)]}'
{
  "commit": "8835113107b2cf75df33a71e265cdeec5d5d13e0",
  "tree": "23b1e505038dc6757b8e2081eb2f56d81daded05",
  "parents": [
    "283fb1db0d4262369c52bb7d774674d0df125b70"
  ],
  "author": {
    "name": "Alexei Starovoitov",
    "email": "ast@kernel.org",
    "time": "Sat Nov 14 23:27:54 2026 +0000"
  },
  "committer": {
    "name": "Alexei Starovoitov",
    "email": "ast@kernel.org",
    "time": "Sun Nov 15 01:55:54 2026 +0000"
  },
  "message": "resolve_btfids: Don\u0027t list -EBUSY next to -EINPROGRESS\n\ncrypto_enqueue_request() returns three codes. Only -ENOSPC is a failure:\n\n\tint err \u003d -EINPROGRESS;\n\n\tif (unlikely(queue-\u003eqlen \u003e\u003d queue-\u003emax_qlen)) {\n\t\tif (!(request-\u003eflags \u0026 CRYPTO_TFM_REQ_MAY_BACKLOG)) {\n\t\t\terr \u003d -ENOSPC;\n\t\t\tgoto out;\n\t\t}\n\t\terr \u003d -EBUSY;\n\t\t..\n\t}\n\tqueue-\u003eqlen++;\n\tlist_add_tail(\u0026request-\u003elist, \u0026queue-\u003elist);\nout:\n\treturn err;\n\n-EBUSY says that the request is queued on the backlog. crypto_wait_req()\nwaits for the completion after -EINPROGRESS and after -EBUSY. The tag of\nthe function is \u0027bpf_fmodret_retval:-28,-16\u0027. An fmod_ret program that\nreturns -EBUSY skips the list_add_tail(). The caller then waits for\na completion that never comes.\n\nmeans_failure() cannot drop -EBUSY for all functions. In most of them it\nis a plain failure. In an x86-64 GCC build 645 of the 11432 tagged\nfunctions list it, and it is the only error of 205 of them.\n\nccp_enqueue_cmd() and qat_alg_send_message_maybacklog() have such\na backlog too. All three functions also return -EINPROGRESS. The rule:\n\n\t-EBUSY is not in the list if the function returns -EINPROGRESS too.\n\ncrypto_enqueue_request() stays a target with -ENOSPC.\n\n\t                list -EBUSY     lose it\n\tx86-64 GCC      645 of 11432    1\n\tx86-64 clang    674 of 12096    2\n\tarm64 clang     638 of 10954    2\n\nNo function loses its tag. The second function is seccomp_notify_ioctl().\nClang inlines seccomp_notify_addfd() there. Its -EBUSY is a plain failure\nnext to an -EINPROGRESS of another kind. It keeps five other errors.\n\nFixes: 4dd950adacad (\"resolve_btfids: Tag the functions that return errors\")\nReported-by: David Windsor \u003cdwindsor@gmail.com\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nLink: https://patch.msgid.link/20261114232754.2732901-1-alexei.starovoitov@gmail.com\nCloses: https://lore.kernel.org/bpf/CAEXv5_i17dJmGp3y_0U_f_xK9-SyC9M4GkLCx\u003dpT\u003d4VoXJ4abg@mail.gmail.com/\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "0751519b95d28342236a8c5e48027aede902b663",
      "old_mode": 33188,
      "old_path": "tools/bpf/resolve_btfids/fmodret.c",
      "new_id": "e1afdb6dedd56606440ed116d898830b27173ccd",
      "new_mode": 33188,
      "new_path": "tools/bpf/resolve_btfids/fmodret.c"
    },
    {
      "type": "modify",
      "old_id": "1db55d8d8d67bd7f75a74432ce4470b124d9026d",
      "old_mode": 33188,
      "old_path": "tools/testing/selftests/bpf/prog_tests/fmodret_retval.c",
      "new_id": "281c6192e5708756b99681b7f718ce02c9663fd7",
      "new_mode": 33188,
      "new_path": "tools/testing/selftests/bpf/prog_tests/fmodret_retval.c"
    },
    {
      "type": "modify",
      "old_id": "c6263d7e65bebb08d3d5eeeb5db009924cd8b852",
      "old_mode": 33188,
      "old_path": "tools/testing/selftests/bpf/test_kmods/bpf_testmod_retval.c",
      "new_id": "85722e48d34cef24b64c2ed80fd2bb1f8dacf16c",
      "new_mode": 33188,
      "new_path": "tools/testing/selftests/bpf/test_kmods/bpf_testmod_retval.c"
    }
  ]
}
