Re: [PATCH 04/12] uprobes: revamp uprobe refcounting and lifetime management

[Andrii Nakryiko <andrii.nakryiko@gmail.com>]

On Tue, Jun 25, 2024 at 11:03 PM kernel test robot <lkp@intel.com> wrote:
>
> Hi Andrii,
>
> kernel test robot noticed the following build warnings:
>
> [auto build test WARNING on next-20240624]
> [also build test WARNING on v6.10-rc5]
> [cannot apply to perf-tools-next/perf-tools-next tip/perf/core perf-tools/perf-tools linus/master acme/perf/core v6.10-rc5 v6.10-rc4 v6.10-rc3]
> [If your patch is applied to the wrong git tree, kindly drop us a note.
> And when submitting patch, we suggest to use '--base' as documented in
> https://git-scm.com/docs/git-format-patch#_base_tree_information]
>
> url:    https://github.com/intel-lab-lkp/linux/commits/Andrii-Nakryiko/uprobes-update-outdated-comment/20240626-001728
> base:   next-20240624
> patch link:    https://lore.kernel.org/r/20240625002144.3485799-5-andrii%40kernel.org
> patch subject: [PATCH 04/12] uprobes: revamp uprobe refcounting and lifetime management
> config: x86_64-defconfig (https://download.01.org/0day-ci/archive/20240626/202406261300.ebbfM0XJ-lkp@intel.com/config)
> compiler: gcc-13 (Ubuntu 13.2.0-4ubuntu3) 13.2.0
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240626/202406261300.ebbfM0XJ-lkp@intel.com/reproduce)
>
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@intel.com>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202406261300.ebbfM0XJ-lkp@intel.com/
>
> All warnings (new ones prefixed by >>):
>
> >> kernel/events/uprobes.c:638: warning: Function parameter or struct member 'uprobe' not described in '__get_uprobe'
> >> kernel/events/uprobes.c:638: warning: expecting prototype for Caller has to make sure that(). Prototype was for __get_uprobe() instead
>
>
> vim +638 kernel/events/uprobes.c
>
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  625
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  626  /**

I shouldn't have used /** here, I'll fix this.

> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  627   * Caller has to make sure that:
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  628   *   a) either uprobe's refcnt is positive before this call;
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  629   *   b) or uprobes_treelock is held (doesn't matter if for read or write),
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  630   *      preventing uprobe's destructor from removing it from uprobes_tree.
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  631   *
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  632   * In the latter case, uprobe's destructor will "resurrect" uprobe instance if
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  633   * it detects that its refcount went back to being positive again inbetween it
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  634   * dropping to zero at some point and (potentially delayed) destructor
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  635   * callback actually running.
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  636   */
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  637  static struct uprobe *__get_uprobe(struct uprobe *uprobe)
> f231722a2b27ee Oleg Nesterov   2015-07-21 @638  {
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  639          s64 v;
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  640
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  641          v = atomic64_add_return(UPROBE_REFCNT_GET, &uprobe->ref);
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  642
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  643          /*
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  644           * If the highest bit is set, we need to clear it. If cmpxchg() fails,
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  645           * we don't retry because there is another CPU that just managed to
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  646           * update refcnt and will attempt the same "fix up". Eventually one of
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  647           * them will succeed to clear highset bit.
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  648           */
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  649          if (unlikely(v < 0))
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  650                  (void)atomic64_cmpxchg(&uprobe->ref, v, v & ~(1ULL << 63));
> b9adadbcb8dfc8 Andrii Nakryiko 2024-06-24  651
> f231722a2b27ee Oleg Nesterov   2015-07-21  652          return uprobe;
> f231722a2b27ee Oleg Nesterov   2015-07-21  653  }
> f231722a2b27ee Oleg Nesterov   2015-07-21  654
>
> --
> 0-DAY CI Kernel Test Service
> https://github.com/intel/lkp-tests/wiki