bpf: no_caller_saved_registers attribute for helper calls

GCC and LLVM define a no_caller_saved_registers function attribute.
This attribute means that function scratches only some of
the caller saved registers defined by ABI.
For BPF the set of such registers could be defined as follows:
- R0 is scratched only if function is non-void;
- R1-R5 are scratched only if corresponding parameter type is defined
  in the function prototype.

This commit introduces flag bpf_func_prot->allow_nocsr.
If this flag is set for some helper function, verifier assumes that
it follows no_caller_saved_registers calling convention.

The contract between kernel and clang allows to simultaneously use
such functions and maintain backwards compatibility with old
kernels that don't understand no_caller_saved_registers calls
(nocsr for short):

- clang generates a simple pattern for nocsr calls, e.g.:

    r1 = 1;
    r2 = 2;
    *(u64 *)(r10 - 8)  = r1;
    *(u64 *)(r10 - 16) = r2;
    call %[to_be_inlined]
    r2 = *(u64 *)(r10 - 16);
    r1 = *(u64 *)(r10 - 8);
    r0 = r1;
    r0 += r2;
    exit;

- kernel removes unnecessary spills and fills, if called function is
  inlined by verifier or current JIT (with assumption that patch
  inserted by verifier or JIT honors nocsr contract, e.g. does not
  scratch r3-r5 for the example above), e.g. the code above would be
  transformed to:

    r1 = 1;
    r2 = 2;
    call %[to_be_inlined]
    r0 = r1;
    r0 += r2;
    exit;

Technically, the transformation is split into the following phases:
- function mark_nocsr_patterns(), called from bpf_check()
  searches and marks potential patterns in instruction auxiliary data;
- upon stack read or write access,
  function check_nocsr_stack_contract() is used to verify if
  stack offsets, presumably reserved for nocsr patterns, are used
  only from those patterns;
- function remove_nocsr_spills_fills(), called from bpf_check(),
  applies the rewrite for valid patterns.

See comment in mark_nocsr_pattern_for_call() for more details.

Suggested-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240722233844.1406874-3-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

2 files changed, 20 insertions, 0 deletions

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index f560ea0c2b36..b9425e410bcb 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -808,6 +808,12 @@ struct bpf_func_proto {
 	bool gpl_only;
 	bool pkt_access;
 	bool might_sleep;
+	/* set to true if helper follows contract for gcc/llvm
+	 * attribute no_caller_saved_registers:
+	 * - void functions do not scratch r0
+	 * - functions taking N arguments scratch only registers r1-rN
+	 */
+	bool allow_nocsr;
 	enum bpf_return_type ret_type;
 	union {
 		struct {
diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
index 731a0a4ac13c..5cea15c81b8a 100644
--- a/include/linux/bpf_verifier.h
+++ b/include/linux/bpf_verifier.h
@@ -576,6 +576,14 @@ struct bpf_insn_aux_data {
 	bool is_iter_next; /* bpf_iter_<type>_next() kfunc call */
 	bool call_with_percpu_alloc_ptr; /* {this,per}_cpu_ptr() with prog percpu alloc */
 	u8 alu_state; /* used in combination with alu_limit */
+	/* true if STX or LDX instruction is a part of a spill/fill
+	 * pattern for a no_caller_saved_registers call.
+	 */
+	u8 nocsr_pattern:1;
+	/* for CALL instructions, a number of spill/fill pairs in the
+	 * no_caller_saved_registers pattern.
+	 */
+	u8 nocsr_spills_num:3;
 
 	/* below fields are initialized once */
 	unsigned int orig_idx; /* original instruction index */
@@ -645,6 +653,10 @@ struct bpf_subprog_info {
 	u32 linfo_idx; /* The idx to the main_prog->aux->linfo */
 	u16 stack_depth; /* max. stack depth used by this function */
 	u16 stack_extra;
+	/* offsets in range [stack_depth .. nocsr_stack_off)
+	 * are used for no_caller_saved_registers spills and fills.
+	 */
+	s16 nocsr_stack_off;
 	bool has_tail_call: 1;
 	bool tail_call_reachable: 1;
 	bool has_ld_abs: 1;
@@ -652,6 +664,8 @@ struct bpf_subprog_info {
 	bool is_async_cb: 1;
 	bool is_exception_cb: 1;
 	bool args_cached: 1;
+	/* true if nocsr stack region is used by functions that can't be inlined */
+	bool keep_nocsr_stack: 1;
 
 	u8 arg_cnt;
 	struct bpf_subprog_arg_info args[MAX_BPF_FUNC_REG_ARGS];