Full reproduction instructions:
1) Install Fedora 26 server. I chose all of the default options,
except I chose to install latest packages from the network. Otherwise,
just run a dnf update after install. I installed into a VM running on
HyperV
2) Install clang, llvm, and kernel-devel:
dnf install -y clang llvm kernel-devel
3) Copy the program to xdp_test_kern.c
4) Compile the program into an elf object file:
clang -S -O2 -Wall -Werror -nostdinc -isystem $(clang
-print-file-name=include) -I /lib/modules/$(uname
-r)/build/arch/x86/include -I /lib/modules/$(uname
-r)/build/arch/x86/include/generated/uapi -I /lib/modules/$(uname
-r)/build/arch/x86/include/generated -I /lib/modules/$(uname
-r)/build/include -I /lib/modules/$(uname
-r)/build/arch/x86/include/uapi -I /lib/modules/$(uname
-r)/build/include/uapi -I /lib/modules/$(uname
-r)/build/include/generated/uapi -include /lib/modules/$(uname
-r)/build/include/linux/kconfig.h -D __KERNEL__ -c -emit-llvm -o
xdp_test_kern.ll xdp_test_kern.c
llc -march bpf -filetype obj -o xdp_test_kern.o xdp_test_kern.ll
5) Attach the program to eth0:
ip link set dev eth0 xdp object xdp_test_kern.o section .text verbose
6) Notice only 0s being printed for data in /sys/kernel/debug/tracing/trace_pipe
As near as I can tell, the driver being used for eth0 is hv_netvsc, in
case it's an issue with the driver.
Thanks!
--Zvi
On Mon, Aug 28, 2017 at 2:16 PM, Y Song <ys114321@xxxxxxxxx> wrote:
> bpf program itself is okay.
> I replaced linux:samples/bpf/xdp1_kern.c with this program and the
> bpf_trace_printk works fine.
> It could be some setup issue. But in any case, data start and data end
> are suspicious.
> It would be good you list all steps which can reproduce the issue.
>
> On Mon, Aug 28, 2017 at 10:20 AM, Zvi Effron <zeffron@xxxxxxxxxxxxx> wrote:
>> Hello,
>>
>> I'm having an issue where bpf_trace_printk seems to always print 0 for
>> values. I'm running the following xdp program:
>>
>>
>> #include <uapi/linux/bpf.h>
>>
>> static int (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) =
>> (void *) BPF_FUNC_trace_printk;
>>
>> int xdp(struct xdp_md *ctx) {
>> char format_string[] = "Data start: %x\tData end: %x\n";
>> bpf_trace_printk(format_string, sizeof(format_string), ctx->data,
>> ctx->data_end);
>>
>> {
>> char format_string[] = "Constant: %x\n";
>> bpf_trace_printk(format_string, sizeof(format_string), 1);
>> }
>>
>> if (ctx->data == 0) {
>> char format_string[] = "Data starts at offset 0";
>> bpf_trace_printk(format_string, sizeof(format_string));
>> }
>> if (ctx->data_end == 0) {
>> char format_string[] = "Data ends at offset 0";
>> bpf_trace_printk(format_string, sizeof(format_string));
>> }
>> return XDP_PASS;
>> }
>>
>> char __attribute__((section("license"), used)) license[] = "GPL";
>>
>>
>> and I get the following output in /sys/kernel/debug/tracing/trace_pipe
>>
>>
>> sshd-4824 [000] ..s1 67372.673714: : Data start: 0 Data end: 0
>> sshd-4824 [000] ..s1 67372.673720: : Constant: 0
>> <idle>-0 [000] ..s. 67372.675062: : Data start: 0 Data end: 0
>> <idle>-0 [000] .Ns. 67372.675090: : Constant: 0
>> <idle>-0 [000] .Ns. 67372.675116: : Data start: 0 Data end: 0
>> <idle>-0 [000] .Ns. 67372.675117: : Constant: 0
>> <idle>-0 [000] .Ns. 67372.675129: : Data start: 0 Data end: 0
>> <idle>-0 [000] .Ns. 67372.675130: : Constant: 0
>> <idle>-0 [000] .Ns. 67372.675136: : Data start: 0 Data end: 0
>> <idle>-0 [000] .Ns. 67372.675137: : Constant: 0
>> <idle>-0 [000] .Ns. 67372.675142: : Data start: 0 Data end: 0
>> <idle>-0 [000] .Ns. 67372.675143: : Constant: 0
>> ip-6458 [000] ..s. 67372.676083: : Data start: 0 Data end: 0
>> ip-6458 [000] ..s. 67372.676084: : Constant: 0
>> ip-6458 [000] ..s. 67372.676093: : Data start: 0 Data end: 0
>> ip-6458 [000] ..s. 67372.676094: : Constant: 0
>>
>>
>> It looks like ctx->data and ctx->data_end are not 0, because the calls
>> to bpf_trace_printk in those ifs aren't being called, and the constant
>> is definitely not 0.
>>
>> What might I be missing or doing incorrectly?
>>
>> Thank you!
>> --Zvi
|