From 1da91ea87aefe2c25b68c9f96947a9271ba6325d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 31 May 2024 14:12:01 -0400 Subject: introduce fd_file(), convert all accessors to it. For any changes of struct fd representation we need to turn existing accesses to fields into calls of wrappers. Accesses to struct fd::flags are very few (3 in linux/file.h, 1 in net/socket.c, 3 in fs/overlayfs/file.c and 3 more in explicit initializers). Those can be dealt with in the commit converting to new layout; accesses to struct fd::file are too many for that. This commit converts (almost) all of f.file to fd_file(f). It's not entirely mechanical ('file' is used as a member name more than just in struct fd) and it does not even attempt to distinguish the uses in pointer context from those in boolean context; the latter will be eventually turned into a separate helper (fd_empty()). NOTE: mass conversion to fd_empty(), tempting as it might be, is a bad idea; better do that piecewise in commit that convert from fdget...() to CLASS(...). [conflicts in fs/fhandle.c, kernel/bpf/syscall.c, mm/memcontrol.c caught by git; fs/stat.c one got caught by git grep] [fs/xattr.c conflict] Reviewed-by: Christian Brauner Signed-off-by: Al Viro --- kernel/sys.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'kernel/sys.c') diff --git a/kernel/sys.c b/kernel/sys.c index 3a2df1bd9f64..a4be1e568ff5 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -1916,10 +1916,10 @@ static int prctl_set_mm_exe_file(struct mm_struct *mm, unsigned int fd) int err; exe = fdget(fd); - if (!exe.file) + if (!fd_file(exe)) return -EBADF; - inode = file_inode(exe.file); + inode = file_inode(fd_file(exe)); /* * Because the original mm->exe_file points to executable file, make @@ -1927,14 +1927,14 @@ static int prctl_set_mm_exe_file(struct mm_struct *mm, unsigned int fd) * overall picture. */ err = -EACCES; - if (!S_ISREG(inode->i_mode) || path_noexec(&exe.file->f_path)) + if (!S_ISREG(inode->i_mode) || path_noexec(&fd_file(exe)->f_path)) goto exit; - err = file_permission(exe.file, MAY_EXEC); + err = file_permission(fd_file(exe), MAY_EXEC); if (err) goto exit; - err = replace_mm_exe_file(mm, exe.file); + err = replace_mm_exe_file(mm, fd_file(exe)); exit: fdput(exe); return err; -- cgit v1.2.3 From ed4fb6d7ef68111bb539283561953e5c6e9a6e38 Mon Sep 17 00:00:00 2001 From: Felix Moessbauer Date: Wed, 14 Aug 2024 14:10:32 +0200 Subject: hrtimer: Use and report correct timerslack values for realtime tasks The timerslack_ns setting is used to specify how much the hardware timers should be delayed, to potentially dispatch multiple timers in a single interrupt. This is a performance optimization. Timers of realtime tasks (having a realtime scheduling policy) should not be delayed. This logic was inconsitently applied to the hrtimers, leading to delays of realtime tasks which used timed waits for events (e.g. condition variables). Due to the downstream override of the slack for rt tasks, the procfs reported incorrect (non-zero) timerslack_ns values. This is changed by setting the timer_slack_ns task attribute to 0 for all tasks with a rt policy. By that, downstream users do not need to specially handle rt tasks (w.r.t. the slack), and the procfs entry shows the correct value of "0". Setting non-zero slack values (either via procfs or PR_SET_TIMERSLACK) on tasks with a rt policy is ignored, as stated in "man 2 PR_SET_TIMERSLACK": Timer slack is not applied to threads that are scheduled under a real-time scheduling policy (see sched_setscheduler(2)). The special handling of timerslack on rt tasks in downstream users is removed as well. Signed-off-by: Felix Moessbauer Signed-off-by: Thomas Gleixner Link: https://lore.kernel.org/all/20240814121032.368444-2-felix.moessbauer@siemens.com --- fs/proc/base.c | 9 +++++---- fs/select.c | 11 ++++------- kernel/sched/syscalls.c | 8 ++++++++ kernel/sys.c | 2 ++ kernel/time/hrtimer.c | 18 +++--------------- 5 files changed, 22 insertions(+), 26 deletions(-) (limited to 'kernel/sys.c') diff --git a/fs/proc/base.c b/fs/proc/base.c index dd579332a7f8..afe573e8d8f1 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -2569,10 +2569,11 @@ static ssize_t timerslack_ns_write(struct file *file, const char __user *buf, } task_lock(p); - if (slack_ns == 0) - p->timer_slack_ns = p->default_timer_slack_ns; - else - p->timer_slack_ns = slack_ns; + if (task_is_realtime(p)) + slack_ns = 0; + else if (slack_ns == 0) + slack_ns = p->default_timer_slack_ns; + p->timer_slack_ns = slack_ns; task_unlock(p); out: diff --git a/fs/select.c b/fs/select.c index 9515c3fa1a03..ad171b7a5c11 100644 --- a/fs/select.c +++ b/fs/select.c @@ -77,19 +77,16 @@ u64 select_estimate_accuracy(struct timespec64 *tv) { u64 ret; struct timespec64 now; + u64 slack = current->timer_slack_ns; - /* - * Realtime tasks get a slack of 0 for obvious reasons. - */ - - if (rt_task(current)) + if (slack == 0) return 0; ktime_get_ts64(&now); now = timespec64_sub(*tv, now); ret = __estimate_accuracy(&now); - if (ret < current->timer_slack_ns) - return current->timer_slack_ns; + if (ret < slack) + return slack; return ret; } diff --git a/kernel/sched/syscalls.c b/kernel/sched/syscalls.c index ae1b42775ef9..195d2f2834a9 100644 --- a/kernel/sched/syscalls.c +++ b/kernel/sched/syscalls.c @@ -406,6 +406,14 @@ static void __setscheduler_params(struct task_struct *p, else if (fair_policy(policy)) p->static_prio = NICE_TO_PRIO(attr->sched_nice); + /* rt-policy tasks do not have a timerslack */ + if (task_is_realtime(p)) { + p->timer_slack_ns = 0; + } else if (p->timer_slack_ns == 0) { + /* when switching back to non-rt policy, restore timerslack */ + p->timer_slack_ns = p->default_timer_slack_ns; + } + /* * __sched_setscheduler() ensures attr->sched_priority == 0 when * !rt_policy. Always setting this ensures that things like diff --git a/kernel/sys.c b/kernel/sys.c index 3a2df1bd9f64..e3c4cffb520c 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -2557,6 +2557,8 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, error = current->timer_slack_ns; break; case PR_SET_TIMERSLACK: + if (task_is_realtime(current)) + break; if (arg2 <= 0) current->timer_slack_ns = current->default_timer_slack_ns; diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index f56ef23aad90..a023946f8558 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -2074,14 +2074,9 @@ long hrtimer_nanosleep(ktime_t rqtp, const enum hrtimer_mode mode, struct restart_block *restart; struct hrtimer_sleeper t; int ret = 0; - u64 slack; - - slack = current->timer_slack_ns; - if (rt_task(current)) - slack = 0; hrtimer_init_sleeper_on_stack(&t, clockid, mode); - hrtimer_set_expires_range_ns(&t.timer, rqtp, slack); + hrtimer_set_expires_range_ns(&t.timer, rqtp, current->timer_slack_ns); ret = do_nanosleep(&t, mode); if (ret != -ERESTART_RESTARTBLOCK) goto out; @@ -2251,7 +2246,7 @@ void __init hrtimers_init(void) /** * schedule_hrtimeout_range_clock - sleep until timeout * @expires: timeout value (ktime_t) - * @delta: slack in expires timeout (ktime_t) for SCHED_OTHER tasks + * @delta: slack in expires timeout (ktime_t) * @mode: timer mode * @clock_id: timer clock to be used */ @@ -2278,13 +2273,6 @@ schedule_hrtimeout_range_clock(ktime_t *expires, u64 delta, return -EINTR; } - /* - * Override any slack passed by the user if under - * rt contraints. - */ - if (rt_task(current)) - delta = 0; - hrtimer_init_sleeper_on_stack(&t, clock_id, mode); hrtimer_set_expires_range_ns(&t.timer, *expires, delta); hrtimer_sleeper_start_expires(&t, mode); @@ -2304,7 +2292,7 @@ EXPORT_SYMBOL_GPL(schedule_hrtimeout_range_clock); /** * schedule_hrtimeout_range - sleep until timeout * @expires: timeout value (ktime_t) - * @delta: slack in expires timeout (ktime_t) for SCHED_OTHER tasks + * @delta: slack in expires timeout (ktime_t) * @mode: timer mode * * Make the current task sleep until the given expiry time has -- cgit v1.2.3