(loop_data.data))
return 0;
bpf_for(i, 0, n) {
/* no rechecking of i against ARRAY_SIZE(loop_data.n) */
sum += loop_data.data[i];
}
return sum;
}
The insn #14 is a sign-extenstion load which is related to 'int i'.
The insn #15 did a subreg comparision. Note that smin=0xffffffff80000000 and this caused later
insn #23 failed verification due to unbounded min value.
Actually insn #15 R1 smin range can be better. Before insn #15, we have
R1_w=scalar(smin=0xffffffff80000000,smax=0x7fffffff)
With the above range, we know for R1, upper 32bit can only be 0xffffffff or 0.
Otherwise, the value range for R1 could be beyond [smin=0xffffffff80000000,smax=0x7fffffff].
After insn #15, for the true patch, we know smin32=0 and smax32=32. With the upper 32bit 0xffffffff,
then the corresponding value is [0xffffffff00000000, 0xffffffff00000020]. The range is
obviously beyond the original range [smin=0xffffffff80000000,smax=0x7fffffff] and the
range is not possible. So the upper 32bit must be 0, which implies smin = smin32 and
smax = smax32.
This patch fixed the issue by adding additional register deduction after 32-bit compare
insn. If the signed 32-bit register range is non-negative then 64-bit smin is
in range of [S32_MIN, S32_MAX], then the actual 64-bit smin/smax should be the same
as 32-bit smin32/smax32.
With this patch, iters/iter_arr_with_actual_elem_count succeeded with better register range:
from 15 to 20: R0=rdonly_mem(id=7,ref_obj_id=2,sz=4) R1_w=scalar(smin=smin32=0,smax=umax=smax32=umax32=31,var_off=(0x0; 0x1f)) R6=scalar(id=1,smin=umin=smin32=umin32=1,smax=umax=smax32=umax32=32,var_off=(0x0; 0x3f)) R7=scalar(id=9,smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) R8=scalar(id=9,smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) R10=fp0 fp-8=iter_num(ref_id=2,state=active,depth=3) refs=2
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20240723162933.2731620-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
The function bpf_get_smp_processor_id() is processed in a different
way, depending on the arch:
- on x86 verifier replaces call to bpf_get_smp_processor_id() with a
sequence of instructions that modify only r0;
- on riscv64 jit replaces call to bpf_get_smp_processor_id() with a
sequence of instructions that modify only r0;
- on arm64 jit replaces call to bpf_get_smp_processor_id() with a
sequence of instructions that modify only r0 and tmp registers.
These rewrites satisfy attribute no_caller_saved_registers contract.
Allow rewrite of no_caller_saved_registers patterns for
bpf_get_smp_processor_id() in order to use this function as a canary
for no_caller_saved_registers tests.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-4-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
The following tests are added to verifier_ldsx.c:
- sign extension of data/data_end/data_meta for tcx programs.
The actual checking is in bpf_skb_is_valid_access() which
is called by sk_filter, cg_skb, lwt, tc(tcx) and sk_skb.
- sign extension of data/data_end/data_meta for xdp programs.
- sign extension of data/data_end for flow_dissector programs.
All newly-added tests have verification failure with message
"invalid bpf_context access". Without previous patch, all these
tests succeeded verification.
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20240723153444.2430365-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
GCC and LLVM define a no_caller_saved_registers function attribute.
This attribute means that function scratches only some of
the caller saved registers defined by ABI.
For BPF the set of such registers could be defined as follows:
- R0 is scratched only if function is non-void;
- R1-R5 are scratched only if corresponding parameter type is defined
in the function prototype.
This commit introduces flag bpf_func_prot->allow_nocsr.
If this flag is set for some helper function, verifier assumes that
it follows no_caller_saved_registers calling convention.
The contract between kernel and clang allows to simultaneously use
such functions and maintain backwards compatibility with old
kernels that don't understand no_caller_saved_registers calls
(nocsr for short):
- clang generates a simple pattern for nocsr calls, e.g.:
r1 = 1;
r2 = 2;
*(u64 *)(r10 - 8) = r1;
*(u64 *)(r10 - 16) = r2;
call %[to_be_inlined]
r2 = *(u64 *)(r10 - 16);
r1 = *(u64 *)(r10 - 8);
r0 = r1;
r0 += r2;
exit;
- kernel removes unnecessary spills and fills, if called function is
inlined by verifier or current JIT (with assumption that patch
inserted by verifier or JIT honors nocsr contract, e.g. does not
scratch r3-r5 for the example above), e.g. the code above would be
transformed to:
r1 = 1;
r2 = 2;
call %[to_be_inlined]
r0 = r1;
r0 += r2;
exit;
Technically, the transformation is split into the following phases:
- function mark_nocsr_patterns(), called from bpf_check()
searches and marks potential patterns in instruction auxiliary data;
- upon stack read or write access,
function check_nocsr_stack_contract() is used to verify if
stack offsets, presumably reserved for nocsr patterns, are used
only from those patterns;
- function remove_nocsr_spills_fills(), called from bpf_check(),
applies the rewrite for valid patterns.
See comment in mark_nocsr_pattern_for_call() for more details.
Suggested-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-3-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
syzbot reported a kernel crash due to
commit 1f1e864b65 ("bpf: Handle sign-extenstin ctx member accesses").
The reason is due to sign-extension of 32-bit load for
packet data/data_end/data_meta uapi field.
The original code looks like:
r2 = *(s32 *)(r1 + 76) /* load __sk_buff->data */
r3 = *(u32 *)(r1 + 80) /* load __sk_buff->data_end */
r0 = r2
r0 += 8
if r3 > r0 goto +1
...
Note that __sk_buff->data load has 32-bit sign extension.
After verification and convert_ctx_accesses(), the final asm code looks like:
r2 = *(u64 *)(r1 +208)
r2 = (s32)r2
r3 = *(u64 *)(r1 +80)
r0 = r2
r0 += 8
if r3 > r0 goto pc+1
...
Note that 'r2 = (s32)r2' may make the kernel __sk_buff->data address invalid
which may cause runtime failure.
Currently, in C code, typically we have
void *data = (void *)(long)skb->data;
void *data_end = (void *)(long)skb->data_end;
...
and it will generate
r2 = *(u64 *)(r1 +208)
r3 = *(u64 *)(r1 +80)
r0 = r2
r0 += 8
if r3 > r0 goto pc+1
If we allow sign-extension,
void *data = (void *)(long)(int)skb->data;
void *data_end = (void *)(long)skb->data_end;
...
the generated code looks like
r2 = *(u64 *)(r1 +208)
r2 <<= 32
r2 s>>= 32
r3 = *(u64 *)(r1 +80)
r0 = r2
r0 += 8
if r3 > r0 goto pc+1
and this will cause verification failure since "r2 <<= 32" is not allowed
as "r2" is a packet pointer.
To fix this issue for case
r2 = *(s32 *)(r1 + 76) /* load __sk_buff->data */
this patch added additional checking in is_valid_access() callback
function for packet data/data_end/data_meta access. If those accesses
are with sign-extenstion, the verification will fail.
[1] https://lore.kernel.org/bpf/000000000000c90eee061d236d37@google.com/
Reported-by: syzbot+ad9ec60c8eaf69e6f99c@syzkaller.appspotmail.com
Fixes: 1f1e864b65 ("bpf: Handle sign-extenstin ctx member accesses")
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20240723153439.2429035-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Make log output incorrectly shows 'test_maps' as the binary name for every
'CLNG-BPF' build step, apparently picking up the last value defined for the
$(TRUNNER_BINARY) variable. Update the 'CLANG_BPF_BUILD_RULE' variants to
fix this confusing output.
Current output:
CLNG-BPF [test_maps] access_map_in_map.bpf.o
GEN-SKEL [test_progs] access_map_in_map.skel.h
...
CLNG-BPF [test_maps] access_map_in_map.bpf.o
GEN-SKEL [test_progs-no_alu32] access_map_in_map.skel.h
...
CLNG-BPF [test_maps] access_map_in_map.bpf.o
GEN-SKEL [test_progs-cpuv4] access_map_in_map.skel.h
After fix:
CLNG-BPF [test_progs] access_map_in_map.bpf.o
GEN-SKEL [test_progs] access_map_in_map.skel.h
...
CLNG-BPF [test_progs-no_alu32] access_map_in_map.bpf.o
GEN-SKEL [test_progs-no_alu32] access_map_in_map.skel.h
...
CLNG-BPF [test_progs-cpuv4] access_map_in_map.bpf.o
GEN-SKEL [test_progs-cpuv4] access_map_in_map.skel.h
Fixes: a5d0c26a27 ("selftests/bpf: Add a cpuv4 test runner for cpu=v4 testing")
Fixes: 89ad7420b2 ("selftests/bpf: Drop the need for LLVM's llc")
Signed-off-by: Tony Ambardar <tony.ambardar@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20240720052535.2185967-1-tony.ambardar@gmail.com
Linking uprobe_multi.c on mips64el fails due to relocation overflows, when
the GOT entries required exceeds the default maximum. Add a specific CFLAGS
(-mxgot) for uprobe_multi.c on MIPS that allows using a larger GOT and
avoids errors such as:
/tmp/ccBTNQzv.o: in function `bench':
uprobe_multi.c:49:(.text+0x1d7720): relocation truncated to fit: R_MIPS_GOT_DISP against `uprobe_multi_func_08188'
uprobe_multi.c:49:(.text+0x1d7730): relocation truncated to fit: R_MIPS_GOT_DISP against `uprobe_multi_func_08189'
...
collect2: error: ld returned 1 exit status
Fixes: 519dfeaf51 ("selftests/bpf: Add uprobe_multi test program")
Signed-off-by: Tony Ambardar <tony.ambardar@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/14eb7b70f8ccef9834874d75eb373cb9292129da.1721692479.git.tony.ambardar@gmail.com
Regular BPF hash map is not mmap-able from user space. However, map-in-map
with outer map of type BPF_MAP_TYPE_HASH_OF_MAPS and mmap-able array as
inner map can perform similar operations as a mmap-able hash map. This
can be used by applications that benefit from fast accesses to some local
data.
Add a selftest to show this use case.
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20240723051455.1589192-1-song@kernel.org
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Geliang Tang says:
====================
This set is part 10 of series "use network helpers" all BPF selftests
wide.
Patches 1-3 drop local functions make_client(), make_socket() and
inetaddr_len() in sk_lookup.c. Patch 4 drops a useless function
__start_server() in network_helpers.c.
====================
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
test_progs-no_alu32 -t libbpf_get_fd_by_id_opts
is being rejected by the verifier with the following error
due to compiler optimization:
6: (67) r0 <<= 62 ; R0_w=scalar(smax=0x4000000000000000,umax=0xc000000000000000,smin32=0,smax32=umax32=0,var_off=(0x0; 0xc000000000000000))
7: (c7) r0 s>>= 63 ; R0_w=scalar(smin=smin32=-1,smax=smax32=0)
; @ test_libbpf_get_fd_by_id_opts.c:0
8: (57) r0 &= -13 ; R0_w=scalar(smax=0x7ffffffffffffff3,umax=0xfffffffffffffff3,smax32=0x7ffffff3,umax32=0xfffffff3,var_off=(0x0; 0xfffffffffffffff3))
; int BPF_PROG(check_access, struct bpf_map *map, fmode_t fmode) @ test_libbpf_get_fd_by_id_opts.c:27
9: (95) exit
At program exit the register R0 has smax=9223372036854775795 should have been in [-4095, 0]
Workaround by adding barrier().
Eventually the verifier will be able to recognize it.
Fixes: 5d99e198be ("bpf, lsm: Add check for BPF LSM return value")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
After checking lsm hook return range in verifier, the test case
"test_progs -t test_lsm" failed, and the failure log says:
libbpf: prog 'test_int_hook': BPF program load failed: Invalid argument
libbpf: prog 'test_int_hook': -- BEGIN PROG LOAD LOG --
0: R1=ctx() R10=fp0
; int BPF_PROG(test_int_hook, struct vm_area_struct *vma, @ lsm.c:89
0: (79) r0 = *(u64 *)(r1 +24) ; R0_w=scalar(smin=smin32=-4095,smax=smax32=0) R1=ctx()
[...]
24: (b4) w0 = -1 ; R0_w=0xffffffff
; int BPF_PROG(test_int_hook, struct vm_area_struct *vma, @ lsm.c:89
25: (95) exit
At program exit the register R0 has smin=4294967295 smax=4294967295 should have been in [-4095, 0]
It can be seen that instruction "w0 = -1" zero extended -1 to 64-bit
register r0, setting both smin and smax values of r0 to 4294967295.
This resulted in a false reject when r0 was checked with range [-4095, 0].
Given bpf lsm does not return 64-bit values, this patch fixes it by changing
the compare between r0 and return range from 64-bit operation to 32-bit
operation for bpf lsm.
Fixes: 8fa4ecd49b ("bpf: enforce exact retval range on subprog/callback exit")
Signed-off-by: Xu Kuohai <xukuohai@huawei.com>
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Link: https://lore.kernel.org/r/20240719110059.797546-5-xukuohai@huaweicloud.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
bpf progs can be attached to kernel functions, and the attached functions
can take different parameters or return different return values. If
prog attached to one kernel function tail calls prog attached to another
kernel function, the ctx access or return value verification could be
bypassed.
For example, if prog1 is attached to func1 which takes only 1 parameter
and prog2 is attached to func2 which takes two parameters. Since verifier
assumes the bpf ctx passed to prog2 is constructed based on func2's
prototype, verifier allows prog2 to access the second parameter from
the bpf ctx passed to it. The problem is that verifier does not prevent
prog1 from passing its bpf ctx to prog2 via tail call. In this case,
the bpf ctx passed to prog2 is constructed from func1 instead of func2,
that is, the assumption for ctx access verification is bypassed.
Another example, if BPF LSM prog1 is attached to hook file_alloc_security,
and BPF LSM prog2 is attached to hook bpf_lsm_audit_rule_known. Verifier
knows the return value rules for these two hooks, e.g. it is legal for
bpf_lsm_audit_rule_known to return positive number 1, and it is illegal
for file_alloc_security to return positive number. So verifier allows
prog2 to return positive number 1, but does not allow prog1 to return
positive number. The problem is that verifier does not prevent prog1
from calling prog2 via tail call. In this case, prog2's return value 1
will be used as the return value for prog1's hook file_alloc_security.
That is, the return value rule is bypassed.
This patch adds restriction for tail call to prevent such bypasses.
Signed-off-by: Xu Kuohai <xukuohai@huawei.com>
Link: https://lore.kernel.org/r/20240719110059.797546-4-xukuohai@huaweicloud.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
A bpf prog returning a positive number attached to file_alloc_security
hook makes kernel panic.
This happens because file system can not filter out the positive number
returned by the LSM prog using IS_ERR, and misinterprets this positive
number as a file pointer.
Given that hook file_alloc_security never returned positive number
before the introduction of BPF LSM, and other BPF LSM hooks may
encounter similar issues, this patch adds LSM return value check
in verifier, to ensure no unexpected value is returned.
Fixes: 520b7aa00d ("bpf: lsm: Initialize the BPF LSM hooks")
Reported-by: Xin Liu <liuxin350@huawei.com>
Signed-off-by: Xu Kuohai <xukuohai@huawei.com>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240719110059.797546-3-xukuohai@huaweicloud.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Martin KaFai Lau says:
====================
bpf: Retire the unsupported_ops usage in struct_ops
From: Martin KaFai Lau <martin.lau@kernel.org>
This series retires the unsupported_ops usage and depends on the
null-ness check on the cfi_stubs instead.
Please see individual patches for details.
v2:
- Fixed a gcc compiler warning on Patch 1.
====================
Link: https://lore.kernel.org/r/20240722183049.2254692-1-martin.lau@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
There is an existing "bpf_tcp_ca/unsupp_cong_op" test to ensure
the unsupported tcp-cc "get_info" struct_ops prog cannot be loaded.
This patch adds a new test in the bpf_testmod such that the
unsupported ops test does not depend on other kernel subsystem
where its supporting ops may be changed in the future.
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://lore.kernel.org/r/20240722183049.2254692-4-martin.lau@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
The bpf_tcp_ca struct_ops currently uses a "u32 unsupported_ops[]"
array to track which ops is not supported.
After cfi_stubs had been added, the function pointer in cfi_stubs is
also NULL for the unsupported ops. Thus, the "u32 unsupported_ops[]"
becomes redundant. This observation was originally brought up in the
bpf/cfi discussion:
https://lore.kernel.org/bpf/CAADnVQJoEkdjyCEJRPASjBw1QGsKYrF33QdMGc1RZa9b88bAEA@mail.gmail.com/
The recent bpf qdisc patch (https://lore.kernel.org/bpf/20240714175130.4051012-6-amery.hung@bytedance.com/)
also needs to specify quite many unsupported ops. It is a good time
to clean it up.
This patch removes the need of "u32 unsupported_ops[]" and tests for null-ness
in the cfi_stubs instead.
Testing the cfi_stubs is done in a new function bpf_struct_ops_supported().
The verifier will call bpf_struct_ops_supported() when loading the
struct_ops program. The ".check_member" is removed from the bpf_tcp_ca
in this patch. ".check_member" could still be useful for other subsytems
to enforce other restrictions (e.g. sched_ext checks for prog->sleepable).
To keep the same error return, ENOTSUPP is used.
Cc: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://lore.kernel.org/r/20240722183049.2254692-2-martin.lau@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Now, attach/detach tcx prog supported in libbpf, so we can add new
command 'bpftool attach/detach tcx' to attach tcx prog with bpftool
for user.
# bpftool prog load tc_prog.bpf.o /sys/fs/bpf/tc_prog
# bpftool prog show
...
192: sched_cls name tc_prog tag 187aeb611ad00cfc gpl
loaded_at 2024-07-11T15:58:16+0800 uid 0
xlated 152B jited 97B memlock 4096B map_ids 100,99,97
btf_id 260
# bpftool net attach tcx_ingress name tc_prog dev lo
# bpftool net
...
tc:
lo(1) tcx/ingress tc_prog prog_id 29
# bpftool net detach tcx_ingress dev lo
# bpftool net
...
tc:
# bpftool net attach tcx_ingress name tc_prog dev lo
# bpftool net
tc:
lo(1) tcx/ingress tc_prog prog_id 29
Test environment: ubuntu_22_04, 6.7.0-060700-generic
Signed-off-by: Tao Chen <chen.dylane@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Quentin Monnet <qmo@kernel.org>
Link: https://lore.kernel.org/bpf/20240721144221.96228-1-chen.dylane@gmail.com
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Leon Hwang says:
====================
bpf: Fix tailcall hierarchy
This patchset fixes a tailcall hierarchy issue.
The issue is confirmed in the discussions of
"bpf, x64: Fix tailcall infinite loop" [0].
The issue has been resolved on both x86_64 and arm64 [1].
I provide a long commit message in the "bpf, x64: Fix tailcall hierarchy"
patch to describe how the issue happens and how this patchset resolves the
issue in details.
How does this patchset resolve the issue?
In short, it stores tail_call_cnt on the stack of main prog, and propagates
tail_call_cnt_ptr to its subprogs.
First, at the prologue of main prog, it initializes tail_call_cnt and
prepares tail_call_cnt_ptr. And at the prologue of subprog, it reuses
the tail_call_cnt_ptr from caller.
Then, when a tailcall happens, it increments tail_call_cnt by its pointer.
v5 -> v6:
* Address comments from Eduard:
* Add JITed dumping along annotating comments
* Rewrite two selftests with RUN_TESTS macro.
v4 -> v5:
* Solution changes from tailcall run ctx to tail_call_cnt and its pointer.
It's because v4 solution is unable to handle the case that there is no
tailcall in subprog but there is tailcall in EXT prog which attaches to
the subprog.
v3 -> v4:
* Solution changes from per-task tail_call_cnt to tailcall run ctx.
As for per-cpu/per-task solution, there is a case it is unable to handle [2].
v2 -> v3:
* Solution changes from percpu tail_call_cnt to tail_call_cnt at task_struct.
v1 -> v2:
* Solution changes from extra run-time call insn to percpu tail_call_cnt.
* Address comments from Alexei:
* Use percpu tail_call_cnt.
* Use asm to make sure no callee saved registers are touched.
RFC v2 -> v1:
* Solution changes from propagating tail_call_cnt with its pointer to extra
run-time call insn.
* Address comments from Maciej:
* Replace all memcpy(prog, x86_nops[5], X86_PATCH_SIZE) with
emit_nops(&prog, X86_PATCH_SIZE)
RFC v1 -> RFC v2:
* Address comments from Stanislav:
* Separate moving emit_nops() as first patch.
Links:
[0] https://lore.kernel.org/bpf/6203dd01-789d-f02c-5293-def4c1b18aef@gmail.com/
[1] https://github.com/kernel-patches/bpf/pull/7350/checks
[2] https://lore.kernel.org/bpf/CAADnVQK1qF+uBjwom2s2W-yEmgd_3rGi5Nr+KiV3cW0T+UPPfA@mail.gmail.com/
====================
Link: https://lore.kernel.org/r/20240714123902.32305-1-hffilwlqm@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
