Skip to contentSkip to conversation

Fix for arm64 jit

PATCH Applied Daniel Borkmann 2 participants 4 messages Nov 26, 2018 – Nov 27, 2018

Subsystems (8)
Networking Network drivers PowerPC BPF +4 more subsystems
lore
Share & tools
lore

Daniel Borkmann

author maintainer
#0 lore raw

cover Fix for arm64 jit

This set contains a fix for arm64 BPF JIT. First patch generalizes ppc64 way of retrieving subprog into bpf_jit_get_func_addr() as core code and uses the same on arm64 in second patch. Tested on both arm64 and ppc64.

Thanks!

Daniel Borkmann (2):
bpf, ppc64: generalize fetching subprog into bpf_jit_get_func_addr bpf, arm64: fix getting subprog addr from aux for calls

 arch/arm64/net/bpf_jit_comp.c     | 26 +++++++++++-------
 arch/powerpc/net/bpf_jit_comp64.c | 57 ++++++++++++++++++++++++++-------------
 include/linux/filter.h            |  4 +++
 kernel/bpf/core.c                 | 34 +++++++++++++++++++++++
 4 files changed, 93 insertions(+), 28 deletions(-)
Signature · 1 line
2.9.5
author maintainer
#1 lore raw

PATCH 1/2 bpf, ppc64: generalize fetching subprog into bpf_jit_get_func_addr

Make fetching of the BPF call address from ppc64 JIT generic. ppc64 was using a slightly different variant rather than through the insns' imm field encoding as the target address would not fit into that space. Therefore, the target subprog number was encoded into the insns' offset and fetched through fp->aux->func[off]->bpf_func instead. Given there are other JITs with this issue and the mechanism of fetching the address is JIT-generic, move it into the core as a helper instead. On the JIT side, we get information on whether the retrieved address is a fixed one, that is, not changing through JIT passes, or a dynamic one. For the former, JITs can optimize their imm emission because this doesn't change jump offsets throughout JIT process.

Signed-off-by Daniel Borkmann <[email protected]>
Reviewed-by Sandipan Das <[email protected]>
Tested-by Sandipan Das <[email protected]>
---
 arch/powerpc/net/bpf_jit_comp64.c | 57 ++++++++++++++++++++++++++-------------
 include/linux/filter.h            |  4 +++
 kernel/bpf/core.c                 | 34 +++++++++++++++++++++++
 3 files changed, 76 insertions(+), 19 deletions(-)
arch/powerpc/net/bpf_jit_comp64.c +38 -19
index 50b1297..17482f5 100644
@@ -166,7 +166,33 @@ static void bpf_jit_build_epilogue(u32 *image, struct codegen_context *ctx)
 	PPC_BLR();
 }
 
-static void bpf_jit_emit_func_call(u32 *image, struct codegen_context *ctx, u64 func)
+static void bpf_jit_emit_func_call_hlp(u32 *image, struct codegen_context *ctx,
+				       u64 func)
+{
+#ifdef PPC64_ELF_ABI_v1
+	/* func points to the function descriptor */
+	PPC_LI64(b2p[TMP_REG_2], func);
+	/* Load actual entry point from function descriptor */
+	PPC_BPF_LL(b2p[TMP_REG_1], b2p[TMP_REG_2], 0);
+	/* ... and move it to LR */
+	PPC_MTLR(b2p[TMP_REG_1]);
+	/*
+	 * Load TOC from function descriptor at offset 8.
+	 * We can clobber r2 since we get called through a
+	 * function pointer (so caller will save/restore r2)
+	 * and since we don't use a TOC ourself.
+	 */
+	PPC_BPF_LL(2, b2p[TMP_REG_2], 8);
+#else
+	/* We can clobber r12 */
+	PPC_FUNC_ADDR(12, func);
+	PPC_MTLR(12);
+#endif
+	PPC_BLRL();
+}
+
+static void bpf_jit_emit_func_call_rel(u32 *image, struct codegen_context *ctx,
+				       u64 func)
 {
 	unsigned int i, ctx_idx = ctx->idx;
 

@@ -273,7 +299,7 @@ static int bpf_jit_build_body(struct bpf_prog *fp, u32 *image,
 {
 	const struct bpf_insn *insn = fp->insnsi;
 	int flen = fp->len;
-	int i;
+	int i, ret;
 
 	/* Start of epilogue code - will only be valid 2nd pass onwards */
 	u32 exit_addr = addrs[flen];

@@ -284,8 +310,9 @@ static int bpf_jit_build_body(struct bpf_prog *fp, u32 *image,
 		u32 src_reg = b2p[insn[i].src_reg];
 		s16 off = insn[i].off;
 		s32 imm = insn[i].imm;
+		bool func_addr_fixed;
+		u64 func_addr;
 		u64 imm64;
-		u8 *func;
 		u32 true_cond;
 		u32 tmp_idx;
 

@@ -711,23 +738,15 @@ static int bpf_jit_build_body(struct bpf_prog *fp, u32 *image,
 		case BPF_JMP | BPF_CALL:
 			ctx->seen |= SEEN_FUNC;
 
-			/* bpf function call */
-			if (insn[i].src_reg == BPF_PSEUDO_CALL)
-				if (!extra_pass)
-					func = NULL;
-				else if (fp->aux->func && off < fp->aux->func_cnt)
-					/* use the subprog id from the off
-					 * field to lookup the callee address
-					 */
-					func = (u8 *) fp->aux->func[off]->bpf_func;
-				else
-					return -EINVAL;
-			/* kernel helper call */
-			else
-				func = (u8 *) __bpf_call_base + imm;
-
-			bpf_jit_emit_func_call(image, ctx, (u64)func);
+			ret = bpf_jit_get_func_addr(fp, &insn[i], extra_pass,
+						    &func_addr, &func_addr_fixed);
+			if (ret < 0)
+				return ret;
 
+			if (func_addr_fixed)
+				bpf_jit_emit_func_call_hlp(image, ctx, func_addr);
+			else
+				bpf_jit_emit_func_call_rel(image, ctx, func_addr);
 			/* move return value from r3 to BPF_REG_0 */
 			PPC_MR(b2p[BPF_REG_0], 3);
 			break;
include/linux/filter.h +4 -0
index de629b7..448dcc4 100644
@@ -866,6 +866,10 @@ void bpf_jit_binary_free(struct bpf_binary_header *hdr);
 
 void bpf_jit_free(struct bpf_prog *fp);
 
+int bpf_jit_get_func_addr(const struct bpf_prog *prog,
+			  const struct bpf_insn *insn, bool extra_pass,
+			  u64 *func_addr, bool *func_addr_fixed);
+
 struct bpf_prog *bpf_jit_blind_constants(struct bpf_prog *fp);
 void bpf_jit_prog_release_other(struct bpf_prog *fp, struct bpf_prog *fp_other);
kernel/bpf/core.c +34 -0
index 1a796e0..b1a3545 100644
@@ -672,6 +672,40 @@ void __weak bpf_jit_free(struct bpf_prog *fp)
 	bpf_prog_unlock_free(fp);
 }
 
+int bpf_jit_get_func_addr(const struct bpf_prog *prog,
+			  const struct bpf_insn *insn, bool extra_pass,
+			  u64 *func_addr, bool *func_addr_fixed)
+{
+	s16 off = insn->off;
+	s32 imm = insn->imm;
+	u8 *addr;
+
+	*func_addr_fixed = insn->src_reg != BPF_PSEUDO_CALL;
+	if (!*func_addr_fixed) {
+		/* Place-holder address till the last pass has collected
+		 * all addresses for JITed subprograms in which case we
+		 * can pick them up from prog->aux.
+		 */
+		if (!extra_pass)
+			addr = NULL;
+		else if (prog->aux->func &&
+			 off >= 0 && off < prog->aux->func_cnt)
+			addr = (u8 *)prog->aux->func[off]->bpf_func;
+		else
+			return -EINVAL;
+	} else {
+		/* Address of a BPF helper call. Since part of the core
+		 * kernel, it's always at a fixed location. __bpf_call_base
+		 * and the helper with imm relative to it are both in core
+		 * kernel.
+		 */
+		addr = (u8 *)__bpf_call_base + imm;
+	}
+
+	*func_addr = (unsigned long)addr;
+	return 0;
+}
+
 static int bpf_jit_blind_insn(const struct bpf_insn *from,
 			      const struct bpf_insn *aux,
 			      struct bpf_insn *to_buff)
Signature · 1 line
2.9.5
author maintainer
#2 lore raw

PATCH 2/2 bpf, arm64: fix getting subprog addr from aux for calls

In reply to Daniel Borkmann

The arm64 JIT has the same issue as ppc64 JIT in that the relative BPF to BPF call offset can be too far away from core kernel in that relative encoding into imm is not sufficient and could potentially be truncated, see also fd045f6cd98e ("arm64: add support for module PLTs") which adds spill-over space for module_alloc() and therefore bpf_jit_binary_alloc(). Therefore, use the recently added bpf_jit_get_func_addr() helper for properly fetching the address through prog->aux->func[off]->bpf_func instead. This also has the benefit to optimize normal helper calls since their address can use the optimized emission. Tested on Cavium ThunderX CN8890.

Fixes db496944fdaa ("bpf: arm64: add JIT support for multi-function programs")
Signed-off-by Daniel Borkmann <[email protected]>
---
 arch/arm64/net/bpf_jit_comp.c | 26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)
arch/arm64/net/bpf_jit_comp.c +17 -9
index a6fdaea..89198017 100644
@@ -351,7 +351,8 @@ static void build_epilogue(struct jit_ctx *ctx)
  * >0 - successfully JITed a 16-byte eBPF instruction.
  * <0 - failed to JIT.
  */
-static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx)
+static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx,
+		      bool extra_pass)
 {
 	const u8 code = insn->code;
 	const u8 dst = bpf2a64[insn->dst_reg];

@@ -625,12 +626,19 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx)
 	case BPF_JMP | BPF_CALL:
 	{
 		const u8 r0 = bpf2a64[BPF_REG_0];
-		const u64 func = (u64)__bpf_call_base + imm;
+		bool func_addr_fixed;
+		u64 func_addr;
+		int ret;
 
-		if (ctx->prog->is_func)
-			emit_addr_mov_i64(tmp, func, ctx);
+		ret = bpf_jit_get_func_addr(ctx->prog, insn, extra_pass,
+					    &func_addr, &func_addr_fixed);
+		if (ret < 0)
+			return ret;
+		if (func_addr_fixed)
+			/* We can use optimized emission here. */
+			emit_a64_mov_i64(tmp, func_addr, ctx);
 		else
-			emit_a64_mov_i64(tmp, func, ctx);
+			emit_addr_mov_i64(tmp, func_addr, ctx);
 		emit(A64_BLR(tmp), ctx);
 		emit(A64_MOV(1, r0, A64_R(0)), ctx);
 		break;

@@ -753,7 +761,7 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx)
 	return 0;
 }
 
-static int build_body(struct jit_ctx *ctx)
+static int build_body(struct jit_ctx *ctx, bool extra_pass)
 {
 	const struct bpf_prog *prog = ctx->prog;
 	int i;

@@ -762,7 +770,7 @@ static int build_body(struct jit_ctx *ctx)
 		const struct bpf_insn *insn = &prog->insnsi[i];
 		int ret;
 
-		ret = build_insn(insn, ctx);
+		ret = build_insn(insn, ctx, extra_pass);
 		if (ret > 0) {
 			i++;
 			if (ctx->image == NULL)

@@ -858,7 +866,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
 	/* 1. Initial fake pass to compute ctx->idx. */
 
 	/* Fake pass to fill in ctx->offset. */
-	if (build_body(&ctx)) {
+	if (build_body(&ctx, extra_pass)) {
 		prog = orig_prog;
 		goto out_off;
 	}

@@ -888,7 +896,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
 
 	build_prologue(&ctx, was_classic);
 
-	if (build_body(&ctx)) {
+	if (build_body(&ctx, extra_pass)) {
 		bpf_jit_binary_free(header);
 		prog = orig_prog;
 		goto out_off;
Signature · 1 line
2.9.5
Applied
#3 lore raw
In reply to Daniel Borkmann
Quoted 6 lines · Daniel Borkmann

This set contains a fix for arm64 BPF JIT. First patch generalizes ppc64 way of retrieving subprog into bpf_jit_get_func_addr() as core code and uses the same on arm64 in second patch. Tested on both arm64 and ppc64.

Thanks!

Applied, Thanks