Am Di., 28. Okt. 2025 um 09:46 Uhr schrieb Christian Brauner
<brauner@kernel.org>:
>
> This is not needed anymore now that we have the new scheme to guarantee
> all-or-nothing information exposure.
>
> Signed-off-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Alexander Mikhalitsyn <aleksandr.mikhalitsyn@canonical.com>
> ---
> fs/pidfs.c | 35 +++++++++++------------------------
> 1 file changed, 11 insertions(+), 24 deletions(-)
>
> diff --git a/fs/pidfs.c b/fs/pidfs.c
> index 0fad0c969b7a..a3b80be3b98b 100644
> --- a/fs/pidfs.c
> +++ b/fs/pidfs.c
> @@ -39,16 +39,6 @@ void pidfs_get_root(struct path *path)
> path_get(path);
> }
>
> -/*
> - * Stashes information that userspace needs to access even after the
> - * process has been reaped.
> - */
> -struct pidfs_exit_info {
> - __u64 cgroupid;
> - __s32 exit_code;
> - __u32 coredump_mask;
> -};
> -
> enum pidfs_attr_mask_bits {
> PIDFS_ATTR_BIT_EXIT = 0,
> };
> @@ -56,8 +46,11 @@ enum pidfs_attr_mask_bits {
> struct pidfs_attr {
> unsigned long attr_mask;
> struct simple_xattrs *xattrs;
> - struct pidfs_exit_info __pei;
> - struct pidfs_exit_info *exit_info;
> + struct /* exit info */ {
> + __u64 cgroupid;
> + __s32 exit_code;
> + };
> + __u32 coredump_mask;
> };
>
> static struct rb_root pidfs_ino_tree = RB_ROOT;
> @@ -313,7 +306,6 @@ static long pidfd_info(struct file *file, unsigned int cmd, unsigned long arg)
> struct pid *pid = pidfd_pid(file);
> size_t usize = _IOC_SIZE(cmd);
> struct pidfd_info kinfo = {};
> - struct pidfs_exit_info *exit_info;
> struct user_namespace *user_ns;
> struct pidfs_attr *attr;
> const struct cred *c;
> @@ -342,15 +334,15 @@ static long pidfd_info(struct file *file, unsigned int cmd, unsigned long arg)
> smp_rmb();
> kinfo.mask |= PIDFD_INFO_EXIT;
> #ifdef CONFIG_CGROUPS
> - kinfo.cgroupid = exit_info->cgroupid;
> + kinfo.cgroupid = attr->cgroupid;
> kinfo.mask |= PIDFD_INFO_CGROUPID;
> #endif
> - kinfo.exit_code = exit_info->exit_code;
> + kinfo.exit_code = attr->exit_code;
> }
> }
>
> if (mask & PIDFD_INFO_COREDUMP) {
> - kinfo.coredump_mask = READ_ONCE(attr->__pei.coredump_mask);
> + kinfo.coredump_mask = READ_ONCE(attr->coredump_mask);
> if (kinfo.coredump_mask)
> kinfo.mask |= PIDFD_INFO_COREDUMP;
> }
> @@ -629,7 +621,6 @@ void pidfs_exit(struct task_struct *tsk)
> {
> struct pid *pid = task_pid(tsk);
> struct pidfs_attr *attr;
> - struct pidfs_exit_info *exit_info;
> #ifdef CONFIG_CGROUPS
> struct cgroup *cgrp;
> #endif
> @@ -657,15 +648,13 @@ void pidfs_exit(struct task_struct *tsk)
> * is put
> */
>
> - exit_info = &attr->__pei;
> -
> #ifdef CONFIG_CGROUPS
> rcu_read_lock();
> cgrp = task_dfl_cgroup(tsk);
> - exit_info->cgroupid = cgroup_id(cgrp);
> + attr->cgroupid = cgroup_id(cgrp);
> rcu_read_unlock();
> #endif
> - exit_info->exit_code = tsk->exit_code;
> + attr->exit_code = tsk->exit_code;
>
> /* Ensure that PIDFD_GET_INFO sees either all or nothing. */
> smp_wmb();
> @@ -676,7 +665,6 @@ void pidfs_exit(struct task_struct *tsk)
> void pidfs_coredump(const struct coredump_params *cprm)
> {
> struct pid *pid = cprm->pid;
> - struct pidfs_exit_info *exit_info;
> struct pidfs_attr *attr;
> __u32 coredump_mask = 0;
>
> @@ -685,14 +673,13 @@ void pidfs_coredump(const struct coredump_params *cprm)
> VFS_WARN_ON_ONCE(!attr);
> VFS_WARN_ON_ONCE(attr == PIDFS_PID_DEAD);
>
> - exit_info = &attr->__pei;
> /* Note how we were coredumped. */
> coredump_mask = pidfs_coredump_mask(cprm->mm_flags);
> /* Note that we actually did coredump. */
> coredump_mask |= PIDFD_COREDUMPED;
> /* If coredumping is set to skip we should never end up here. */
> VFS_WARN_ON_ONCE(coredump_mask & PIDFD_COREDUMP_SKIP);
> - smp_store_release(&exit_info->coredump_mask, coredump_mask);
> + smp_store_release(&attr->coredump_mask, coredump_mask);
> }
> #endif
>
>
> --
> 2.47.3
>