From: Yonghong Song <yhs@fb.com>
To: Brendan Jackman <jackmanb@google.com>, <bpf@vger.kernel.org>
Cc: Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
KP Singh <kpsingh@chromium.org>,
Florent Revest <revest@chromium.org>,
<linux-kernel@vger.kernel.org>, Jann Horn <jannh@google.com>
Subject: Re: [PATCH v2 bpf-next 11/13] bpf: Add bitwise atomic instructions
Date: Fri, 27 Nov 2020 21:39:10 -0800 [thread overview]
Message-ID: <d2e093c3-79bc-0a6b-8919-c5a07667926a@fb.com> (raw)
In-Reply-To: <20201127175738.1085417-12-jackmanb@google.com>
On 11/27/20 9:57 AM, Brendan Jackman wrote:
> This adds instructions for
>
> atomic[64]_[fetch_]and
> atomic[64]_[fetch_]or
> atomic[64]_[fetch_]xor
>
> All these operations are isomorphic enough to implement with the same
> verifier, interpreter, and x86 JIT code, hence being a single commit.
>
> The main interesting thing here is that x86 doesn't directly support
> the fetch_ version these operations, so we need to generate a CMPXCHG
> loop in the JIT. This requires the use of two temporary registers,
> IIUC it's safe to use BPF_REG_AX and x86's AUX_REG for this purpose.
similar to previous xsub (atomic[64]_sub), should we implement
xand, xor, xxor in llvm?
>
> Signed-off-by: Brendan Jackman <jackmanb@google.com>
> ---
> arch/x86/net/bpf_jit_comp.c | 49 ++++++++++++++++++++++++++++-
> include/linux/filter.h | 60 ++++++++++++++++++++++++++++++++++++
> kernel/bpf/core.c | 5 ++-
> kernel/bpf/disasm.c | 7 +++--
> kernel/bpf/verifier.c | 6 ++++
> tools/include/linux/filter.h | 60 ++++++++++++++++++++++++++++++++++++
> 6 files changed, 183 insertions(+), 4 deletions(-)
>
> diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
> index a8a9fab13fcf..46b977ee21c4 100644
> --- a/arch/x86/net/bpf_jit_comp.c
> +++ b/arch/x86/net/bpf_jit_comp.c
> @@ -823,8 +823,11 @@ static int emit_atomic(u8 **pprog, u8 atomic_op,
>
> /* emit opcode */
> switch (atomic_op) {
> - case BPF_SUB:
> case BPF_ADD:
> + case BPF_SUB:
> + case BPF_AND:
> + case BPF_OR:
> + case BPF_XOR:
> /* lock *(u32/u64*)(dst_reg + off) <op>= src_reg */
> EMIT1(simple_alu_opcodes[atomic_op]);
> break;
> @@ -1307,6 +1310,50 @@ st: if (is_imm8(insn->off))
>
> case BPF_STX | BPF_ATOMIC | BPF_W:
> case BPF_STX | BPF_ATOMIC | BPF_DW:
> + if (insn->imm == (BPF_AND | BPF_FETCH) ||
> + insn->imm == (BPF_OR | BPF_FETCH) ||
> + insn->imm == (BPF_XOR | BPF_FETCH)) {
> + u8 *branch_target;
> + bool is64 = BPF_SIZE(insn->code) == BPF_DW;
> +
> + /*
> + * Can't be implemented with a single x86 insn.
> + * Need to do a CMPXCHG loop.
> + */
> +
> + /* Will need RAX as a CMPXCHG operand so save R0 */
> + emit_mov_reg(&prog, true, BPF_REG_AX, BPF_REG_0);
> + branch_target = prog;
> + /* Load old value */
> + emit_ldx(&prog, BPF_SIZE(insn->code),
> + BPF_REG_0, dst_reg, insn->off);
> + /*
> + * Perform the (commutative) operation locally,
> + * put the result in the AUX_REG.
> + */
> + emit_mov_reg(&prog, is64, AUX_REG, BPF_REG_0);
> + maybe_emit_rex(&prog, AUX_REG, src_reg, is64);
> + EMIT2(simple_alu_opcodes[BPF_OP(insn->imm)],
> + add_2reg(0xC0, AUX_REG, src_reg));
> + /* Attempt to swap in new value */
> + err = emit_atomic(&prog, BPF_CMPXCHG,
> + dst_reg, AUX_REG, insn->off,
> + BPF_SIZE(insn->code));
> + if (WARN_ON(err))
> + return err;
> + /*
> + * ZF tells us whether we won the race. If it's
> + * cleared we need to try again.
> + */
> + EMIT2(X86_JNE, -(prog - branch_target) - 2);
> + /* Return the pre-modification value */
> + emit_mov_reg(&prog, is64, src_reg, BPF_REG_0);
> + /* Restore R0 after clobbering RAX */
> + emit_mov_reg(&prog, true, BPF_REG_0, BPF_REG_AX);
> + break;
> +
> + }
> +
> if (insn->imm == (BPF_SUB | BPF_FETCH)) {
> /*
> * x86 doesn't have an XSUB insn, so we negate
> diff --git a/include/linux/filter.h b/include/linux/filter.h
> index a20a3a536bf5..cb5d865cce3c 100644
> --- a/include/linux/filter.h
> +++ b/include/linux/filter.h
> @@ -300,6 +300,66 @@ static inline bool insn_is_zext(const struct bpf_insn *insn)
> .off = OFF, \
> .imm = BPF_SUB | BPF_FETCH })
>
> +/* Atomic memory and, *(uint *)(dst_reg + off16) -= src_reg */
> +
> +#define BPF_ATOMIC_AND(SIZE, DST, SRC, OFF) \
> + ((struct bpf_insn) { \
> + .code = BPF_STX | BPF_SIZE(SIZE) | BPF_ATOMIC, \
> + .dst_reg = DST, \
> + .src_reg = SRC, \
> + .off = OFF, \
> + .imm = BPF_AND })
> +
> +/* Atomic memory and with fetch, src_reg = atomic_fetch_and(*(dst_reg + off), src_reg); */
> +
> +#define BPF_ATOMIC_FETCH_AND(SIZE, DST, SRC, OFF) \
> + ((struct bpf_insn) { \
> + .code = BPF_STX | BPF_SIZE(SIZE) | BPF_ATOMIC, \
> + .dst_reg = DST, \
> + .src_reg = SRC, \
> + .off = OFF, \
> + .imm = BPF_AND | BPF_FETCH })
> +
[...]
next prev parent reply other threads:[~2020-11-28 21:51 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-27 17:57 [PATCH v2 bpf-next 00/13] Atomics for eBPF Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 01/13] bpf: x86: Factor out emission of ModR/M for *(reg + off) Brendan Jackman
2020-11-29 1:15 ` Alexei Starovoitov
2020-12-01 12:14 ` Brendan Jackman
2020-12-02 5:50 ` Alexei Starovoitov
2020-12-02 10:52 ` Brendan Jackman
2020-12-02 17:35 ` Alexei Starovoitov
2020-11-27 17:57 ` [PATCH v2 bpf-next 02/13] bpf: x86: Factor out emission of REX byte Brendan Jackman
2020-11-29 1:14 ` Alexei Starovoitov
2020-12-01 12:12 ` Brendan Jackman
2020-12-02 5:48 ` Alexei Starovoitov
2020-12-02 10:54 ` Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 03/13] bpf: x86: Factor out function to emit NEG Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 04/13] bpf: x86: Factor out a lookup table for some ALU opcodes Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 05/13] bpf: Rename BPF_XADD and prepare to encode other atomics in .imm Brendan Jackman
2020-11-28 3:43 ` Yonghong Song
2020-12-01 12:17 ` Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 06/13] bpf: Move BPF_STX reserved field check into BPF_STX verifier code Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 07/13] bpf: Add BPF_FETCH field / create atomic_fetch_add instruction Brendan Jackman
2020-11-28 4:15 ` Yonghong Song
2020-12-01 12:22 ` Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 08/13] bpf: Add instructions for atomic_[cmp]xchg Brendan Jackman
2020-11-28 5:25 ` Yonghong Song
2020-12-01 12:27 ` Brendan Jackman
2020-11-29 1:27 ` Alexei Starovoitov
2020-12-01 12:32 ` Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 09/13] bpf: Pull out a macro for interpreting atomic ALU operations Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 10/13] bpf: Add instructions for atomic[64]_[fetch_]sub Brendan Jackman
2020-11-27 21:39 ` kernel test robot
2020-11-27 21:39 ` [RFC PATCH] bpf: bpf_atomic_alu_string[] can be static kernel test robot
2020-11-28 5:35 ` [PATCH v2 bpf-next 10/13] bpf: Add instructions for atomic[64]_[fetch_]sub Yonghong Song
2020-11-29 1:34 ` Alexei Starovoitov
2020-11-30 17:18 ` Yonghong Song
2020-12-01 12:38 ` Brendan Jackman
2020-12-02 5:55 ` Alexei Starovoitov
2020-12-02 11:19 ` Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 11/13] bpf: Add bitwise atomic instructions Brendan Jackman
2020-11-28 5:39 ` Yonghong Song [this message]
2020-11-29 1:36 ` Alexei Starovoitov
2020-11-30 17:20 ` Yonghong Song
2020-11-27 17:57 ` [PATCH v2 bpf-next 12/13] bpf: Add tests for new BPF atomic operations Brendan Jackman
2020-12-01 3:55 ` Yonghong Song
2020-12-01 12:56 ` Brendan Jackman
2020-12-01 17:24 ` Yonghong Song
2020-12-02 2:22 ` Andrii Nakryiko
2020-12-02 12:26 ` Brendan Jackman
2020-11-27 17:57 ` [PATCH v2 bpf-next 13/13] bpf: Document new atomic instructions Brendan Jackman
2020-11-28 5:53 ` [PATCH v2 bpf-next 00/13] Atomics for eBPF Yonghong Song
2020-11-29 1:40 ` Alexei Starovoitov
2020-11-30 17:22 ` Yonghong Song
2020-12-01 3:48 ` Yonghong Song
2020-12-02 2:00 ` Andrii Nakryiko
2020-12-02 5:05 ` Yonghong Song
2020-12-02 5:53 ` John Fastabend
2020-12-02 5:59 ` Andrii Nakryiko
2020-12-02 6:27 ` John Fastabend
2020-12-02 8:03 ` Yonghong Song
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=d2e093c3-79bc-0a6b-8919-c5a07667926a@fb.com \
--to=yhs@fb.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=jackmanb@google.com \
--cc=jannh@google.com \
--cc=kpsingh@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=revest@chromium.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
all inboxes | Powered by JetHome®