Merge tag 'bpf-next-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next

Pull bpf updates from Alexei Starovoitov:

 - Convert selftests/bpf/test_tc_edt and test_tc_tunnel from .sh to
   test_progs runner (Alexis Lothoré)

 - Convert selftests/bpf/test_xsk to test_progs runner (Bastien
   Curutchet)

 - Replace bpf memory allocator with kmalloc_nolock() in
   bpf_local_storage (Amery Hung), and in bpf streams and range tree
   (Puranjay Mohan)

 - Introduce support for indirect jumps in BPF verifier and x86 JIT
   (Anton Protopopov) and arm64 JIT (Puranjay Mohan)

 - Remove runqslower bpf tool (Hoyeon Lee)

 - Fix corner cases in the verifier to close several syzbot reports
   (Eduard Zingerman, KaFai Wan)

 - Several improvements in deadlock detection in rqspinlock (Kumar
   Kartikeya Dwivedi)

 - Implement "jmp" mode for BPF trampoline and corresponding
   DYNAMIC_FTRACE_WITH_JMP. It improves "fexit" program type performance
   from 80 M/s to 136 M/s. With Steven's Ack. (Menglong Dong)

 - Add ability to test non-linear skbs in BPF_PROG_TEST_RUN (Paul
   Chaignon)

 - Do not let BPF_PROG_TEST_RUN emit invalid GSO types to stack (Daniel
   Borkmann)

 - Generalize buildid reader into bpf_dynptr (Mykyta Yatsenko)

 - Optimize bpf_map_update_elem() for map-in-map types (Ritesh
   Oedayrajsingh Varma)

 - Introduce overwrite mode for BPF ring buffer (Xu Kuohai)

* tag 'bpf-next-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (169 commits)
  bpf: optimize bpf_map_update_elem() for map-in-map types
  bpf: make kprobe_multi_link_prog_run always_inline
  selftests/bpf: do not hardcode target rate in test_tc_edt BPF program
  selftests/bpf: remove test_tc_edt.sh
  selftests/bpf: integrate test_tc_edt into test_progs
  selftests/bpf: rename test_tc_edt.bpf.c section to expose program type
  selftests/bpf: Add success stats to rqspinlock stress test
  rqspinlock: Precede non-head waiter queueing with AA check
  rqspinlock: Disable spinning for trylock fallback
  rqspinlock: Use trylock fallback when per-CPU rqnode is busy
  rqspinlock: Perform AA checks immediately
  rqspinlock: Enclose lock/unlock within lock entry acquisitions
  bpf: Remove runqslower tool
  selftests/bpf: Remove usage of lsm/file_alloc_security in selftest
  bpf: Disable file_alloc_security hook
  bpf: check for insn arrays in check_ptr_alignment
  bpf: force BPF_F_RDONLY_PROG on insn array creation
  bpf: Fix exclusive map memory leak
  selftests/bpf: Make CS length configurable for rqspinlock stress test
  selftests/bpf: Add lock wait time stats to rqspinlock stress test
  ...

1 files changed, 28 insertions, 39 deletions

diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index c2fcd0cd51e5..c8a9b27f8663 100644
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -215,19 +215,6 @@ static bool htab_has_extra_elems(struct bpf_htab *htab)
 	return !htab_is_percpu(htab) && !htab_is_lru(htab) && !is_fd_htab(htab);
 }
 
-static void htab_free_internal_structs(struct bpf_htab *htab, struct htab_elem *elem)
-{
-	if (btf_record_has_field(htab->map.record, BPF_TIMER))
-		bpf_obj_free_timer(htab->map.record,
-				   htab_elem_value(elem, htab->map.key_size));
-	if (btf_record_has_field(htab->map.record, BPF_WORKQUEUE))
-		bpf_obj_free_workqueue(htab->map.record,
-				       htab_elem_value(elem, htab->map.key_size));
-	if (btf_record_has_field(htab->map.record, BPF_TASK_WORK))
-		bpf_obj_free_task_work(htab->map.record,
-				       htab_elem_value(elem, htab->map.key_size));
-}
-
 static void htab_free_prealloced_internal_structs(struct bpf_htab *htab)
 {
 	u32 num_entries = htab->map.max_entries;
@@ -240,7 +227,8 @@ static void htab_free_prealloced_internal_structs(struct bpf_htab *htab)
 		struct htab_elem *elem;
 
 		elem = get_htab_elem(htab, i);
-		htab_free_internal_structs(htab, elem);
+		bpf_map_free_internal_structs(&htab->map,
+					      htab_elem_value(elem, htab->map.key_size));
 		cond_resched();
 	}
 }
@@ -669,8 +657,7 @@ static void *__htab_map_lookup_elem(struct bpf_map *map, void *key)
 	struct htab_elem *l;
 	u32 hash, key_size;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -947,15 +934,21 @@ static void free_htab_elem(struct bpf_htab *htab, struct htab_elem *l)
 static void pcpu_copy_value(struct bpf_htab *htab, void __percpu *pptr,
 			    void *value, bool onallcpus)
 {
+	void *ptr;
+
 	if (!onallcpus) {
 		/* copy true value_size bytes */
-		copy_map_value(&htab->map, this_cpu_ptr(pptr), value);
+		ptr = this_cpu_ptr(pptr);
+		copy_map_value(&htab->map, ptr, value);
+		bpf_obj_free_fields(htab->map.record, ptr);
 	} else {
 		u32 size = round_up(htab->map.value_size, 8);
 		int off = 0, cpu;
 
 		for_each_possible_cpu(cpu) {
-			copy_map_value_long(&htab->map, per_cpu_ptr(pptr, cpu), value + off);
+			ptr = per_cpu_ptr(pptr, cpu);
+			copy_map_value_long(&htab->map, ptr, value + off);
+			bpf_obj_free_fields(htab->map.record, ptr);
 			off += size;
 		}
 	}
@@ -1098,8 +1091,7 @@ static long htab_map_update_elem(struct bpf_map *map, void *key, void *value,
 		/* unknown flags */
 		return -EINVAL;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -1206,8 +1198,7 @@ static long htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value
 		/* unknown flags */
 		return -EINVAL;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -1275,8 +1266,7 @@ static long htab_map_update_elem_in_place(struct bpf_map *map, void *key,
 		/* unknown flags */
 		return -EINVAL;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -1338,8 +1328,7 @@ static long __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
 		/* unknown flags */
 		return -EINVAL;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -1416,8 +1405,7 @@ static long htab_map_delete_elem(struct bpf_map *map, void *key)
 	u32 hash, key_size;
 	int ret;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -1452,8 +1440,7 @@ static long htab_lru_map_delete_elem(struct bpf_map *map, void *key)
 	u32 hash, key_size;
 	int ret;
 
-	WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
-		     !rcu_read_lock_bh_held());
+	WARN_ON_ONCE(!bpf_rcu_lock_held());
 
 	key_size = map->key_size;
 
@@ -1509,8 +1496,9 @@ static void htab_free_malloced_internal_structs(struct bpf_htab *htab)
 		struct htab_elem *l;
 
 		hlist_nulls_for_each_entry(l, n, head, hash_node) {
-			/* We only free timer on uref dropping to zero */
-			htab_free_internal_structs(htab, l);
+			/* We only free internal structs on uref dropping to zero */
+			bpf_map_free_internal_structs(&htab->map,
+						      htab_elem_value(l, htab->map.key_size));
 		}
 		cond_resched_rcu();
 	}
@@ -1521,13 +1509,14 @@ static void htab_map_free_internal_structs(struct bpf_map *map)
 {
 	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
 
-	/* We only free timer and workqueue on uref dropping to zero */
-	if (btf_record_has_field(htab->map.record, BPF_TIMER | BPF_WORKQUEUE | BPF_TASK_WORK)) {
-		if (!htab_is_prealloc(htab))
-			htab_free_malloced_internal_structs(htab);
-		else
-			htab_free_prealloced_internal_structs(htab);
-	}
+	/* We only free internal structs on uref dropping to zero */
+	if (!bpf_map_has_internal_structs(map))
+		return;
+
+	if (htab_is_prealloc(htab))
+		htab_free_prealloced_internal_structs(htab);
+	else
+		htab_free_malloced_internal_structs(htab);
 }
 
 /* Called when map->refcnt goes to zero, either from workqueue or from syscall */