From: Peilin Ye <yepeilin@google.com>
To: bpf@vger.kernel.org, linux-arm-kernel@lists.infradead.org
Cc: Peilin Ye <yepeilin@google.com>,
bpf@ietf.org, Alexei Starovoitov <ast@kernel.org>,
Xu Kuohai <xukuohai@huaweicloud.com>,
Eduard Zingerman <eddyz87@gmail.com>,
David Vernet <void@manifault.com>,
Daniel Borkmann <daniel@iogearbox.net>,
Andrii Nakryiko <andrii@kernel.org>,
Martin KaFai Lau <martin.lau@linux.dev>,
Song Liu <song@kernel.org>,
Yonghong Song <yonghong.song@linux.dev>,
John Fastabend <john.fastabend@gmail.com>,
KP Singh <kpsingh@kernel.org>,
Stanislav Fomichev <sdf@fomichev.me>, Hao Luo <haoluo@google.com>,
Jiri Olsa <jolsa@kernel.org>, Jonathan Corbet <corbet@lwn.net>,
"Paul E. McKenney" <paulmck@kernel.org>,
Puranjay Mohan <puranjay@kernel.org>,
Ilya Leoshkevich <iii@linux.ibm.com>,
Heiko Carstens <hca@linux.ibm.com>,
Vasily Gorbik <gor@linux.ibm.com>,
Catalin Marinas <catalin.marinas@arm.com>,
Will Deacon <will@kernel.org>, Quentin Monnet <qmo@kernel.org>,
Mykola Lysenko <mykolal@fb.com>, Shuah Khan <shuah@kernel.org>,
Ihor Solodrai <ihor.solodrai@linux.dev>,
Yingchi Long <longyingchi24s@ict.ac.cn>,
Josh Don <joshdon@google.com>, Barret Rhoden <brho@google.com>,
Neel Natu <neelnatu@google.com>,
Benjamin Segall <bsegall@google.com>,
linux-kernel@vger.kernel.org
Subject: [PATCH bpf-next v5 5/6] bpf, x86: Support load-acquire and store-release instructions
Date: Tue, 4 Mar 2025 00:14:22 +0000 [thread overview]
Message-ID: <c4b2b97086e214ee24aafeb5f57cb1275a4469ef.1741046028.git.yepeilin@google.com> (raw)
In-Reply-To: <cover.1741046028.git.yepeilin@google.com>
Recently we introduced BPF load-acquire (BPF_LOAD_ACQ) and store-release
(BPF_STORE_REL) instructions. For x86-64, simply implement them as
regular BPF_LDX/BPF_STX loads and stores. The verifier always rejects
misaligned load-acquires/store-releases (even if BPF_F_ANY_ALIGNMENT is
set), so emitted MOV* instructions are guaranteed to be atomic.
Arena accesses are supported. 8- and 16-bit load-acquires are
zero-extending (i.e., MOVZBQ, MOVZWQ).
Rename emit_atomic{,_index}() to emit_atomic_rmw{,_index}() to make it
clear that they only handle read-modify-write atomics, and extend their
@atomic_op parameter from u8 to u32, since we are starting to use more
than the lowest 8 bits of the 'imm' field.
Signed-off-by: Peilin Ye <yepeilin@google.com>
---
arch/x86/net/bpf_jit_comp.c | 99 ++++++++++++++++++++++++++++++-------
1 file changed, 82 insertions(+), 17 deletions(-)
diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index f0c31c940fb8..0263d98d92b0 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -1242,8 +1242,8 @@ static void emit_st_r12(u8 **pprog, u32 size, u32 dst_reg, int off, int imm)
emit_st_index(pprog, size, dst_reg, X86_REG_R12, off, imm);
}
-static int emit_atomic(u8 **pprog, u8 atomic_op,
- u32 dst_reg, u32 src_reg, s16 off, u8 bpf_size)
+static int emit_atomic_rmw(u8 **pprog, u32 atomic_op,
+ u32 dst_reg, u32 src_reg, s16 off, u8 bpf_size)
{
u8 *prog = *pprog;
@@ -1283,8 +1283,9 @@ static int emit_atomic(u8 **pprog, u8 atomic_op,
return 0;
}
-static int emit_atomic_index(u8 **pprog, u8 atomic_op, u32 size,
- u32 dst_reg, u32 src_reg, u32 index_reg, int off)
+static int emit_atomic_rmw_index(u8 **pprog, u32 atomic_op, u32 size,
+ u32 dst_reg, u32 src_reg, u32 index_reg,
+ int off)
{
u8 *prog = *pprog;
@@ -1297,7 +1298,7 @@ static int emit_atomic_index(u8 **pprog, u8 atomic_op, u32 size,
EMIT1(add_3mod(0x48, dst_reg, src_reg, index_reg));
break;
default:
- pr_err("bpf_jit: 1 and 2 byte atomics are not supported\n");
+ pr_err("bpf_jit: 1- and 2-byte RMW atomics are not supported\n");
return -EFAULT;
}
@@ -1331,6 +1332,49 @@ static int emit_atomic_index(u8 **pprog, u8 atomic_op, u32 size,
return 0;
}
+static int emit_atomic_ld_st(u8 **pprog, u32 atomic_op, u32 dst_reg,
+ u32 src_reg, s16 off, u8 bpf_size)
+{
+ switch (atomic_op) {
+ case BPF_LOAD_ACQ:
+ /* dst_reg = smp_load_acquire(src_reg + off16) */
+ emit_ldx(pprog, bpf_size, dst_reg, src_reg, off);
+ break;
+ case BPF_STORE_REL:
+ /* smp_store_release(dst_reg + off16, src_reg) */
+ emit_stx(pprog, bpf_size, dst_reg, src_reg, off);
+ break;
+ default:
+ pr_err("bpf_jit: unknown atomic load/store opcode %02x\n",
+ atomic_op);
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
+static int emit_atomic_ld_st_index(u8 **pprog, u32 atomic_op, u32 size,
+ u32 dst_reg, u32 src_reg, u32 index_reg,
+ int off)
+{
+ switch (atomic_op) {
+ case BPF_LOAD_ACQ:
+ /* dst_reg = smp_load_acquire(src_reg + idx_reg + off16) */
+ emit_ldx_index(pprog, size, dst_reg, src_reg, index_reg, off);
+ break;
+ case BPF_STORE_REL:
+ /* smp_store_release(dst_reg + idx_reg + off16, src_reg) */
+ emit_stx_index(pprog, size, dst_reg, src_reg, index_reg, off);
+ break;
+ default:
+ pr_err("bpf_jit: unknown atomic load/store opcode %02x\n",
+ atomic_op);
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
#define DONT_CLEAR 1
bool ex_handler_bpf(const struct exception_table_entry *x, struct pt_regs *regs)
@@ -2113,6 +2157,13 @@ st: if (is_imm8(insn->off))
}
break;
+ case BPF_STX | BPF_ATOMIC | BPF_B:
+ case BPF_STX | BPF_ATOMIC | BPF_H:
+ if (!bpf_atomic_is_load_store(insn)) {
+ pr_err("bpf_jit: 1- and 2-byte RMW atomics are not supported\n");
+ return -EFAULT;
+ }
+ fallthrough;
case BPF_STX | BPF_ATOMIC | BPF_W:
case BPF_STX | BPF_ATOMIC | BPF_DW:
if (insn->imm == (BPF_AND | BPF_FETCH) ||
@@ -2148,10 +2199,10 @@ st: if (is_imm8(insn->off))
EMIT2(simple_alu_opcodes[BPF_OP(insn->imm)],
add_2reg(0xC0, AUX_REG, real_src_reg));
/* Attempt to swap in new value */
- err = emit_atomic(&prog, BPF_CMPXCHG,
- real_dst_reg, AUX_REG,
- insn->off,
- BPF_SIZE(insn->code));
+ err = emit_atomic_rmw(&prog, BPF_CMPXCHG,
+ real_dst_reg, AUX_REG,
+ insn->off,
+ BPF_SIZE(insn->code));
if (WARN_ON(err))
return err;
/*
@@ -2166,17 +2217,35 @@ st: if (is_imm8(insn->off))
break;
}
- err = emit_atomic(&prog, insn->imm, dst_reg, src_reg,
- insn->off, BPF_SIZE(insn->code));
+ if (bpf_atomic_is_load_store(insn))
+ err = emit_atomic_ld_st(&prog, insn->imm, dst_reg, src_reg,
+ insn->off, BPF_SIZE(insn->code));
+ else
+ err = emit_atomic_rmw(&prog, insn->imm, dst_reg, src_reg,
+ insn->off, BPF_SIZE(insn->code));
if (err)
return err;
break;
+ case BPF_STX | BPF_PROBE_ATOMIC | BPF_B:
+ case BPF_STX | BPF_PROBE_ATOMIC | BPF_H:
+ if (!bpf_atomic_is_load_store(insn)) {
+ pr_err("bpf_jit: 1- and 2-byte RMW atomics are not supported\n");
+ return -EFAULT;
+ }
+ fallthrough;
case BPF_STX | BPF_PROBE_ATOMIC | BPF_W:
case BPF_STX | BPF_PROBE_ATOMIC | BPF_DW:
start_of_ldx = prog;
- err = emit_atomic_index(&prog, insn->imm, BPF_SIZE(insn->code),
- dst_reg, src_reg, X86_REG_R12, insn->off);
+
+ if (bpf_atomic_is_load_store(insn))
+ err = emit_atomic_ld_st_index(&prog, insn->imm,
+ BPF_SIZE(insn->code), dst_reg,
+ src_reg, X86_REG_R12, insn->off);
+ else
+ err = emit_atomic_rmw_index(&prog, insn->imm, BPF_SIZE(insn->code),
+ dst_reg, src_reg, X86_REG_R12,
+ insn->off);
if (err)
return err;
goto populate_extable;
@@ -3771,12 +3840,8 @@ bool bpf_jit_supports_insn(struct bpf_insn *insn, bool in_arena)
if (!in_arena)
return true;
switch (insn->code) {
- case BPF_STX | BPF_ATOMIC | BPF_B:
- case BPF_STX | BPF_ATOMIC | BPF_H:
case BPF_STX | BPF_ATOMIC | BPF_W:
case BPF_STX | BPF_ATOMIC | BPF_DW:
- if (bpf_atomic_is_load_store(insn))
- return false;
if (insn->imm == (BPF_AND | BPF_FETCH) ||
insn->imm == (BPF_OR | BPF_FETCH) ||
insn->imm == (BPF_XOR | BPF_FETCH))
--
2.48.1.711.g2feabab25a-goog
next prev parent reply other threads:[~2025-03-04 0:14 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-04 0:13 [PATCH bpf-next v5 0/6] Introduce load-acquire and store-release BPF instructions Peilin Ye
2025-03-04 0:13 ` [PATCH bpf-next v5 1/6] bpf: Introduce load-acquire and store-release instructions Peilin Ye
2025-03-04 0:24 ` Alexei Starovoitov
2025-03-04 0:39 ` Peilin Ye
2025-03-04 0:45 ` Alexei Starovoitov
2025-03-04 0:49 ` Peilin Ye
2025-03-04 0:13 ` [PATCH bpf-next v5 2/6] arm64: insn: Add BIT(23) to {load,store}_ex's mask Peilin Ye
2025-03-04 0:13 ` [PATCH bpf-next v5 3/6] arm64: insn: Add load-acquire and store-release instructions Peilin Ye
2025-03-04 0:14 ` [PATCH bpf-next v5 4/6] bpf, arm64: Support " Peilin Ye
2025-03-04 0:14 ` Peilin Ye [this message]
2025-03-04 0:14 ` [PATCH bpf-next v5 6/6] selftests/bpf: Add selftests for " Peilin Ye
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=c4b2b97086e214ee24aafeb5f57cb1275a4469ef.1741046028.git.yepeilin@google.com \
--to=yepeilin@google.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@ietf.org \
--cc=bpf@vger.kernel.org \
--cc=brho@google.com \
--cc=bsegall@google.com \
--cc=catalin.marinas@arm.com \
--cc=corbet@lwn.net \
--cc=daniel@iogearbox.net \
--cc=eddyz87@gmail.com \
--cc=gor@linux.ibm.com \
--cc=haoluo@google.com \
--cc=hca@linux.ibm.com \
--cc=ihor.solodrai@linux.dev \
--cc=iii@linux.ibm.com \
--cc=john.fastabend@gmail.com \
--cc=jolsa@kernel.org \
--cc=joshdon@google.com \
--cc=kpsingh@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=longyingchi24s@ict.ac.cn \
--cc=martin.lau@linux.dev \
--cc=mykolal@fb.com \
--cc=neelnatu@google.com \
--cc=paulmck@kernel.org \
--cc=puranjay@kernel.org \
--cc=qmo@kernel.org \
--cc=sdf@fomichev.me \
--cc=shuah@kernel.org \
--cc=song@kernel.org \
--cc=void@manifault.com \
--cc=will@kernel.org \
--cc=xukuohai@huaweicloud.com \
--cc=yonghong.song@linux.dev \
/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®