| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20171215015517.409513-13-ast@kernel.org>
Date: Thu, 14 Dec 2017 17:55:16 -0800
From: Alexei Starovoitov <ast@...nel.org>
To: "David S . Miller" <davem@...emloft.net>
CC: Daniel Borkmann <daniel@...earbox.net>,
John Fastabend <john.fastabend@...il.com>,
Edward Cree <ecree@...arflare.com>,
Jakub Kicinski <jakub.kicinski@...ronome.com>,
<netdev@...r.kernel.org>, <kernel-team@...com>
Subject: [PATCH bpf-next 12/13] bpf: arm64: add JIT support for multi-function programs
From: Alexei Starovoitov <ast@...com>
similar to x64 add support for bpf-to-bpf calls.
When program has calls to in-kernel helpers the target call offset
is known at JIT time and arm64 architecture needs 2 passes.
With bpf-to-bpf calls the dynamically allocated function start
is unknown until all functions of the program are JITed.
Therefore (just like x64) arm64 JIT needs one extra pass over
the program to emit correct call offsets.
Implementation detail:
Avoid being too clever in 64-bit immediate moves and
always use 4 instructions (instead of 3-4 depending on the address)
to make sure only one extra pass is needed.
If some future optimization would make it worth while to optimize
'call 64-bit imm' further, the JIT would need to do 4 passes
over the program instead of 3 as in this patch.
For typical bpf program address the mov needs 3 or 4 insns,
so unconditional 4 insns to save extra pass is a worthy trade off
at this state of JIT.
Signed-off-by: Alexei Starovoitov <ast@...nel.org>
Acked-by: Daniel Borkmann <daniel@...earbox.net>
---
arch/arm64/net/bpf_jit_comp.c | 68 ++++++++++++++++++++++++++++++++++++++++---
1 file changed, 64 insertions(+), 4 deletions(-)
diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c
index 288137cb0871..396490cf7316 100644
--- a/arch/arm64/net/bpf_jit_comp.c
+++ b/arch/arm64/net/bpf_jit_comp.c
@@ -99,6 +99,20 @@ static inline void emit_a64_mov_i64(const int reg, const u64 val,
}
}
+static inline void emit_addr_mov_i64(const int reg, const u64 val,
+ struct jit_ctx *ctx)
+{
+ u64 tmp = val;
+ int shift = 0;
+
+ emit(A64_MOVZ(1, reg, tmp & 0xffff, shift), ctx);
+ for (;shift < 48;) {
+ tmp >>= 16;
+ shift += 16;
+ emit(A64_MOVK(1, reg, tmp & 0xffff, shift), ctx);
+ }
+}
+
static inline void emit_a64_mov_i(const int is64, const int reg,
const s32 val, struct jit_ctx *ctx)
{
@@ -603,7 +617,10 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx)
const u8 r0 = bpf2a64[BPF_REG_0];
const u64 func = (u64)__bpf_call_base + imm;
- emit_a64_mov_i64(tmp, func, ctx);
+ if (ctx->prog->is_func)
+ emit_addr_mov_i64(tmp, func, ctx);
+ else
+ emit_a64_mov_i64(tmp, func, ctx);
emit(A64_BLR(tmp), ctx);
emit(A64_MOV(1, r0, A64_R(0)), ctx);
break;
@@ -835,11 +852,19 @@ static inline void bpf_flush_icache(void *start, void *end)
flush_icache_range((unsigned long)start, (unsigned long)end);
}
+struct arm64_jit_data {
+ struct bpf_binary_header *header;
+ u8 *image;
+ struct jit_ctx ctx;
+};
+
struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
{
struct bpf_prog *tmp, *orig_prog = prog;
struct bpf_binary_header *header;
+ struct arm64_jit_data *jit_data;
bool tmp_blinded = false;
+ bool extra_pass = false;
struct jit_ctx ctx;
int image_size;
u8 *image_ptr;
@@ -858,13 +883,29 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
prog = tmp;
}
+ jit_data = prog->aux->jit_data;
+ if (!jit_data) {
+ jit_data = kzalloc(sizeof(*jit_data), GFP_KERNEL);
+ if (!jit_data) {
+ prog = orig_prog;
+ goto out;
+ }
+ prog->aux->jit_data = jit_data;
+ }
+ if (jit_data->ctx.offset) {
+ ctx = jit_data->ctx;
+ image_ptr = jit_data->image;
+ header = jit_data->header;
+ extra_pass = true;
+ goto skip_init_ctx;
+ }
memset(&ctx, 0, sizeof(ctx));
ctx.prog = prog;
ctx.offset = kcalloc(prog->len, sizeof(int), GFP_KERNEL);
if (ctx.offset == NULL) {
prog = orig_prog;
- goto out;
+ goto out_off;
}
/* 1. Initial fake pass to compute ctx->idx. */
@@ -895,6 +936,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
/* 2. Now, the actual pass. */
ctx.image = (__le32 *)image_ptr;
+skip_init_ctx:
ctx.idx = 0;
build_prologue(&ctx);
@@ -920,13 +962,31 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
bpf_flush_icache(header, ctx.image + ctx.idx);
- bpf_jit_binary_lock_ro(header);
+ if (!prog->is_func || extra_pass) {
+ if (extra_pass && ctx.idx != jit_data->ctx.idx) {
+ pr_err_once("multi-func JIT bug %d != %d\n",
+ ctx.idx, jit_data->ctx.idx);
+ bpf_jit_binary_free(header);
+ prog->bpf_func = NULL;
+ prog->jited = 0;
+ goto out_off;
+ }
+ bpf_jit_binary_lock_ro(header);
+ } else {
+ jit_data->ctx = ctx;
+ jit_data->image = image_ptr;
+ jit_data->header = header;
+ }
prog->bpf_func = (void *)ctx.image;
prog->jited = 1;
prog->jited_len = image_size;
+ if (!prog->is_func || extra_pass) {
out_off:
- kfree(ctx.offset);
+ kfree(ctx.offset);
+ kfree(jit_data);
+ prog->aux->jit_data = NULL;
+ }
out:
if (tmp_blinded)
bpf_jit_prog_release_other(prog, prog == orig_prog ?
--
2.9.5
Powered by blists - more mailing lists