From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: Andrei Matei <andreimatei1@gmail.com>
Cc: bpf <bpf@vger.kernel.org>, Alexei Starovoitov <ast@kernel.org>
Subject: Re: [PATCH bpf-next v2 5/5] selftest/bpf: add test for var-offset stack access
Date: Mon, 25 Jan 2021 18:37:36 -0800 [thread overview]
Message-ID: <CAEf4Bzb3cw1zPfvdpZg2X+N5SS+H-NJJzrsDjW-9xEDFjRKA1g@mail.gmail.com> (raw)
In-Reply-To: <20210124194909.453844-6-andreimatei1@gmail.com>
On Sun, Jan 24, 2021 at 11:54 AM Andrei Matei <andreimatei1@gmail.com> wrote:
>
> Add a higher-level test (C BPF program) for the new functionality -
> variable access stack reads and writes.
>
> Signed-off-by: Andrei Matei <andreimatei1@gmail.com>
> ---
> .../selftests/bpf/prog_tests/stack_var_off.c | 56 +++++++++++++++++++
> .../selftests/bpf/progs/test_stack_var_off.c | 43 ++++++++++++++
> 2 files changed, 99 insertions(+)
> create mode 100644 tools/testing/selftests/bpf/prog_tests/stack_var_off.c
> create mode 100644 tools/testing/selftests/bpf/progs/test_stack_var_off.c
>
> diff --git a/tools/testing/selftests/bpf/prog_tests/stack_var_off.c b/tools/testing/selftests/bpf/prog_tests/stack_var_off.c
> new file mode 100644
> index 000000000000..c4c47fb0f0af
> --- /dev/null
> +++ b/tools/testing/selftests/bpf/prog_tests/stack_var_off.c
> @@ -0,0 +1,56 @@
> +// SPDX-License-Identifier: GPL-2.0
> +#include <test_progs.h>
> +#include "test_stack_var_off.skel.h"
> +
> +int dummy;
> +
> +noinline void uprobed_function(char *s, int len)
> +{
> + /* Do something to keep the compiler from removing the function.
> + */
> + dummy++;
> +}
> +
> +void test_stack_var_off(void)
> +{
> + int duration = 0;
> + struct bpf_link *uprobe_link;
> + struct test_stack_var_off *skel;
> + size_t uprobe_offset;
> + ssize_t base_addr;
> + char s[100];
> +
> + base_addr = get_base_addr();
> + if (CHECK(base_addr < 0, "get_base_addr",
> + "failed to find base addr: %zd", base_addr))
> + return;
> + uprobe_offset = (size_t)&uprobed_function - base_addr;
> +
> + skel = test_stack_var_off__open_and_load();
> + if (CHECK(!skel, "skel_open", "failed to open skeleton\n"))
> + return;
> + if (CHECK(!skel->bss, "check_bss", ".bss wasn't mmap()-ed\n"))
> + goto cleanup;
> +
> + uprobe_link = bpf_program__attach_uprobe(skel->progs.uprobe,
> + false /* retprobe */,
> + 0 /* self pid */,
> + "/proc/self/exe",
> + uprobe_offset);
> + if (CHECK(IS_ERR(uprobe_link), "attach_uprobe",
> + "err %ld\n", PTR_ERR(uprobe_link)))
> + goto cleanup;
> + skel->links.uprobe = uprobe_link;
> +
> + /* trigger uprobe */
> + s[0] = 1;
> + s[1] = 10;
> + uprobed_function(&s[0], 2);
I don't think uprobe() is essential to this test and just obscured
what is being tested. I'd just use a global variable to pass whatever
input data you need and use usleep(1), just like lots of other tests.
> +
> + if (CHECK(skel->bss->uprobe_res != 10, "check_uprobe_res",
> + "wrong uprobe res: %d\n", skel->bss->uprobe_res))
> + goto cleanup;
> +
> +cleanup:
> + test_stack_var_off__destroy(skel);
> +}
> diff --git a/tools/testing/selftests/bpf/progs/test_stack_var_off.c b/tools/testing/selftests/bpf/progs/test_stack_var_off.c
> new file mode 100644
> index 000000000000..44f982684541
> --- /dev/null
> +++ b/tools/testing/selftests/bpf/progs/test_stack_var_off.c
> @@ -0,0 +1,43 @@
> +// SPDX-License-Identifier: GPL-2.0
> +// Copyright (c) 2017 Facebook
> +
> +#include <linux/ptrace.h>
> +#include <linux/bpf.h>
> +#include <bpf/bpf_helpers.h>
> +#include <bpf/bpf_tracing.h>
> +
> +int uprobe_res;
> +
> +SEC("uprobe/func")
> +int BPF_KPROBE(uprobe, char *s, int len)
> +{
> + /* This BPF program performs variable-offset reads and writes on a
> + * stack-allocated buffer.
> + */
> + char buf[16];
> + unsigned long idx;
> + char out;
> +
> + /* Zero-out the buffer so we can read anywhere inside it. */
> + __builtin_memset(&buf, 0, 16);
> + /* Copy the contents of s from user-space. */
> + len &= 0xf;
> + if (bpf_probe_read_user(&buf, len, s)) {
> + bpf_printk("error reading user mem\n");
> + return 1;
> + }
> + /* Index into the buffer at an unknown offset that comes from the
> + * buffer itself. This is a variable-offset stack read.
> + */
> + idx = buf[0];
> + idx &= 0xf;
> + out = buf[idx];
> + /* Append something to the buffer. The position where we append it
> + * is unknown. This is a variable-offset stack write.
> + */
> + buf[len] = buf[idx];
> + uprobe_res = out;
> + return 0;
> +}
> +
> +char _license[] SEC("license") = "GPL";
> --
> 2.27.0
>
next prev parent reply other threads:[~2021-01-26 5:38 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-24 19:49 [PATCH bpf-next v2 0/5] bpf: allow variable-offset stack access Andrei Matei
2021-01-24 19:49 ` [PATCH bpf-next v2 1/5] " Andrei Matei
2021-01-27 22:58 ` Alexei Starovoitov
2021-01-30 22:55 ` Andrei Matei
2021-02-02 0:22 ` Alexei Starovoitov
2021-02-07 1:11 ` Andrei Matei
2021-01-24 19:49 ` [PATCH bpf-next v2 2/5] selftest/bpf: adjust expected verifier errors Andrei Matei
2021-01-24 19:49 ` [PATCH bpf-next v2 3/5] selftest/bpf: verifier tests for var-off access Andrei Matei
2021-01-24 19:49 ` [PATCH bpf-next v2 4/5] selftest/bpf: move utility function to tests header Andrei Matei
2021-01-26 2:33 ` Andrii Nakryiko
2021-01-24 19:49 ` [PATCH bpf-next v2 5/5] selftest/bpf: add test for var-offset stack access Andrei Matei
2021-01-26 2:37 ` Andrii Nakryiko [this message]
2021-02-07 1:11 ` Andrei Matei
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAEf4Bzb3cw1zPfvdpZg2X+N5SS+H-NJJzrsDjW-9xEDFjRKA1g@mail.gmail.com \
--to=andrii.nakryiko@gmail.com \
--cc=andreimatei1@gmail.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).