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, 53 insertions, 0 deletions

diff --git a/tools/testing/selftests/bpf/progs/verifier_subprog_precision.c b/tools/testing/selftests/bpf/progs/verifier_subprog_precision.c
index ac3e418c2a96..61886ed554de 100644
--- a/tools/testing/selftests/bpf/progs/verifier_subprog_precision.c
+++ b/tools/testing/selftests/bpf/progs/verifier_subprog_precision.c
@@ -793,4 +793,57 @@ __naked int stack_slot_aliases_precision(void)
 	);
 }
 
+struct {
+        __uint(type, BPF_MAP_TYPE_PROG_ARRAY);
+        __uint(max_entries, 1);
+        __type(key, __u32);
+        __type(value, __u32);
+} map_array SEC(".maps");
+
+__naked __noinline __used
+static unsigned long identity_tail_call(void)
+{
+	/* the simplest identity function involving a tail call */
+        asm volatile (
+		"r6 = r2;"
+		"r2 = %[map_array] ll;"
+		"r3 = 0;"
+		"call %[bpf_tail_call];"
+		"r0 = r6;"
+		"exit;"
+		:
+		: __imm(bpf_tail_call),
+		  __imm_addr(map_array)
+		: __clobber_all);
+}
+
+SEC("?raw_tp")
+__failure __log_level(2)
+__msg("13: (85) call bpf_tail_call#12")
+__msg("mark_precise: frame1: last_idx 13 first_idx 0 subseq_idx -1 ")
+__msg("returning from callee:")
+__msg("frame1: R0=scalar() R6=3 R10=fp0")
+__msg("to caller at 4:")
+__msg("R0=scalar() R6=map_value(map=.data.vals,ks=4,vs=16) R10=fp0")
+__msg("6: (0f) r1 += r0")
+__msg("mark_precise: frame0: regs=r0 stack= before 5: (bf) r1 = r6")
+__msg("mark_precise: frame0: regs=r0 stack= before 4: (27) r0 *= 4")
+__msg("mark_precise: frame0: parent state regs=r0 stack=:  R0=Pscalar() R6=map_value(map=.data.vals,ks=4,vs=16) R10=fp0")
+__msg("math between map_value pointer and register with unbounded min value is not allowed")
+__naked int subprog_result_tail_call(void)
+{
+	asm volatile (
+		"r2 = 3;"
+		"call identity_tail_call;"
+		"r0 *= 4;"
+		"r1 = %[vals];"
+		"r1 += r0;"
+		"r0 = *(u32 *)(r1 + 0);"
+		"exit;"
+		:
+		: __imm_ptr(vals)
+		: __clobber_common
+	);
+}
+
 char _license[] SEC("license") = "GPL";