len = bpf_probe_read_str(); bpf_perf_event_output(... len) == FAIL
Subsystems (6)
Hi,
In a5e8c07059d0 ("bpf: add bpf_probe_read_str helper") youstate:
"This is suboptimal because the size of the string needs to be estimated
at compile time, causing more memory to be copied than often necessary, and can become more problematic if further processing on buf is done, for example by pushing it to userspace via bpf_perf_event_output(), since the real length of the string is unknown and the entire buffer must be copied (and defining an unrolled strnlen() inside the bpf program is a very inefficient and unfeasible approach)."
So I went on to try this with 'perf trace' but it isn't working if I use the return from bpf_probe_read_str(), I must be missing something here...
I.e. this works:
[root@jouet bpf]# cat open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
int prog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, 32);
return 1;
}
[root@jouet bpf]# perf trace -e open,open.c touch /etc/passwd
bpf: builtin compilation failed: -95, try external compiler
0.000 ( 0.013 ms): touch/14403 open(filename: 0x2ff7ce37, flags: CLOEXEC ) ...
0.013 ( ): __bpf_stdout__:/etc/ld.so.cache..B.................)
0.015 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7f7a2ff7ce37)
0.000 ( 0.021 ms): touch/14403 ... [continued]: open()) = 3
0.042 ( 0.002 ms): touch/14403 open(filename: 0x30180640, flags: CLOEXEC ) ...
0.044 ( ): __bpf_stdout__:/lib64/libc.so.6.. .......G.........)
0.045 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7f7a30180640)
0.042 ( 0.010 ms): touch/14403 ... [continued]: open()) = 3
0.301 ( 0.003 ms): touch/14403 open(filename: 0x2fd26c70, flags: CLOEXEC ) ...
0.305 ( ): __bpf_stdout__:/usr/lib/locale/locale-archive......)
0.306 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7f7a2fd26c70)
0.301 ( 0.011 ms): touch/14403 ... [continued]: open()) = 3
0.360 ( 0.002 ms): touch/14403 open(filename: 0x681f20f3, flags: CREAT|NOCTTY|NONBLOCK|WRONLY, mode: IRUGO|IWUGO) ...
0.362 ( ): __bpf_stdout__:/etc/passwd....... .......D.........)
0.363 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7ffe681f20f3)
0.360 ( 0.010 ms): touch/14403 ... [continued]: open()) = 3[root@jouet bpf]#
That bpf.h will set up the maps, etc, its attached if that may be needed to help figure this out.
But then if I use the return value to push just the string lenght, it doesn't work:
[root@jouet bpf]# cat open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
int prog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);
return 1;
}[root@jouet bpf]# perf trace -e open,open.c touch /etc/passwd
bpf: builtin compilation failed: -95, try external compiler event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
[root@jouet bpf]#
When running this with -v we get the tools/lib/libbpf.c debug that may help here:
Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: p:perf_bpf_probe/prog _text+2493152 filename=%si:x64 In map_prologue, ntevs=1
mapping[0]=0
libbpf: create map __bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r7 = r10
4: (07) r7 += -128
5: (bf) r1 = r7
6: (b7) r2 = 128
7: (85) call bpf_probe_read_str#45
8: (bf) r8 = r0
9: (67) r8 <<= 32
10: (77) r8 >>= 32
11: (85) call bpf_get_smp_processor_id#8
12: (bf) r1 = r6
13: (18) r2 = 0xffffa0b5958e16c0
15: (bf) r3 = r0
16: (bf) r4 = r7
17: (bf) r5 = r8
18: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0
libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
I tried adding checks for len to try to somehow make sure its all bounds checked, but couldn't get past that "invalid stack type R4", the problem seems to be that access_size=0...
Ideas?
- Arnaldo
Daniel Borkmann
maintainerQuoted 22 lines · Arnaldo Carvalho de Melo
Hi,
In a5e8c07059d0 ("bpf: add bpf_probe_read_str helper") youstate:
"This is suboptimal because the size of the string needs to be estimated
at compile time, causing more memory to be copied than often necessary, and can become more problematic if further processing on buf is done, for example by pushing it to userspace viabpf_perf_event_output(), since the real length of the string is unknown and the entire buffer must be copied (and defining an unrolledstrnlen()inside the bpf program is a very inefficient and unfeasible approach)."So I went on to try this with 'perf trace' but it isn't working if I use the return from
bpf_probe_read_str(), I must be missing something here...I.e. this works:
[root@jouet bpf]# cat
open.c
#include "bpf.h"
Quoted 6 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, 32);
By the way, you can just use BPF_F_CURRENT_CPU flag instead of the helper call get_smp_processor_id() to get current CPU.
Quoted 31 lines
return 1; } [root@jouet bpf]# perf trace -e open,open.c touch /etc/passwd bpf: builtin compilation failed: -95, try external compiler 0.000 ( 0.013 ms): touch/14403 open(filename: 0x2ff7ce37, flags: CLOEXEC ) ... 0.013 ( ): __bpf_stdout__:/etc/ld.so.cache..B.................) 0.015 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7f7a2ff7ce37) 0.000 ( 0.021 ms): touch/14403 ... [continued]: open()) = 3 0.042 ( 0.002 ms): touch/14403 open(filename: 0x30180640, flags: CLOEXEC ) ... 0.044 ( ): __bpf_stdout__:/lib64/libc.so.6.. .......G.........) 0.045 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7f7a30180640) 0.042 ( 0.010 ms): touch/14403 ... [continued]: open()) = 3 0.301 ( 0.003 ms): touch/14403 open(filename: 0x2fd26c70, flags: CLOEXEC ) ... 0.305 ( ): __bpf_stdout__:/usr/lib/locale/locale-archive......) 0.306 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7f7a2fd26c70) 0.301 ( 0.011 ms): touch/14403 ... [continued]: open()) = 3 0.360 ( 0.002 ms): touch/14403 open(filename: 0x681f20f3, flags: CREAT|NOCTTY|NONBLOCK|WRONLY, mode: IRUGO|IWUGO) ... 0.362 ( ): __bpf_stdout__:/etc/passwd....... .......D.........) 0.363 ( ): perf_bpf_probe:prog:(ffffffffb4260ae0) filename=0x7ffe681f20f3) 0.360 ( 0.010 ms): touch/14403 ... [continued]: open()) = 3[root@jouet bpf]#
That
bpf.hwill set up the maps, etc, its attached if that may be needed to help figure this out.But then if I use the return value to push just the string lenght, it doesn't work:
[root@jouet bpf]# cat
open.c
#include "bpf.h"
Quoted 6 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);
The below issue 'invalid stack type R4 off=-128 access_size=0' is basically that unsigned len is unknown at verification time, thus unbounded. Can you try the following to see if that passes?
if (len > 0 && len <= sizeof(filename))
perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);
Quoted 67 lines
return 1; }[root@jouet bpf]# perf trace -e open,open.c touch
/etc/passwd
bpf: builtin compilation failed: -95, try external compiler event syntax error: 'open.c'
\___ Kernel verifier blocks program loading(add -v to see detail)
Run 'perf list' for a list of valid eventsUsage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]-e, --event <event> event/syscall selector. use 'perf list' to list available events
[root@jouet bpf]#When running this with -v we get the
tools/lib/libbpf.cdebug that may help here:Opening
/sys/kernel/debug/tracing//kprobe_eventswrite=1 Writing event: p:perf_bpf_probe/prog _text+2493152 filename=%si:x64 Inmap_prologue, ntevs=1
mapping[0]=0
libbpf: create map__bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r7 = r10
4: (07) r7 += -128
5: (bf) r1 = r7
6: (b7) r2 = 128
7: (85) call bpf_probe_read_str#45
8: (bf) r8 = r0
9: (67) r8 <<= 32
10: (77) r8 >>= 32
11: (85) call bpf_get_smp_processor_id#8
12: (bf) r1 = r6
13: (18) r2 =0xffffa0b5958e16c0
15: (bf) r3 = r0
16: (bf) r4 = r7
17: (bf) r5 = r8
18: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loadingI tried adding checks for len to try to somehow make sure its all bounds checked, but couldn't get past that "invalid stack type R4", the problem seems to be that access_size=0...
Ideas?
- Arnaldo
Em Mon, Nov 13, 2017 at 03:56:14PM +0100, Daniel Borkmann escreveu:
Quoted 30 lines · Arnaldo Carvalho de Melo
Quoted 22 lines
Hi,
In a5e8c07059d0 ("bpf: add bpf_probe_read_str helper") youstate:
"This is suboptimal because the size of the string needs to be estimated
at compile time, causing more memory to be copied than often necessary, and can become more problematic if further processing on buf is done, for example by pushing it to userspace viabpf_perf_event_output(), since the real length of the string is unknown and the entire buffer must be copied (and defining an unrolledstrnlen()inside the bpf program is a very inefficient and unfeasible approach)."So I went on to try this with 'perf trace' but it isn't working if I use the return from
bpf_probe_read_str(), I must be missing something here...I.e. this works:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 6 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, 32);
Quoted 2 lines
By the way, you can just use
BPF_F_CURRENT_CPUflag instead of the helper callget_smp_processor_id()to get current CPU.
Thanks, switched to it.
Quoted 13 lines
Quoted 6 lines
But then if I use the return value to push just the string lenght, it doesn't work:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 6 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);
Quoted 6 lines
The below issue 'invalid stack type R4 off=-128 access_size=0' is basically that unsigned len is unknown at verification time, thus unbounded. Can you try the following to see if that passes?
if (len > 0 && len <= sizeof(filename))
perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);
I had it like:
if (len > 0 && len < 32)
And it didn't helped, now I did exactly as you suggested:
[root@jouet bpf]# cat open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
int prog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
if (len > 0 && len <= sizeof(filename))
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len);
return 1;
}[root@jouet bpf]# trace -e open,open.c touch /etc/passwd
bpf: builtin compilation failed: -95, try external compiler event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
<SNIP>
[root@jouet bpf]#
The -v output looks the same:
[root@jouet bpf]# trace -v -e open,open.c touch /etc/passwd
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to /lib/modules/4.14.0-rc6+/build
set env: KBUILD_DIR=/lib/modules/4.14.0-rc6+/build
unset env: KBUILD_OPTS
include option is set to -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40e00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env: CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: WORKING_DIR=/lib/modules/4.14.0-rc6+/build
set env: CLANG_SOURCE=/home/acme/bpf/open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'open.c' from buffer
libbpf: section .strtab, size 103, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section prog=do_sys_open filename, size 184, link 0, flags 6, type=1
libbpf: found program prog=do_sys_open filename
libbpf: section .relprog=do_sys_open filename, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license of open.c is GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version of open.c is 40e00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps in open.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'prog=do_sys_open filename'
libbpf: relocation: insn_idx=15
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 15
bpf: config program 'prog=do_sys_open filename'
symbol:do_sys_open file:(null) line:0 offset:0 return:0 lazy:(null)
parsing arg: filename into filename
bpf: config 'prog=do_sys_open filename' is ok
Looking at the vmlinux_path (8 entries long)
Using /lib/modules/4.14.0-rc6+/build/vmlinux for symbols
Open Debuginfo file: /lib/modules/4.14.0-rc6+/build/vmlinux
Try to find probe point from debuginfo.
Matched function: do_sys_open [2a2bbbe]
Probe point found: do_sys_open+0
Searching 'filename' variable in context.
Converting variable filename into trace event.
filename type is (null).
Opening /sys/kernel/debug/tracing//README write=0
Found 1 probe_trace_events.
Opening /sys/kernel/debug/tracing//kprobe_events write=1
Writing event: p:perf_bpf_probe/prog _text+2493152 filename=%si:x64
In map_prologue, ntevs=1
mapping[0]=0
libbpf: create map __bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 > 0x7f goto pc+11
R0=inv(id=0) R1=inv(id=0,umax_value=127,var_off=(0x0; 0x7f)) R6=ctx(id=0,off=0,imm=0) R10=fp0
12: (67) r0 <<= 32
13: (77) r0 >>= 32
14: (bf) r4 = r10
15: (07) r4 += -128
16: (bf) r1 = r6
17: (18) r2 = 0xffffa0b74ba91000
19: (18) r3 = 0xffffffff
21: (bf) r5 = r0
22: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0
libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
Also:
[root@jouet bpf]# clang -v
clang version 4.0.0 (http://llvm.org/git/clang.git f5be8ba13adc4ba1011a7ccd60c844bd60427c1c) (http://llvm.org/git/llvm.git efca1a37676f4cd276d947658cf90b0fb625abfd)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir: /usr/local/bin
Found candidate GCC installation: /usr/lib/gcc/x86_64-redhat-linux/7
Selected GCC installation: /usr/lib/gcc/x86_64-redhat-linux/7
Candidate multilib: .;@m64
Candidate multilib: 32;@m32
Selected multilib: .;@m64
[root@jouet bpf]#
And now I've really attached that bpf.h header I use.
- Arnaldo
#include <uapi/linux/bpf.h>
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
#define SEC(NAME) __attribute__((section(NAME), used))
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
(void *)BPF_FUNC_perf_event_output;
static int (*bpf_probe_read_str)(void *dst, int size, const void *unsafe_addr) =
(void *)BPF_FUNC_probe_read_str;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;struct bpf_map_def SEC("maps") __bpf_stdout__ = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(u32),
.max_entries = __NR_CPUS__,
};
#define printf(msg) \
({ char str[] = msg; perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), &str, sizeof(str)); })char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
Daniel Borkmann
maintainerQuoted 143 lines · Arnaldo Carvalho de Melo
Em Mon, Nov 13, 2017 at 03:56:14PM +0100, Daniel Borkmann escreveu:
Quoted 30 lines · Arnaldo Carvalho de Melo
Quoted 22 lines
Hi,
In a5e8c07059d0 ("bpf: add bpf_probe_read_str helper") youstate:
"This is suboptimal because the size of the string needs to be estimated
at compile time, causing more memory to be copied than often necessary, and can become more problematic if further processing on buf is done, for example by pushing it to userspace viabpf_perf_event_output(), since the real length of the string is unknown and the entire buffer must be copied (and defining an unrolledstrnlen()inside the bpf program is a very inefficient and unfeasible approach)."So I went on to try this with 'perf trace' but it isn't working if I use the return from
bpf_probe_read_str(), I must be missing something here...I.e. this works:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 6 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, 32);Quoted 2 lines
By the way, you can just use
BPF_F_CURRENT_CPUflag instead of the helper callget_smp_processor_id()to get current CPU.Thanks, switched to it.
Quoted 13 lines
Quoted 6 lines
But then if I use the return value to push just the string lenght, it doesn't work:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 6 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);Quoted 6 lines
The below issue 'invalid stack type R4 off=-128 access_size=0' is basically that unsigned len is unknown at verification time, thus unbounded. Can you try the following to see if that passes?
if (len > 0 && len <= sizeof(filename))
perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(), filename, len);I had it like:
if (len > 0 && len < 32)And it didn't helped, now I did exactly as you suggested:
[root@jouet bpf]# cat
open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (len > 0 && len <= sizeof(filename)) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len); return 1; }[root@jouet bpf]# trace -e open,open.c touch
/etc/passwd
bpf: builtin compilation failed: -95, try external compiler event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
<SNIP>
[root@jouet bpf]#The -v output looks the same:
[root@jouet bpf]# trace -v -e open,open.c touch
/etc/passwd
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to/lib/modules/4.14.0-rc6+/build
set env: KBUILD_DIR=/lib/modules/4.14.0-rc6+/build
unset env:KBUILD_OPTS
include option is set to -nostdinc -isystem/usr/lib/gcc/x86_64-redhat-linux/7/include-I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include/home/acme/git/linux/include/linux/kconfig.hset env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40e00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env:CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem/usr/lib/gcc/x86_64-redhat-linux/7/include-I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include/home/acme/git/linux/include/linux/kconfig.hset env: WORKING_DIR=/lib/modules/4.14.0-rc6+/build
set env: CLANG_SOURCE=/home/acme/bpf/open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'open.c' from buffer
libbpf: section .strtab, size 103, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section prog=do_sys_open filename, size 184, link 0, flags 6, type=1
libbpf: found program prog=do_sys_open filename
libbpf: section .relprog=do_sys_open filename, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license ofopen.cis GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version ofopen.cis 40e00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps inopen.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'prog=do_sys_open filename'
libbpf: relocation: insn_idx=15
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 15
bpf: config program 'prog=do_sys_open filename'
symbol:do_sys_open file:(null) line:0 offset:0 return:0 lazy:(null)
parsing arg: filename into filename
bpf: config 'prog=do_sys_open filename' is ok
Looking at thevmlinux_path(8 entries long)
Using/lib/modules/4.14.0-rc6+/build/vmlinuxfor symbols
Open Debuginfo file:/lib/modules/4.14.0-rc6+/build/vmlinux
Try to find probe point from debuginfo.
Matched function:do_sys_open[2a2bbbe]
Probe point found: do_sys_open+0
Searching 'filename' variable in context.
Converting variable filename into trace event.
filename type is (null).
Opening/sys/kernel/debug/tracing//READMEwrite=0
Found 1probe_trace_events.
Opening/sys/kernel/debug/tracing//kprobe_eventswrite=1
Writing event: p:perf_bpf_probe/prog _text+2493152 filename=%si:x64
Inmap_prologue, ntevs=1
mapping[0]=0
libbpf: create map__bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10
Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11
Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int prog([...])
{
char filename[128];
int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
if (ret > 0)
bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename,
ret & (sizeof(filename) - 1));
return 1;
}r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1
1: bf a1 00 00 00 00 00 00 r1 = r10
2: 07 01 00 00 80 ff ff ff r1 += -128
3: b7 02 00 00 80 00 00 00 r2 = 128
4: 85 00 00 00 2d 00 00 00 call 45
5: 67 00 00 00 20 00 00 00 r0 <<= 32
6: c7 00 00 00 20 00 00 00 r0 s>>= 32
7: b7 01 00 00 01 00 00 00 r1 = 1
8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10
9: 57 00 00 00 7f 00 00 00 r0 &= 127
10: bf a4 00 00 00 00 00 00 r4 = r10
11: 07 04 00 00 80 ff ff ff r4 += -128
12: bf 61 00 00 00 00 00 00 r1 = r6
13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll
15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll
17: bf 05 00 00 00 00 00 00 r5 = r0
18: 85 00 00 00 19 00 00 00 call 25[1] http://patchwork.ozlabs.org/project/netdev/list/?series=13211
Quoted 38 lines
R0=inv(id=0) R1=inv(id=0,umax_value=127,var_off=(0x0;
0x7f)) R6=ctx(id=0,off=0,imm=0) R10=fp0
12: (67) r0 <<= 32
13: (77) r0 >>= 32
14: (bf) r4 = r10
15: (07) r4 += -128
16: (bf) r1 = r6
17: (18) r2 =0xffffa0b74ba91000
19: (18) r3 =0xffffffff
21: (bf) r5 = r0
22: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loadingAlso:
[root@jouet bpf]# clang -v
clang version 4.0.0 (http://llvm.org/git/clang.gitf5be8ba13adc4ba1011a7ccd60c844bd60427c1c) (http://llvm.org/git/llvm.gitefca1a37676f4cd276d947658cf90b0fb625abfd)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir:/usr/local/bin
Found candidate GCC installation:/usr/lib/gcc/x86_64-redhat-linux/7
Selected GCC installation:/usr/lib/gcc/x86_64-redhat-linux/7
Candidate multilib: .;@m64
Candidate multilib: 32;@m32
Selected multilib: .;@m64
[root@jouet bpf]#And now I've really attached that
bpf.hheader I use.- Arnaldo
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...]){ char filename[128]; int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1; }r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1 1: bf a1 00 00 00 00 00 00 r1 = r10 2: 07 01 00 00 80 ff ff ff r1 += -128 3: b7 02 00 00 80 00 00 00 r2 = 128 4: 85 00 00 00 2d 00 00 00 call 45 5: 67 00 00 00 20 00 00 00 r0 <<= 32 6: c7 00 00 00 20 00 00 00 r0 s>>= 32 7: b7 01 00 00 01 00 00 00 r1 = 1 8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10 9: 57 00 00 00 7f 00 00 00 r0 &= 127 10: bf a4 00 00 00 00 00 00 r4 = r10 11: 07 04 00 00 80 ff ff ff r4 += -128 12: bf 61 00 00 00 00 00 00 r1 = r6 13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0 18: 85 00 00 00 19 00 00 00 call 25
Quoted 1 line
[1] http://patchwork.ozlabs.org/project/netdev/list/?series=13211
Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 == 0x0 goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 = 0xffff92bfc2aba840
17: (18) r3 = 0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0
I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
int prog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
if (len > 0)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename,
len & (sizeof(filename) - 1));
return 1;
}[root@jouet bpf]# perf trace -v -e *open,open.c usleep 2
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to /lib/modules/4.14.0+/build
set env: KBUILD_DIR=/lib/modules/4.14.0+/build
unset env: KBUILD_OPTS
include option is set to -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40e00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env: CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: WORKING_DIR=/lib/modules/4.14.0+/build
set env: CLANG_SOURCE=/home/acme/bpf/open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'open.c' from buffer
libbpf: section .strtab, size 103, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section prog=do_sys_open filename, size 168, link 0, flags 6, type=1
libbpf: found program prog=do_sys_open filename
libbpf: section .relprog=do_sys_open filename, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license of open.c is GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version of open.c is 40e00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps in open.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'prog=do_sys_open filename'
libbpf: relocation: insn_idx=13
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 13
bpf: config program 'prog=do_sys_open filename'
symbol:do_sys_open file:(null) line:0 offset:0 return:0 lazy:(null)
parsing arg: filename into filename
bpf: config 'prog=do_sys_open filename' is ok
Failed to open cache(-1): /root/.debug/[kernel.kallsyms]/b5e25bb34abb3ddefa85250978beb530cc2eebd5/probes
Cache open error: -1
Looking at the vmlinux_path (8 entries long)
Using /lib/modules/4.14.0+/build/vmlinux for symbols
Open Debuginfo file: /lib/modules/4.14.0+/build/vmlinux
Try to find probe point from debuginfo.
Matched function: do_sys_open [2a2e5db]
Probe point found: do_sys_open+0
Searching 'filename' variable in context.
Converting variable filename into trace event.
filename type is (null).
Opening /sys/kernel/debug/tracing//README write=0
Found 1 probe_trace_events.
Opening /sys/kernel/debug/tracing//kprobe_events write=1
Writing event: p:perf_bpf_probe/prog _text+2493840 filename=%si:x64
In map_prologue, ntevs=1
mapping[0]=0
libbpf: create map __bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 == 0x0 goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 = 0xffff92bfc2aba840
17: (18) r3 = 0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0
libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
Opening /sys/kernel/debug/tracing//kprobe_events write=1 Opening /sys/kernel/debug/tracing//uprobe_events write=1 Parsing probe_events: p:perf_bpf_probe/prog _text+2493840 filename=%si:x64 Group:perf_bpf_probe Event:prog probe:p
Writing event: -:perf_bpf_probe/prog
[root@jouet bpf]#
Daniel Borkmann
maintainerQuoted 78 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...]){ char filename[128]; int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1; }r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1 1: bf a1 00 00 00 00 00 00 r1 = r10 2: 07 01 00 00 80 ff ff ff r1 += -128 3: b7 02 00 00 80 00 00 00 r2 = 128 4: 85 00 00 00 2d 00 00 00 call 45 5: 67 00 00 00 20 00 00 00 r0 <<= 32 6: c7 00 00 00 20 00 00 00 r0 s>>= 32 7: b7 01 00 00 01 00 00 00 r1 = 1 8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10 9: 57 00 00 00 7f 00 00 00 r0 &= 127 10: bf a4 00 00 00 00 00 00 r4 = r10 11: 07 04 00 00 80 ff ff ff r4 += -128 12: bf 61 00 00 00 00 00 00 r1 = r6 13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0 18: 85 00 00 00 19 00 00 00 call 25Quoted 1 line
[1] http://patchwork.ozlabs.org/project/netdev/list/?series=13211
Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
Quoted 8 lines
[root@jouet bpf]# cat
open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
Btw, I was using 'int' here above instead of 'unsigned' as strncpy_from_unsafe() could potentially return errors like -EFAULT.
Currently having a version compiled from the git tree:
# llc --version LLVM (http://llvm.org/): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylake
Registered Targets:
bpf - BPF (host endian)
bpfeb - BPF (big endian)
bpfel - BPF (little endian)
x86 - 32-bit X86: Pentium-Pro and above
x86-64 - 64-bit X86: EM64T and AMD64Quoted 5 lines
if (len > 0)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename,
len & (sizeof(filename) - 1));
return 1;
}Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...]){ char filename[128]; int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1; }r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1 1: bf a1 00 00 00 00 00 00 r1 = r10 2: 07 01 00 00 80 ff ff ff r1 += -128 3: b7 02 00 00 80 00 00 00 r2 = 128 4: 85 00 00 00 2d 00 00 00 call 45 5: 67 00 00 00 20 00 00 00 r0 <<= 32 6: c7 00 00 00 20 00 00 00 r0 s>>= 32 7: b7 01 00 00 01 00 00 00 r1 = 1 8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10 9: 57 00 00 00 7f 00 00 00 r0 &= 127 10: bf a4 00 00 00 00 00 00 r4 = r10 11: 07 04 00 00 80 ff ff ff r4 += -128 12: bf 61 00 00 00 00 00 00 r1 = r6 13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0 18: 85 00 00 00 19 00 00 00 call 25Quoted 1 line
[1] http://patchwork.ozlabs.org/project/netdev/list/?series=13211
Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"
Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.
I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (http://llvm.org/): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylake
[root@jouet bpf]# llc --version
LLVM (http://llvm.org/):
LLVM version 4.0.0svn
Old stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
- Arnaldo
Quoted 12 lines
Registered Targets:
bpf - BPF (host endian) bpfeb - BPF (big endian) bpfel - BPF (little endian) x86 - 32-bit X86: Pentium-Pro and above x86-64 - 64-bit X86: EM64T and AMD64Quoted 5 lines
if (len > 0) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len & (sizeof(filename) - 1)); return 1; }
Daniel Borkmann
maintainerQuoted 103 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...]){ char filename[128]; int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1; }r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1 1: bf a1 00 00 00 00 00 00 r1 = r10 2: 07 01 00 00 80 ff ff ff r1 += -128 3: b7 02 00 00 80 00 00 00 r2 = 128 4: 85 00 00 00 2d 00 00 00 call 45 5: 67 00 00 00 20 00 00 00 r0 <<= 32 6: c7 00 00 00 20 00 00 00 r0 s>>= 32 7: b7 01 00 00 01 00 00 00 r1 = 1 8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10 9: 57 00 00 00 7f 00 00 00 r0 &= 127 10: bf a4 00 00 00 00 00 00 r4 = r10 11: 07 04 00 00 80 ff ff ff r4 += -128 12: bf 61 00 00 00 00 00 00 r1 = r6 13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0 18: 85 00 00 00 19 00 00 00 call 25Quoted 1 line
[1] http://patchwork.ozlabs.org/project/netdev/list/?series=13211
Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840u
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (http://llvm.org/): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylake
Quoted 6 lines
[root@jouet bpf]# llc --version
LLVM (http://llvm.org/):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 == 0x0 goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
[...]
... the shifts on r1 might be due to using 32 bit type, so if you find a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
Quoted 14 lines
- Arnaldo
Quoted 12 lines
Registered Targets:
bpf - BPF (host endian) bpfeb - BPF (big endian) bpfel - BPF (little endian) x86 - 32-bit X86: Pentium-Pro and above x86-64 - 64-bit X86: EM64T and AMD64Quoted 5 lines
if (len > 0) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len & (sizeof(filename) - 1)); return 1; }
Em Tue, Nov 14, 2017 at 03:19:51PM +0100, Daniel Borkmann escreveu:
Quoted 35 lines · Arnaldo Carvalho de Melo
Quoted 11 lines
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 9 lines · Arnaldo Carvalho de Melo
Currently having a version compiled from the git tree:
# llc --version LLVM (http://llvm.org/): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylakeQuoted 6 lines
[root@jouet bpf]# llc --version
LLVM (http://llvm.org/):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
[...]
Quoted 5 lines
... the shifts on r1 might be due to using 32 bit type, so if you find a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
I changed len/ret to u64, didn't help, updating clang and llvm to see if that helps...
Will end up working directly with eBPF bytecode, which is what I really need in 'perf trace', but lets get this sorted out first.
- Arnaldo
Quoted 127 lines · Daniel Borkmann
Quoted 103 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...]){ char filename[128]; int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1; }r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1 1: bf a1 00 00 00 00 00 00 r1 = r10 2: 07 01 00 00 80 ff ff ff r1 += -128 3: b7 02 00 00 80 00 00 00 r2 = 128 4: 85 00 00 00 2d 00 00 00 call 45 5: 67 00 00 00 20 00 00 00 r0 <<= 32 6: c7 00 00 00 20 00 00 00 r0 s>>= 32 7: b7 01 00 00 01 00 00 00 r1 = 1 8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10 9: 57 00 00 00 7f 00 00 00 r0 &= 127 10: bf a4 00 00 00 00 00 00 r4 = r10 11: 07 04 00 00 80 ff ff ff r4 += -128 12: bf 61 00 00 00 00 00 00 r1 = r6 13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0 18: 85 00 00 00 19 00 00 00 call 25Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840u
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; const unsigned len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylakeQuoted 6 lines
[root@jouet bpf]# llc --version
LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
[...]
Quoted 5 lines
... the shifts on r1 might be due to using 32 bit type, so if you find a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
It is tricky to do in the bpf_program. Compiler tries hard to optimize :-).
The issue is at "r0 &= 127".
9: (6d) if r1 s> r0 goto pc+10
R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0
10: R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0; 0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: (57) r0 &= 127
11: R0=inv(id=0,umax_value=127,var_off=(0x0; 0x7f)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0
One possible solution for this problem is to relax the arg4 type from ARG_CONST_SIZE to ARG_CONST_SIZE_OR_ZERO.
kernel/trace/bpf_trace.c +4 -1
@@ -393,6 +393,9 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, }, }; +if (unlikely (size ==0 )) +return 0 ; +if (unlikely (flags & ~(BPF_F_INDEX_MASK )))return -EINVAL ;
@@ -407,7 +410,7 @@ static const struct bpf_func_proto bpf_perf_event_output_proto = { .arg2_type = ARG_CONST_MAP_PTR, .arg3_type = ARG_ANYTHING, .arg4_type = ARG_PTR_TO_MEM, - .arg5_type = ARG_CONST_SIZE, + .arg5_type = ARG_CONST_SIZE_OR_ZERO, };
Quoted 16 lines
Quoted 15 lines
- Arnaldo
Quoted 12 lines
Registered Targets:
bpf - BPF (host endian) bpfeb - BPF (big endian) bpfel - BPF (little endian) x86 - 32-bit X86: Pentium-Pro and above x86-64 - 64-bit X86: EM64T and AMD64Quoted 5 lines
if (len > 0) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len & (sizeof(filename) - 1)); return 1; }
Daniel Borkmann
maintainerQuoted 138 lines · Yonghong Song
Quoted 127 lines · Daniel Borkmann
Quoted 103 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...])
{
char filename[128];
int ret =bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0)
bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1;
}r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1
1: bf a1 00 00 00 00 00 00 r1 = r10
2: 07 01 00 00 80 ff ff ff r1 += -128
3: b7 02 00 00 80 00 00 00 r2 = 128
4: 85 00 00 00 2d 00 00 00 call 45
5: 67 00 00 00 20 00 00 00 r0 <<= 32
6: c7 00 00 00 20 00 00 00 r0 s>>= 32
7: b7 01 00 00 01 00 00 00 r1 = 1
8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10
9: 57 00 00 00 7f 00 00 00 r0 &= 127
10: bf a4 00 00 00 00 00 00 r4 = r10
11: 07 04 00 00 80 ff ff ff r4 += -128
12: bf 61 00 00 00 00 00 00 r1 = r6
13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0
18: 85 00 00 00 19 00 00 00 call 25Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0 11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840u
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len =bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylakeQuoted 6 lines
[root@jouet bpf]# llc --version
LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0 11: (57) r0 &= 127
[...]Quoted 5 lines
... the shifts on r1 might be due to using 32 bit type, so if you find a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
It is tricky to do in the
bpf_program. Compiler tries hard to optimize :-).The issue is at "r0 &= 127".
Quoted 8 lines
9: (6d) if r1 s> r0 goto pc+10
R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: (57) r0 &= 127
11: R0=inv(id=0,umax_value=127,var_off=(0x0;0x7f)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0One possible solution for this problem is to relax the arg4 type from
ARG_CONST_SIZEtoARG_CONST_SIZE_OR_ZERO.
Yeah, I know, that's what I mentioned earlier in this thread to resolve it, but do we really want to add this hack everywhere? :( Potentially any function having ARG_CONST_SIZE would need to handle size 0 and bail out again in their helper implementation and it ends up that progs start relying on this runtime check where we won't be able to get rid of it later on anymore.
Quoted 21 lines
kernel/trace/bpf_trace.c +4
-1@@ -393,6 +393,9 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, }, }; + if (unlikely(size == 0)) + return 0; + if (unlikely(flags & ~(BPF_F_INDEX_MASK))) return -EINVAL;@@ -407,7 +410,7 @@ static const struct bpf_func_proto bpf_perf_event_output_proto = { .arg2_type = ARG_CONST_MAP_PTR, .arg3_type = ARG_ANYTHING, .arg4_type = ARG_PTR_TO_MEM, - .arg5_type = ARG_CONST_SIZE, + .arg5_type = ARG_CONST_SIZE_OR_ZERO, };
Quoted 145 lines · Daniel Borkmann
Quoted 144 lines · Yonghong Song
Quoted 127 lines · Daniel Borkmann
Quoted 103 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...])
{
char filename[128];
int ret =bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0)
bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1;
}r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1
1: bf a1 00 00 00 00 00 00 r1 = r10
2: 07 01 00 00 80 ff ff ff r1 += -128
3: b7 02 00 00 80 00 00 00 r2 = 128
4: 85 00 00 00 2d 00 00 00 call 45
5: 67 00 00 00 20 00 00 00 r0 <<= 32
6: c7 00 00 00 20 00 00 00 r0 s>>= 32
7: b7 01 00 00 01 00 00 00 r1 = 1
8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10
9: 57 00 00 00 7f 00 00 00 r0 &= 127
10: bf a4 00 00 00 00 00 00 r4 = r10
11: 07 04 00 00 80 ff ff ff r4 += -128
12: bf 61 00 00 00 00 00 00 r1 = r6
13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0
18: 85 00 00 00 19 00 00 00 call 25Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840u
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len =bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylakeQuoted 6 lines
[root@jouet bpf]# llc --version
LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0 11: (57) r0 &= 127
[...]Quoted 5 lines
... the shifts on r1 might be due to using 32 bit type, so if you find a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
It is tricky to do in the
bpf_program. Compiler tries hard to optimize :-).The issue is at "r0 &= 127".
9: (6d) if r1 s> r0 goto pc+10
R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0
10: R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: (57) r0 &= 127
11: R0=inv(id=0,umax_value=127,var_off=(0x0;0x7f)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0
Quoted 8 lines
Quoted 2 lines
One possible solution for this problem is to relax the arg4 type from
ARG_CONST_SIZEtoARG_CONST_SIZE_OR_ZERO.Yeah, I know, that's what I mentioned earlier in this thread to resolve it, but do we really want to add this hack everywhere? :( Potentially any function having
ARG_CONST_SIZEwould need to handle size 0 and bail out again in their helper implementation and it ends up that progs start relying on this runtime check where we won't be able to get rid of it later on anymore.
The compiler actually does the right thing for the below code:
int ret = bpf_probe_read_str(filename, sizeof(filename),
filename_ptr);
if (ret > 0)
bpf_perf_event_output(ctx, &__bpf_stdout__,BPF_F_CURRENT_CPU,
filename, ret & (sizeof(filename) - 1));Just from the above code without consulting bpf_probe_read_str internals, it is totally possible that ret = 128, then
ret & (sizeof(filename) - 1) = 0.
The issue is that the verifier did not set the "ret" initial range as (-inf, sizeof(filename) - 1). We could have this information associated with helper and feed back to verifier.
If we have this range, later for ret & (sizeof(filename) - 1) with ret >= 1, the verifier should be able to conclude
ret & (sizeof(filename) - 1) >= 1.
To workaround the immediate problem, I tested the following hack with bcc and it works fine.
BPF_PERF_OUTPUT(events);
int trace(struct pt_regs *ctx) {
char filename[128];
int ret = bpf_probe_read_str(filename, sizeof(filename), 0);
if (ret > 0) {
if (ret == 1)
events.perf_submit(ctx, filename, ret);
else if (ret < 128)
events.perf_submit(ctx, filename, ret); }
return 1;
}
The idea is to make control flow more complex to prevent llvm do certain optimizations.
Quoted 23 lines
Quoted 22 lines
kernel/trace/bpf_trace.c +4
-1@@ -393,6 +393,9 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, }, }; + if (unlikely(size == 0)) + return 0; + if (unlikely(flags & ~(BPF_F_INDEX_MASK))) return -EINVAL;@@ -407,7 +410,7 @@ static const struct bpf_func_proto bpf_perf_event_output_proto = { .arg2_type = ARG_CONST_MAP_PTR, .arg3_type = ARG_ANYTHING, .arg4_type = ARG_PTR_TO_MEM, - .arg5_type = ARG_CONST_SIZE, + .arg5_type = ARG_CONST_SIZE_OR_ZERO, };
Em Tue, Nov 14, 2017 at 02:58:24PM -0800, Yonghong Song escreveu:
Quoted 38 lines · Daniel Borkmann
Quoted 5 lines
Yeah, I know, that's what I mentioned earlier in this thread to resolve it, but do we really want to add this hack everywhere? :( Potentially any function having
ARG_CONST_SIZEwould need to handle size 0 and bail out again in their helper implementation and it ends up that progs start relying on this runtime check where we won't be able to get rid of it later on anymore.The compiler actually does the right thing for the below code:
int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__,BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1));Just from the above code without consulting
bpf_probe_read_strinternals, it is totally possible that ret = 128, then
ret & (sizeof(filename) - 1) = 0.The issue is that the verifier did not set the "ret" initial range as (-inf, sizeof(filename) - 1). We could have this information associated with helper and feed back to verifier.
If we have this range, later for ret & (sizeof(filename) - 1) with ret >= 1, the verifier should be able to conclude
ret & (sizeof(filename) - 1) >= 1.To workaround the immediate problem, I tested the following hack with bcc and it works fine.
BPF_PERF_OUTPUT(events);
inttrace(struct pt_regs *ctx){
char filename[128];
int ret =bpf_probe_read_str(filename, sizeof(filename), 0);if (ret > 0) { if (ret == 1) events.perf_submit(ctx, filename, ret); else if (ret < 128) events.perf_submit(ctx, filename, ret);
Quoted 6 lines
}
return 1;
}The idea is to make control flow more complex to prevent llvm do certain optimizations.
So, the hack makes it work for me, using clang 6.0:
set env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40e00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env: CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: WORKING_DIR=/lib/modules/4.14.0+/build
set env: CLANG_SOURCE=/home/acme/bpf/open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
[root@jouet bpf]# perf probe -V do_sys_open
Available variables at do_sys_open
@<do_sys_open+0>
char* filename
int dfd
int flags
struct open_flags op
umode_t mode
[root@jouet bpf]# cat open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
int prog(void *ctx, int err, char *filename_ptr)
{
char filename[128];
int len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
if (len > 0) {
if (len == 1)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len);
else if (len < 128)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len);
}
return 1;
}
[root@jouet bpf]#
[root@jouet bpf]# perf trace -e *open,open.c touch /tmp/Thanks.Yonghong.Song\!
LLVM: dumping open.o
0.000 ( 0.009 ms): touch/9034 open(filename: 0x5b678e37, flags: CLOEXEC ) ...
0.009 ( ): __bpf_stdout__:/etc/ld.so.cache....)
0.011 ( ): perf_bpf_probe:prog:(ffffffff8f260da0) filename=0x7f805b678e37)
0.000 ( 0.016 ms): touch/9034 ... [continued]: open()) = 3
0.034 ( 0.002 ms): touch/9034 open(filename: 0x5b87c640, flags: CLOEXEC ) ...
0.036 ( ): __bpf_stdout__:/lib64/libc.so.6....)
0.037 ( ): perf_bpf_probe:prog:(ffffffff8f260da0) filename=0x7f805b87c640)
0.034 ( 0.009 ms): touch/9034 ... [continued]: open()) = 3
0.251 ( 0.002 ms): touch/9034 open(filename: 0x5b422c70, flags: CLOEXEC ) ...
0.253 ( ): __bpf_stdout__:/usr/lib/locale/locale-archive......)
0.254 ( ): perf_bpf_probe:prog:(ffffffff8f260da0) filename=0x7f805b422c70)
0.251 ( 0.009 ms): touch/9034 ... [continued]: open()) = 3
0.296 ( 0.002 ms): touch/9034 open(filename: 0x1d3a00f1, flags: CREAT|NOCTTY|NONBLOCK|WRONLY, mode: IRUGO|IWUGO) ...
0.298 ( ): __bpf_stdout__:/tmp/Thanks.Yonghong.Song!..)
0.299 ( ): perf_bpf_probe:prog:(ffffffff8f260da0) filename=0x7ffd1d3a00f1)
0.296 ( 0.009 ms): touch/9034 ... [continued]: open()) = 3[root@jouet bpf]#
Quoted 39 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 02:58:24PM -0800, Yonghong Song escreveu:
Quoted 38 lines · Daniel Borkmann
Quoted 5 lines
Yeah, I know, that's what I mentioned earlier in this thread to resolve it, but do we really want to add this hack everywhere? :( Potentially any function having
ARG_CONST_SIZEwould need to handle size 0 and bail out again in their helper implementation and it ends up that progs start relying on this runtime check where we won't be able to get rid of it later on anymore.The compiler actually does the right thing for the below code:
int ret = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0) bpf_perf_event_output(ctx, &__bpf_stdout__,BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1));Just from the above code without consulting
bpf_probe_read_strinternals, it is totally possible that ret = 128, then
ret & (sizeof(filename) - 1) = 0.The issue is that the verifier did not set the "ret" initial range as (-inf, sizeof(filename) - 1). We could have this information associated with helper and feed back to verifier.
If we have this range, later for ret & (sizeof(filename) - 1) with ret >= 1, the verifier should be able to conclude
ret & (sizeof(filename) - 1) >= 1.To workaround the immediate problem, I tested the following hack with bcc and it works fine.
BPF_PERF_OUTPUT(events);
inttrace(struct pt_regs *ctx){
char filename[128];
int ret =bpf_probe_read_str(filename, sizeof(filename), 0);if (ret > 0) { if (ret == 1) events.perf_submit(ctx, filename, ret); else if (ret < 128) events.perf_submit(ctx, filename, ret);
...
Quoted 11 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, char *filename_ptr){ char filename[128]; int len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (len > 0) { if (len == 1) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len); else if (len < 128) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len);
yeah sorry about this hack. Gianluca reported this issue as well. Yonghong fixed it for bpf_probe_read only. We will extend the fix to bpf_probe_read_str() and bpf_perf_event_output() asap. The above workaround gets too much into llvm and verifier details we should strive to make bpf program writing as easy as possible.
Quoted 7 lines · Alexei Starovoitov
yeah sorry about this hack. Gianluca reported this issue as well. Yonghong fixed it for
bpf_probe_readonly. We will extend the fix tobpf_probe_read_str()andbpf_perf_event_output()asap. The above workaround gets too much into llvm and verifier details we should strive to make bpf program writing as easy as possible.
Hi Arnaldo
With the help of Alexei, Daniel and Yonghong I just submitted a new series ("bpf: fix semantics issues with helpers receiving NULL arguments") that includes a fix in bpf_perf_event_output. This should simplify the way you write your bpf programs, so you shouldn't be required to write those convoluted branches anymore (there are a few usage examples in the commit log).
In my case it made writing the code much easier, after applying it I haven't been surprised by the compiler output in a while, and I hope your experience will be improved as well.
Thanks
Em Wed, Nov 22, 2017 at 10:42:22AM -0800, Gianluca Borello escreveu:
Quoted 19 lines
Quoted 7 lines · Alexei Starovoitov
yeah sorry about this hack. Gianluca reported this issue as well. Yonghong fixed it for
bpf_probe_readonly. We will extend the fix tobpf_probe_read_str()andbpf_perf_event_output()asap. The above workaround gets too much into llvm and verifier details we should strive to make bpf program writing as easy as possible.Hi Arnaldo
With the help of Alexei, Daniel and Yonghong I just submitted a new series ("bpf: fix semantics issues with helpers receiving NULL arguments") that includes a fix in
bpf_perf_event_output. This should simplify the way you write your bpf programs, so you shouldn't be required to write those convoluted branches anymore (there are a few usage examples in the commit log).
Quoted 3 lines
In my case it made writing the code much easier, after applying it I haven't been surprised by the compiler output in a while, and I hope your experience will be improved as well.
Trying to work with this again, and I still need to trick clang into not doing some optimizations that end up getting the resulting eBPF object rejected by the kernel verifier:
[root@jouet bpf]# uname -a
Linux jouet 4.15.0-rc8+ #1 SMP Wed Jan 17 11:01:34 -03 2018 x86_64 x86_64 x86_64 GNU/Linux [root@jouet bpf]# grep -i bpf /lib/modules/uname -r/build/.config CONFIG_CGROUP_BPF=yCONFIG_BPF=yCONFIG_BPF_SYSCALL=yCONFIG_BPF_JIT_ALWAYS_ON=y
# CONFIG_NETFILTER_XT_MATCH_BPF is not set
# CONFIG_NET_CLS_BPF is not set
# CONFIG_NET_ACT_BPF is not setCONFIG_BPF_JIT=yCONFIG_BPF_STREAM_PARSER=yCONFIG_LWTUNNEL_BPF=yCONFIG_HAVE_EBPF_JIT=yCONFIG_BPF_EVENTS=y
# CONFIG_TEST_BPF is not set
[root@jouet bpf]# cat sys_enter_open.c
#include "bpf.h"
SEC("syscalls:sys_enter_open")
int func(void *ctx)
{
struct {
char *ptr;
char path[256];
} filename = {
/*
* /sys/kernel/debug/tracing/events/syscalls/sys_enter_open/format:
*
* ...
* field:const char * filename; offset:16; size:8; signed:0;
* ...
* ctx + 16 selects 'filename'
*/
.ptr = *((char **)(ctx + 16)),
};
int len = bpf_probe_read_str(filename.path, sizeof(filename.path), filename.ptr);
if (len > 0 && len < 256)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr));
return 0;
}[root@jouet bpf]#
[root@jouet bpf]# perf trace -v -e open,sys_enter_open.c
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to /lib/modules/4.15.0-rc8+/build
set env: KBUILD_DIR=/lib/modules/4.15.0-rc8+/build
unset env: KBUILD_OPTS
include option is set to -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40f00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env: CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: WORKING_DIR=/lib/modules/4.15.0-rc8+/build
set env: CLANG_SOURCE=/home/acme/bpf/sys_enter_open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'sys_enter_open.c' from buffer
libbpf: section .strtab, size 101, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section syscalls:sys_enter_open, size 472, link 0, flags 6, type=1
libbpf: found program syscalls:sys_enter_open
libbpf: section .relsyscalls:sys_enter_open, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license of sys_enter_open.c is GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version of sys_enter_open.c is 40f00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps in sys_enter_open.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'syscalls:sys_enter_open'
libbpf: relocation: insn_idx=51
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 51
LLVM: dumping sys_enter_open.o
bpf: config program 'syscalls:sys_enter_open'
libbpf: create map __bpf_stdout__: fd=3
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (bf) r6 = r1
1: (b7) r1 = 0
2: (7b) *(u64 *)(r10 -8) = r1
3: (7b) *(u64 *)(r10 -16) = r1
4: (7b) *(u64 *)(r10 -24) = r1
5: (7b) *(u64 *)(r10 -32) = r1
6: (7b) *(u64 *)(r10 -40) = r1
7: (7b) *(u64 *)(r10 -48) = r1
8: (7b) *(u64 *)(r10 -56) = r1
9: (7b) *(u64 *)(r10 -64) = r1
10: (7b) *(u64 *)(r10 -72) = r1
11: (7b) *(u64 *)(r10 -80) = r1
12: (7b) *(u64 *)(r10 -88) = r1
13: (7b) *(u64 *)(r10 -96) = r1
14: (7b) *(u64 *)(r10 -104) = r1
15: (7b) *(u64 *)(r10 -112) = r1
16: (7b) *(u64 *)(r10 -120) = r1
17: (7b) *(u64 *)(r10 -128) = r1
18: (7b) *(u64 *)(r10 -136) = r1
19: (7b) *(u64 *)(r10 -144) = r1
20: (7b) *(u64 *)(r10 -152) = r1
21: (7b) *(u64 *)(r10 -160) = r1
22: (7b) *(u64 *)(r10 -168) = r1
23: (7b) *(u64 *)(r10 -176) = r1
24: (7b) *(u64 *)(r10 -184) = r1
25: (7b) *(u64 *)(r10 -192) = r1
26: (7b) *(u64 *)(r10 -200) = r1
27: (7b) *(u64 *)(r10 -208) = r1
28: (7b) *(u64 *)(r10 -216) = r1
29: (7b) *(u64 *)(r10 -224) = r1
30: (7b) *(u64 *)(r10 -232) = r1
31: (7b) *(u64 *)(r10 -240) = r1
32: (7b) *(u64 *)(r10 -248) = r1
33: (7b) *(u64 *)(r10 -256) = r1
34: (79) r3 = *(u64 *)(r6 +16)
35: (7b) *(u64 *)(r10 -264) = r3
36: (bf) r1 = r10
37: (07) r1 += -256
38: (b7) r2 = 256
39: (85) call bpf_probe_read_str#45
40: (bf) r1 = r0
41: (07) r1 += -1
42: (67) r1 <<= 32
43: (77) r1 >>= 32
44: (25) if r1 > 0xfe goto pc+12
R0=inv(id=0) R1=inv(id=0,umax_value=254,var_off=(0x0; 0xff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
45: (67) r0 <<= 32
46: (c7) r0 s>>= 32
47: (07) r0 += 8
48: (bf) r4 = r10
49: (07) r4 += -264
50: (bf) r1 = r6
51: (18) r2 = 0xffff957fc90e1000
53: (18) r3 = 0xffffffff
55: (bf) r5 = r0
56: (85) call bpf_perf_event_output#25
R5 min value is negative, either use unsigned or 'var &= const'
libbpf: -- END LOG --
libbpf: failed to load program 'syscalls:sys_enter_open'
libbpf: failed to load object 'sys_enter_open.c'
bpf: load objects failed
event syntax error: 'sys_enter_open.c'
\___ Kernel verifier blocks program loading
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
[root@jouet bpf]#
If I use the version we came up before, it works:
[root@jouet bpf]# trace -e open,sys_enter_open.c
LLVM: dumping sys_enter_open.o
2171.820 ( ): __bpf_stdout__:@......./proc/self/task/14683/comm..)
2171.845 ( 0.339 ms): qemu-system-x8/6721 open(filename: /proc/self/task/14683/comm, flags: RDWR ) = 91
^C[root@jouet bpf]
[root@jouet bpf]# trace -e open,sys_enter_open.o
2485.416 ( ): __bpf_stdout__:Vm..u.../proc/loadavg.......)
2485.434 ( 0.210 ms): lighttpd/25120 open(filename: /proc/loadavg, mode: ISGID|IXOTH ) = 8
^C[root@jouet bpf]#
[root@jouet bpf]cat sys_enter_open.c
#include "bpf.h"
SEC("syscalls:sys_enter_open")
int func(void *ctx)
{
struct {
char *ptr;
char path[256];
} filename = {
/*
* /sys/kernel/debug/tracing/events/syscalls/sys_enter_open/format:
*
* ...
* field:const char * filename; offset:16; size:8; signed:0;
* ...
* ctx + 16 selects 'filename'
*/
.ptr = *((char **)(ctx + 16)),
};
int len = bpf_probe_read_str(filename.path, sizeof(filename.path), filename.ptr);
if (len > 0) {
if (len == 1)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr));
else if (len < 256)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr));
}
return 0;
}[root@jouet bpf]#
[acme@jouet perf]$ llc --version | head -16
LLVM (http://llvm.org/):
LLVM version 6.0.0svn
DEBUG build with assertions.
Default target: x86_64-unknown-linux-gnu
Host CPU: broadwell
Registered Targets:
aarch64 - AArch64 (little endian)
aarch64_be - AArch64 (big endian)
amdgcn - AMD GCN GPUs
arm - ARM
arm64 - ARM64 (little endian)
armeb - ARM (big endian)
bpf - BPF (host endian)
bpfeb - BPF (big endian)
bpfel - BPF (little endian)
[acme@jouet perf]$
Quoted 156 lines · Arnaldo Carvalho de Melo
Em Wed, Nov 22, 2017 at 10:42:22AM -0800, Gianluca Borello escreveu:
Quoted 19 lines
Quoted 7 lines · Alexei Starovoitov
yeah sorry about this hack. Gianluca reported this issue as well. Yonghong fixed it for
bpf_probe_readonly. We will extend the fix tobpf_probe_read_str()andbpf_perf_event_output()asap. The above workaround gets too much into llvm and verifier details we should strive to make bpf program writing as easy as possible.Hi Arnaldo
With the help of Alexei, Daniel and Yonghong I just submitted a new series ("bpf: fix semantics issues with helpers receiving NULL arguments") that includes a fix in
bpf_perf_event_output. This should simplify the way you write your bpf programs, so you shouldn't be required to write those convoluted branches anymore (there are a few usage examples in the commit log).Quoted 3 lines
In my case it made writing the code much easier, after applying it I haven't been surprised by the compiler output in a while, and I hope your experience will be improved as well.
Trying to work with this again, and I still need to trick clang into not doing some optimizations that end up getting the resulting eBPF object rejected by the kernel verifier:
[root@jouet bpf]# uname -a
Linux jouet 4.15.0-rc8+ #1 SMP Wed Jan 17 11:01:34 -03 2018x86_64x86_64x86_64GNU/Linux [root@jouet bpf]# grep -i bpf/lib/modules/uname -r/build/.configCONFIG_CGROUP_BPF=yCONFIG_BPF=yCONFIG_BPF_SYSCALL=yCONFIG_BPF_JIT_ALWAYS_ON=y
#CONFIG_NETFILTER_XT_MATCH_BPFis not set
#CONFIG_NET_CLS_BPFis not set
#CONFIG_NET_ACT_BPFis not setCONFIG_BPF_JIT=yCONFIG_BPF_STREAM_PARSER=yCONFIG_LWTUNNEL_BPF=yCONFIG_HAVE_EBPF_JIT=yCONFIG_BPF_EVENTS=y
#CONFIG_TEST_BPFis not set
[root@jouet bpf]# catsys_enter_open.c
#include "bpf.h"
SEC("syscalls:sys_enter_open")
intfunc(void *ctx){ struct { char *ptr; char path[256]; } filename = { /* * /sys/kernel/debug/tracing/events/syscalls/sys_enter_open/format: * * ... * field:const char * filename; offset:16; size:8; signed:0; * ... * ctx + 16 selects 'filename' */ .ptr = *((char **)(ctx + 16)), }; int len = bpf_probe_read_str(filename.path, sizeof(filename.path), filename.ptr); if (len > 0 && len < 256) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr)); return 0; }[root@jouet bpf]#
[root@jouet bpf]# perf trace -v -e open,sys_enter_open.c
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to/lib/modules/4.15.0-rc8+/build
set env: KBUILD_DIR=/lib/modules/4.15.0-rc8+/build
unset env:KBUILD_OPTS
include option is set to -nostdinc -isystem/usr/lib/gcc/x86_64-redhat-linux/7/include-I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include/home/acme/git/linux/include/linux/kconfig.hset env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40f00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env:CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem/usr/lib/gcc/x86_64-redhat-linux/7/include-I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include/home/acme/git/linux/include/linux/kconfig.hset env: WORKING_DIR=/lib/modules/4.15.0-rc8+/build
set env: CLANG_SOURCE=/home/acme/bpf/sys_enter_open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'sys_enter_open.c' from buffer
libbpf: section .strtab, size 101, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section syscalls:sys_enter_open, size 472, link 0, flags 6, type=1
libbpf: found program syscalls:sys_enter_open
libbpf: section .relsyscalls:sys_enter_open, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license ofsys_enter_open.cis GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version ofsys_enter_open.cis 40f00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps insys_enter_open.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'syscalls:sys_enter_open'
libbpf: relocation: insn_idx=51
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 51
LLVM: dumping sys_enter_open.o
bpf: config program 'syscalls:sys_enter_open'
libbpf: create map__bpf_stdout__: fd=3
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (bf) r6 = r1
1: (b7) r1 = 0
2: (7b) *(u64 *)(r10 -8) = r1
3: (7b) *(u64 *)(r10 -16) = r1
4: (7b) *(u64 *)(r10 -24) = r1
5: (7b) *(u64 *)(r10 -32) = r1
6: (7b) *(u64 *)(r10 -40) = r1
7: (7b) *(u64 *)(r10 -48) = r1
8: (7b) *(u64 *)(r10 -56) = r1
9: (7b) *(u64 *)(r10 -64) = r1
10: (7b) *(u64 *)(r10 -72) = r1
11: (7b) *(u64 *)(r10 -80) = r1
12: (7b) *(u64 *)(r10 -88) = r1
13: (7b) *(u64 *)(r10 -96) = r1
14: (7b) *(u64 *)(r10 -104) = r1
15: (7b) *(u64 *)(r10 -112) = r1
16: (7b) *(u64 *)(r10 -120) = r1
17: (7b) *(u64 *)(r10 -128) = r1
18: (7b) *(u64 *)(r10 -136) = r1
19: (7b) *(u64 *)(r10 -144) = r1
20: (7b) *(u64 *)(r10 -152) = r1
21: (7b) *(u64 *)(r10 -160) = r1
22: (7b) *(u64 *)(r10 -168) = r1
23: (7b) *(u64 *)(r10 -176) = r1
24: (7b) *(u64 *)(r10 -184) = r1
25: (7b) *(u64 *)(r10 -192) = r1
26: (7b) *(u64 *)(r10 -200) = r1
27: (7b) *(u64 *)(r10 -208) = r1
28: (7b) *(u64 *)(r10 -216) = r1
29: (7b) *(u64 *)(r10 -224) = r1
30: (7b) *(u64 *)(r10 -232) = r1
31: (7b) *(u64 *)(r10 -240) = r1
32: (7b) *(u64 *)(r10 -248) = r1
33: (7b) *(u64 *)(r10 -256) = r1
34: (79) r3 = *(u64 *)(r6 +16)
35: (7b) *(u64 *)(r10 -264) = r3
36: (bf) r1 = r10
37: (07) r1 += -256
38: (b7) r2 = 256
39: (85) call bpf_probe_read_str#45
40: (bf) r1 = r0
41: (07) r1 += -1
42: (67) r1 <<= 32
43: (77) r1 >>= 32
44: (25) if r1 >0xfegoto pc+12
R0=inv(id=0) R1=inv(id=0,umax_value=254,var_off=(0x0;0xff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
45: (67) r0 <<= 32
46: (c7) r0 s>>= 32
47: (07) r0 += 8
Quoted 8 lines
48: (bf) r4 = r10
49: (07) r4 += -264
50: (bf) r1 = r6
51: (18) r2 =0xffff957fc90e1000
53: (18) r3 =0xffffffff
55: (bf) r5 = r0
56: (85) call bpf_perf_event_output#25
R5 min value is negative, either use unsigned or 'var &= const'
Yes, I can reproduce the issue as well. The reason is again due to compiler optimization,
Quoted 7 lines
39: (85) call bpf_probe_read_str#45
40: (bf) r1 = r0
41: (07) r1 += -1
42: (67) r1 <<= 32
43: (77) r1 >>= 32
44: (25) if r1 >0xfegoto pc+12
R0=inv(id=0) R1=inv(id=0,umax_value=254,var_off=(0x0;0xff))
R6=ctx(id=0,off=0,imm=0) R10=fp0
Quoted 11 lines
45: (67) r0 <<= 32
46: (c7) r0 s>>= 32
47: (07) r0 += 8
48: (bf) r4 = r10
49: (07) r4 += -264
50: (bf) r1 = r6
51: (18) r2 =0xffff957fc90e1000
53: (18) r3 =0xffffffff
55: (bf) r5 = r0
56: (85) call bpf_perf_event_output#25
R5 min value is negative, either use unsigned or 'var &= const'
The compiler did "40: (bf) r1 = r0" and then uses "r1" for branch comparison, the original "r0" is left with complete unknown integer value and later used to calculate the buffer size "55: (bf) r5 = r0" where "r5" could be negative value and the verifier rightfully
complains.
There is no easy way to fix this in verifier unless verifier starts to track correlations between registers which is a big task. So your below workaround is okay. The below workaround should also work:
int len = bpf_probe_read_str(filename.path,
sizeof(filename.path), filename.ptr);
if (len > 0 && len < 256)
bpf_perf_event_output(ctx, &my_map, BPF_F_CURRENT_CPU,
&filename, (len & 0xff) + sizeof(filename.ptr));
return 0;Thanks,
Quoted 81 lines
libbpf: -- END LOG --
libbpf: failed to load program 'syscalls:sys_enter_open'
libbpf: failed to load object 'sys_enter_open.c'
bpf: load objects failed
event syntax error: 'sys_enter_open.c'
\___ Kernel verifier blocks program loading(add -v to see detail)
Run 'perf list' for a list of valid eventsUsage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]-e, --event <event> event/syscall selector. use 'perf list' to list available events
[root@jouet bpf]#If I use the version we came up before, it works:
[root@jouet bpf]# trace -e open,sys_enter_open.c
LLVM: dumping sys_enter_open.o
2171.820 ( ): __bpf_stdout__:@......./proc/self/task/14683/comm..)
2171.845 ( 0.339 ms): qemu-system-x8/6721open(filename: /proc/self/task/14683/comm, flags: RDWR )= 91
^C[root@jouet bpf]
[root@jouet bpf]# trace -e open,sys_enter_open.o
2485.416 ( ): __bpf_stdout__:Vm..u.../proc/loadavg.......)
2485.434 ( 0.210 ms): lighttpd/25120open(filename: /proc/loadavg, mode: ISGID|IXOTH )= 8
^C[root@jouet bpf]#[root@jouet bpf]cat
sys_enter_open.c
#include "bpf.h"
SEC("syscalls:sys_enter_open")
intfunc(void *ctx){ struct { char *ptr; char path[256]; } filename = { /* * /sys/kernel/debug/tracing/events/syscalls/sys_enter_open/format: * * ... * field:const char * filename; offset:16; size:8; signed:0; * ... * ctx + 16 selects 'filename' */ .ptr = *((char **)(ctx + 16)), }; int len = bpf_probe_read_str(filename.path, sizeof(filename.path), filename.ptr); if (len > 0) { if (len == 1) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr)); else if (len < 256) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr)); } return 0; }[root@jouet bpf]#
[acme@jouet perf]$ llc --version | head -16
LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIBAg&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=kbmGig2gAUdNIT0zFKDollApvFW0R95_U5Ntz4dvRbM&s=zQgAZ9vJnfkoNy-c53Y3H_2nMVwMzV-LyMA2f6bSFrw&e=):
LLVM version 6.0.0svn
DEBUG build with assertions.
Default target: x86_64-unknown-linux-gnu
Host CPU: broadwellRegistered Targets:
aarch64 - AArch64 (little endian)
aarch64_be- AArch64 (big endian)
amdgcn - AMD GCN GPUs
arm - ARM
arm64 - ARM64 (little endian)
armeb - ARM (big endian)
bpf - BPF (host endian)
bpfeb - BPF (big endian)
bpfel - BPF (little endian)
[acme@jouet perf]$
Em Mon, Jan 22, 2018 at 10:28:11AM -0800, Yonghong Song escreveu:
Quoted 10 lines
The compiler did "40: (bf) r1 = r0" and then uses "r1" for branch comparison, the original "r0" is left with complete unknown integer value and later used to calculate the buffer size "55: (bf) r5 = r0" where "r5" could be negative value and the verifier rightfully complains.
There is no easy way to fix this in verifier unless verifier starts to track correlations between registers which is a big task. So your below workaround is okay. The below workaround should also work:
Quoted 6 lines
int len = bpf_probe_read_str(filename.path, sizeof(filename.path),
filename.ptr);
if (len > 0 && len < 256)
bpf_perf_event_output(ctx,&my_map,BPF_F_CURRENT_CPU,&filename, (len & 0xff) + sizeof(filename.ptr)); return 0;
Ok, thanks for one more time doing the analysis of the optimizations emitted and suggesting something more compact, that I can confirm works:
[root@jouet bpf]# perf trace -a -e open,sys_enter_open.c sleep 0.1
LLVM: dumping sys_enter_open.o
1.212 ( ): __bpf_stdout__:......../usr/lib/locale/locale-archive......)
1.218 ( 0.021 ms): sleep/9872 open(filename: /usr/lib/locale/locale-archive, flags: CLOEXEC) = 3
2.905 ( ): __bpf_stdout__:..:.F.../usr/lib/locale/locale-archive......)
2.910 ( 0.013 ms): rm/9873 open(filename: /usr/lib/locale/locale-archive, flags: CLOEXEC) = 3
7.562 ( ): __bpf_stdout__:..ul..../usr/lib/locale/locale-archive......)
7.564 ( 0.013 ms): mv/9874 open(filename: /usr/lib/locale/locale-archive, flags: CLOEXEC) = 3
11.275 ( ): __bpf_stdout__:...d..../usr/lib/locale/locale-archive......)
11.278 ( 0.012 ms): sh/9875 open(filename: /usr/lib/locale/locale-archive, flags: CLOEXEC) = 3
11.945 ( ): __bpf_stdout__:...d..../usr/lib64/gconv/gconv-modules.cache........)
11.953 ( 0.018 ms): sh/9875 open(filename: /usr/lib64/gconv/gconv-modules.cache) = 3
17.906 ( ): __bpf_stdout__:..T.p.../usr/lib/locale/locale-archive......)
17.913 ( 0.319 ms): gcc/9877 open(filename: /usr/lib/locale/locale-archive, flags: CLOEXEC) = 4
18.389 ( ): __bpf_stdout__:...l..../usr/share/locale/locale.alias......)
18.394 ( 0.266 ms): gcc/9877 open(filename: /usr/share/locale/locale.alias, flags: CLOEXEC) = 4
18.777 ( ): __bpf_stdout__:@......./usr/share/locale/en_US.UTF-8/LC_MESSAGES/gcc.mo....)
18.782 ( 0.318 ms): gcc/9877 open(filename: /usr/share/locale/en_US.UTF-8/LC_MESSAGES/gcc.mo, mode: IFBLK|IFIFO|ISGID|ISVTX|IRUSR|IXUSR|0xb5cc0000) = -1 ENOENT No such file or directory[root@jouet bpf]# cat sys_enter_open.c
#include "bpf.h"
SEC("syscalls:sys_enter_open")
int func(void *ctx)
{
struct {
char *ptr;
char path[256];
} filename = {
.ptr = *((char **)(ctx + 16)),
};
int len = bpf_probe_read_str(filename.path, sizeof(filename.path), filename.ptr);
if (len > 0 && len < 256)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, (len & 0xff) + sizeof(filename.ptr));
return 0;
}[root@jouet bpf]#
- Arnaldo
Em Tue, Nov 14, 2017 at 09:25:17PM +0100, Daniel Borkmann escreveu:
Quoted 129 lines · Yonghong Song
Quoted 128 lines · Daniel Borkmann
Quoted 127 lines · Arnaldo Carvalho de Melo
Quoted 103 lines
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...])
{
char filename[128];
int ret =bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0)
bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1;
}r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1
1: bf a1 00 00 00 00 00 00 r1 = r10
2: 07 01 00 00 80 ff ff ff r1 += -128
3: b7 02 00 00 80 00 00 00 r2 = 128
4: 85 00 00 00 2d 00 00 00 call 45
5: 67 00 00 00 20 00 00 00 r0 <<= 32
6: c7 00 00 00 20 00 00 00 r0 s>>= 32
7: b7 01 00 00 01 00 00 00 r1 = 1
8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10
9: 57 00 00 00 7f 00 00 00 r0 &= 127
10: bf a4 00 00 00 00 00 00 r4 = r10
11: 07 04 00 00 80 ff ff ff r4 += -128
12: bf 61 00 00 00 00 00 00 r1 = r6
13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0
18: 85 00 00 00 19 00 00 00 call 25Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0 11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840u
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len =bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylakeQuoted 6 lines
[root@jouet bpf]# llc --version
LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0 11: (57) r0 &= 127
[...]
Quoted 1 line
Quoted 1 line
Quoted 1 line
... the shifts on r1 might be due to using 32 bit type, so if you find
Where is it using a 32 bit type?
Here is it again, now using clang 6:
[root@jouet bpf]# trace -v -e *open,open.c usleep 2
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to /lib/modules/4.14.0+/build
set env: KBUILD_DIR=/lib/modules/4.14.0+/build
unset env: KBUILD_OPTS
include option is set to -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40e00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env: CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/7/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h set env: WORKING_DIR=/lib/modules/4.14.0+/build
set env: CLANG_SOURCE=/home/acme/bpf/open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'open.c' from buffer
libbpf: section .strtab, size 103, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section prog=do_sys_open filename, size 168, link 0, flags 6, type=1
libbpf: found program prog=do_sys_open filename
libbpf: section .relprog=do_sys_open filename, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license of open.c is GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version of open.c is 40e00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps in open.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'prog=do_sys_open filename'
libbpf: relocation: insn_idx=13
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 13
bpf: config program 'prog=do_sys_open filename'
symbol:do_sys_open file:(null) line:0 offset:0 return:0 lazy:(null)
parsing arg: filename into filename
bpf: config 'prog=do_sys_open filename' is ok
Looking at the vmlinux_path (8 entries long)
Using /lib/modules/4.14.0+/build/vmlinux for symbols
Open Debuginfo file: /lib/modules/4.14.0+/build/vmlinux
Try to find probe point from debuginfo.
Matched function: do_sys_open [2a2e5f8]
Probe point found: do_sys_open+0
Searching 'filename' variable in context.
Converting variable filename into trace event.
filename type is (null).
Opening /sys/kernel/debug/tracing//README write=0
Found 1 probe_trace_events.
Opening /sys/kernel/debug/tracing//kprobe_events write=1
Writing event: p:perf_bpf_probe/prog _text+2493856 filename=%si:x64
In map_prologue, ntevs=1
mapping[0]=0
libbpf: create map __bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 == 0x0 goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 = 0xffff9b56ca5a2a80
17: (18) r3 = 0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0
libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loading
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
Opening /sys/kernel/debug/tracing//kprobe_events write=1
Opening /sys/kernel/debug/tracing//uprobe_events write=1
Parsing probe_events: p:perf_bpf_probe/prog _text+2493856 filename=%si:x64
Group:perf_bpf_probe Event:prog probe:p
Writing event: -:perf_bpf_probe/prog
[root@jouet bpf]# /usr/local/bin/clang --version
clang version 6.0.0 (http://llvm.org/git/clang.git 56cc8f8880db2ebc433eeb6b6a707c101467a186) (http://llvm.org/git/llvm.git 3656d83960a4f3fedf6d8f19043abf52379f78c3)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir: /usr/local/bin
[root@jouet bpf]# cat open.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
int prog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
u64 len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr);
if (len > 0)
perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename,
len & (sizeof(filename) - 1));
return 1;
}[root@jouet bpf]#
Quoted 45 lines
Quoted 9 lines
Quoted 4 lines
a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
It is tricky to do in the
bpf_program. Compiler tries hard to optimize :-).The issue is at "r0 &= 127".
Quoted 8 lines
9: (6d) if r1 s> r0 goto pc+10
R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: (57) r0 &= 127
11: R0=inv(id=0,umax_value=127,var_off=(0x0;0x7f)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0One possible solution for this problem is to relax the arg4 type from
ARG_CONST_SIZEtoARG_CONST_SIZE_OR_ZERO.Yeah, I know, that's what I mentioned earlier in this thread to resolve it, but do we really want to add this hack everywhere? :( Potentially any function having
ARG_CONST_SIZEwould need to handle size 0 and bail out again in their helper implementation and it ends up that progs start relying on this runtime check where we won't be able to get rid of it later on anymore.Quoted 21 lines
kernel/trace/bpf_trace.c +4
-1@@ -393,6 +393,9 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, }, }; + if (unlikely(size == 0)) + return 0; + if (unlikely(flags & ~(BPF_F_INDEX_MASK))) return -EINVAL;@@ -407,7 +410,7 @@ static const struct bpf_func_proto bpf_perf_event_output_proto = { .arg2_type = ARG_CONST_MAP_PTR, .arg3_type = ARG_ANYTHING, .arg4_type = ARG_PTR_TO_MEM, - .arg5_type = ARG_CONST_SIZE, + .arg5_type = ARG_CONST_SIZE_OR_ZERO, };
Quoted 194 lines · Arnaldo Carvalho de Melo
Em Tue, Nov 14, 2017 at 09:25:17PM +0100, Daniel Borkmann escreveu:
Quoted 129 lines · Yonghong Song
Quoted 128 lines · Daniel Borkmann
Quoted 127 lines · Arnaldo Carvalho de Melo
Quoted 103 lines
Em Tue, Nov 14, 2017 at 02:09:34PM +0100, Daniel Borkmann escreveu:
Quoted 82 lines · Arnaldo Carvalho de Melo
Quoted 81 lines
Em Tue, Nov 14, 2017 at 01:09:39AM +0100, Daniel Borkmann escreveu:
Quoted 53 lines · Arnaldo Carvalho de Melo
Quoted 6 lines
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
2: (bf) r6 = r1
3: (bf) r1 = r10Quoted 8 lines
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (07) r1 += -1
9: (67) r1 <<= 32
10: (77) r1 >>= 32
11: (25) if r1 >0x7fgoto pc+11Right, so the compiler is optimizing the two tests into a single one above, which means lower bound cannot properly be derived again by the verifier due to this and thus you'll get the error. Similar issue was seen recently [1].
Does the below hack work for you?
int
prog([...])
{
char filename[128];
int ret =bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (ret > 0)
bpf_perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, ret & (sizeof(filename) - 1)); return 1;
}r0 should keep on tracking bounds here at least:
prog:
0: bf 16 00 00 00 00 00 00 r6 = r1
1: bf a1 00 00 00 00 00 00 r1 = r10
2: 07 01 00 00 80 ff ff ff r1 += -128
3: b7 02 00 00 80 00 00 00 r2 = 128
4: 85 00 00 00 2d 00 00 00 call 45
5: 67 00 00 00 20 00 00 00 r0 <<= 32
6: c7 00 00 00 20 00 00 00 r0 s>>= 32
7: b7 01 00 00 01 00 00 00 r1 = 1
8: 6d 01 0a 00 00 00 00 00 if r1 s> r0 goto 10
9: 57 00 00 00 7f 00 00 00 r0 &= 127
10: bf a4 00 00 00 00 00 00 r4 = r10
11: 07 04 00 00 80 ff ff ff r4 += -128
12: bf 61 00 00 00 00 00 00 r1 = r6
13: 18 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 r2 = 0ll 15: 18 03 00 00 ff ff ff ff 00 00 00 00 00 00 00 00 r3 = 4294967295ll 17: bf 05 00 00 00 00 00 00 r5 = r0
18: 85 00 00 00 19 00 00 00 call 25Not yet:
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff92bfc2aba840u
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0I'll try updating clang/llvm...
Full details:
[root@jouet bpf]# cat
open.c
#include "bpf.h"Quoted 8 lines
Quoted 5 lines
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr)
{
char filename[128];
const unsigned len =bpf_probe_read_str(filename, sizeof(filename), filename_ptr);Btw, I was using 'int' here above instead of 'unsigned' as
strncpy_from_unsafe()could potentially return errors like-EFAULT.I changed to int, didn't help
Quoted 8 lines
Currently having a version compiled from the git tree:
# llc --version LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=): LLVM version 6.0.0git-2d810c2 Optimized build. Default target: x86_64-unknown-linux-gnu Host CPU: skylakeQuoted 6 lines
[root@jouet bpf]# llc --version
LLVM (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_&d=DwIDaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=Qp3xFfXEz-CT8rzYtrHeXbow2M6FlsUzwcY32i3_2Q0&s=BKC_Gu9s1hw0v13OCgCpfsGtAY2hE7dujFqg8LNaK2I&e=):
LLVM version 4.0.0svnOld stuff! ;-) Will change, but improving these messages should be on the radar, I think :-)
Yep, agree, I think we need a generic, better solution for this type of issue instead of converting individual helpers to handle 0 min bound and then only bailing out in such case; need to brainstorm a bit on that.
I think for the above in your case ...
[...]
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
10: (15) if r1 ==0x0goto pc+10
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0 11: (57) r0 &= 127
[...]Quoted 1 line
Quoted 1 line
Quoted 1 line
... the shifts on r1 might be due to using 32 bit type, so if you find
Where is it using a 32 bit type?
Here is it again, now using clang 6:
[root@jouet bpf]# trace -v -e *open,open.c usleep 2
bpf: builtin compilation failed: -95, try external compiler
Kernel build dir is set to/lib/modules/4.14.0+/build
set env: KBUILD_DIR=/lib/modules/4.14.0+/build
unset env:KBUILD_OPTS
include option is set to -nostdinc -isystem/usr/lib/gcc/x86_64-redhat-linux/7/include-I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include/home/acme/git/linux/include/linux/kconfig.hset env: NR_CPUS=4
set env: LINUX_VERSION_CODE=0x40e00
set env: CLANG_EXEC=/usr/local/bin/clang
unset env:CLANG_OPTIONS
set env: KERNEL_INC_OPTIONS= -nostdinc -isystem/usr/lib/gcc/x86_64-redhat-linux/7/include-I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include/home/acme/git/linux/include/linux/kconfig.hset env: WORKING_DIR=/lib/modules/4.14.0+/build
set env: CLANG_SOURCE=/home/acme/bpf/open.c
llvm compiling command template: $CLANG_EXEC -D__KERNEL__ -D__NR_CPUS__=$NR_CPUS -DLINUX_VERSION_CODE=$LINUX_VERSION_CODE $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o -
libbpf: loading object 'open.c' from buffer
libbpf: section .strtab, size 103, link 0, flags 0, type=3
libbpf: section .text, size 0, link 0, flags 6, type=1
libbpf: section prog=do_sys_open filename, size 168, link 0, flags 6, type=1
libbpf: found program prog=do_sys_open filename
libbpf: section .relprog=do_sys_open filename, size 16, link 8, flags 0, type=9
libbpf: section maps, size 16, link 0, flags 3, type=1
libbpf: section license, size 4, link 0, flags 3, type=1
libbpf: license ofopen.cis GPL
libbpf: section version, size 4, link 0, flags 3, type=1
libbpf: kernel version ofopen.cis 40e00
libbpf: section .symtab, size 144, link 1, flags 0, type=2
libbpf: maps inopen.c: 1 maps in 16 bytes
libbpf: map 0 is "__bpf_stdout__"
libbpf: collecting relocating info for: 'prog=do_sys_open filename'
libbpf: relocation: insn_idx=13
libbpf: relocation: find map 0 (__bpf_stdout__) for insn 13
bpf: config program 'prog=do_sys_open filename'
symbol:do_sys_open file:(null) line:0 offset:0 return:0 lazy:(null)
parsing arg: filename into filename
bpf: config 'prog=do_sys_open filename' is ok
Looking at thevmlinux_path(8 entries long)
Using/lib/modules/4.14.0+/build/vmlinuxfor symbols
Open Debuginfo file:/lib/modules/4.14.0+/build/vmlinux
Try to find probe point from debuginfo.
Matched function:do_sys_open[2a2e5f8]
Probe point found: do_sys_open+0
Searching 'filename' variable in context.
Converting variable filename into trace event.
filename type is (null).
Opening/sys/kernel/debug/tracing//READMEwrite=0
Found 1probe_trace_events.
Opening/sys/kernel/debug/tracing//kprobe_eventswrite=1
Writing event: p:perf_bpf_probe/prog _text+2493856 filename=%si:x64
Inmap_prologue, ntevs=1
mapping[0]=0
libbpf: create map__bpf_stdout__: fd=3
prologue: pass validation
prologue: fast path
libbpf: load bpf program failed: Permission denied
libbpf: -- BEGIN DUMP LOG ---
libbpf:
0: (79) r3 = *(u64 *)(r1 +104)
1: (b7) r2 = 0
Quoted 8 lines
2: (bf) r6 = r1
3: (bf) r1 = r10
4: (07) r1 += -128
5: (b7) r2 = 128
6: (85) call bpf_probe_read_str#45
7: (bf) r1 = r0
8: (67) r1 <<= 32
9: (77) r1 >>= 32
This shift is due to the return type of bpf_probe_read_str which is int. The compiler does a sign extension to get it to u64 type.
Quoted 1 line
10: (15) if r1 ==
0x0goto pc+10
Since it is u64 type, the compiler uses "==" instead of ">".
Quoted 8 lines
R0=inv(id=0) R1=inv(id=0,umax_value=4294967295,var_off=(0x0;
0xffffffff)) R6=ctx(id=0,off=0,imm=0) R10=fp0
11: (57) r0 &= 127
12: (bf) r4 = r10
13: (07) r4 += -128
14: (bf) r1 = r6
15: (18) r2 =0xffff9b56ca5a2a80
17: (18) r3 =0xffffffff
19: (bf) r5 = r0
r5 = r0, and r0 is still possbily 0.
r0 = bpf_probe_read_str
...
r0 &= 127
...
The same symptom as "int len" type.
Quoted 90 lines
20: (85) call bpf_perf_event_output#25
invalid stack type R4 off=-128 access_size=0libbpf: -- END LOG --
libbpf: Loading the 0th instance of program 'prog=do_sys_open filename' failed
libbpf: failed to load program 'prog=do_sys_open filename'
libbpf: failed to load object 'open.c'
bpf: load objects failed
event syntax error: 'open.c'
\___ Kernel verifier blocks program loading(add -v to see detail)
Run 'perf list' for a list of valid eventsUsage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]-e, --event <event> event/syscall selector. use 'perf list' to list available events
Opening/sys/kernel/debug/tracing//kprobe_eventswrite=1
Opening/sys/kernel/debug/tracing//uprobe_eventswrite=1
Parsingprobe_events: p:perf_bpf_probe/prog _text+2493856 filename=%si:x64
Group:perf_bpf_probe Event:prog probe:p
Writing event: -:perf_bpf_probe/prog
[root@jouet bpf]#/usr/local/bin/clang--version
clang version 6.0.0 (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_git_clang.git&d=DwIDAw&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=lALOhwxSDXp7fA6co6UyLBnLfqv0e2pf97GCuu5hGSw&s=FGoodXMmAdx4dqZJtvq6Vosu03to1-GFIRbagBiz0fg&e=56cc8f8880db2ebc433eeb6b6a707c101467a186) (https://urldefense.proofpoint.com/v2/url?u=http-3A__llvm.org_git_llvm.git&d=DwIDAw&c=5VD0RTtNlTh3ycd41b3MUw&r=DA8e1B5r073vIqRrFz7MRA&m=lALOhwxSDXp7fA6co6UyLBnLfqv0e2pf97GCuu5hGSw&s=_pQdnamFeu6n_um7WGrLwaCMtXqIVGLLEmYbcss_0aU&e=3656d83960a4f3fedf6d8f19043abf52379f78c3)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir:/usr/local/bin
[root@jouet bpf]# catopen.c
#include "bpf.h"
SEC("prog=do_sys_open filename")
intprog(void *ctx, int err, const char __user *filename_ptr){ char filename[128]; u64 len = bpf_probe_read_str(filename, sizeof(filename), filename_ptr); if (len > 0) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, filename, len & (sizeof(filename) - 1)); return 1; }[root@jouet bpf]#
Quoted 45 lines
Quoted 9 lines
Quoted 4 lines
a way to avoid these and have the test on r0 directly, we might get there. Perhaps keep using a 64 bit type to avoid them. It would be useful to propagate the deduced bound information back to r0 when we know that neither r0 nor r1 has changed in the meantime.
It is tricky to do in the
bpf_program. Compiler tries hard to optimize :-).The issue is at "r0 &= 127".
Quoted 8 lines
9: (6d) if r1 s> r0 goto pc+10
R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0
10: R0=inv(id=0,umin_value=1,umax_value=9223372036854775807,var_off=(0x0;0x7fffffffffffffff)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0 10: (57) r0 &= 127
11: R0=inv(id=0,umax_value=127,var_off=(0x0;0x7f)) R1=inv1 R6=ctx(id=0,off=0,imm=0) R10=fp0One possible solution for this problem is to relax the arg4 type from
ARG_CONST_SIZEtoARG_CONST_SIZE_OR_ZERO.Yeah, I know, that's what I mentioned earlier in this thread to resolve it, but do we really want to add this hack everywhere? :( Potentially any function having
ARG_CONST_SIZEwould need to handle size 0 and bail out again in their helper implementation and it ends up that progs start relying on this runtime check where we won't be able to get rid of it later on anymore.Quoted 21 lines
kernel/trace/bpf_trace.c +4
-1@@ -393,6 +393,9 @@ BPF_CALL_5(bpf_perf_event_output, struct pt_regs *, regs, struct bpf_map *, map, }, }; + if (unlikely(size == 0)) + return 0; + if (unlikely(flags & ~(BPF_F_INDEX_MASK))) return -EINVAL;@@ -407,7 +410,7 @@ static const struct bpf_func_proto bpf_perf_event_output_proto = { .arg2_type = ARG_CONST_MAP_PTR, .arg3_type = ARG_ANYTHING, .arg4_type = ARG_PTR_TO_MEM, - .arg5_type = ARG_CONST_SIZE, + .arg5_type = ARG_CONST_SIZE_OR_ZERO, };