From 7d59065aa4bfb3f0a9d9ce70f8bdad3f894b9d1c Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 14 Nov 2023 09:55:50 -0700 Subject: [PATCH] BACKPORT: io_uring/fdinfo: remove need for sqpoll lock for thread/pid retrieval A previous commit added a trylock for getting the SQPOLL thread info via fdinfo, but this introduced a regression where we often fail to get it if the thread is busy. For that case, we end up not printing the current CPU and PID info. Rather than rely on this lock, just print the pid we already stored in the io_sq_data struct, and ensure we update the current CPU every time we've slept or potentially rescheduled. The latter won't potentially be 100% accurate, but that wasn't the case before either as the task can get migrated at any time unless it has been pinned at creation time. We retain keeping the io_sq_data dereference inside the ctx->uring_lock, as it has always been, as destruction of the thread and data happen below that. We could make this RCU safe, but there's little point in doing that. With this, we always print the last valid information we had, rather than have spurious outputs with missing information. Bug: 254441685 Fixes: 7644b1a1c9a7 ("io_uring/fdinfo: lock SQ thread while retrieving thread cpu/pid") Signed-off-by: Jens Axboe (cherry picked from commit a0d45c3f596be53c1bd8822a1984532d14fdcea9) [Lee: PF_NO_SETAFFINITY has been removed upstream causing surrounding diff to change] Signed-off-by: Lee Jones Change-Id: Iab4aeffb443ed5140b33dd7274fee57c36f40f9e --- io_uring/fdinfo.c | 9 ++------- io_uring/sqpoll.c | 12 ++++++++++-- 2 files changed, 12 insertions(+), 9 deletions(-) diff --git a/io_uring/fdinfo.c b/io_uring/fdinfo.c index ea2c2ded4e41..42d8461ac546 100644 --- a/io_uring/fdinfo.c +++ b/io_uring/fdinfo.c @@ -141,13 +141,8 @@ static __cold void __io_uring_show_fdinfo(struct io_ring_ctx *ctx, if (has_lock && (ctx->flags & IORING_SETUP_SQPOLL)) { struct io_sq_data *sq = ctx->sq_data; - if (mutex_trylock(&sq->lock)) { - if (sq->thread) { - sq_pid = task_pid_nr(sq->thread); - sq_cpu = task_cpu(sq->thread); - } - mutex_unlock(&sq->lock); - } + sq_pid = sq->task_pid; + sq_cpu = sq->sq_cpu; } seq_printf(m, "SqThread:\t%d\n", sq_pid); diff --git a/io_uring/sqpoll.c b/io_uring/sqpoll.c index 7b6facf529b8..6f9c78bdd4ec 100644 --- a/io_uring/sqpoll.c +++ b/io_uring/sqpoll.c @@ -214,6 +214,7 @@ static bool io_sqd_handle_event(struct io_sq_data *sqd) did_sig = get_signal(&ksig); cond_resched(); mutex_lock(&sqd->lock); + sqd->sq_cpu = raw_smp_processor_id(); } return did_sig || test_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state); } @@ -229,10 +230,15 @@ static int io_sq_thread(void *data) snprintf(buf, sizeof(buf), "iou-sqp-%d", sqd->task_pid); set_task_comm(current, buf); - if (sqd->sq_cpu != -1) + /* reset to our pid after we've set task_comm, for fdinfo */ + sqd->task_pid = current->pid; + + if (sqd->sq_cpu != -1) { set_cpus_allowed_ptr(current, cpumask_of(sqd->sq_cpu)); - else + } else { set_cpus_allowed_ptr(current, cpu_online_mask); + sqd->sq_cpu = raw_smp_processor_id(); + } current->flags |= PF_NO_SETAFFINITY; mutex_lock(&sqd->lock); @@ -262,6 +268,7 @@ static int io_sq_thread(void *data) mutex_unlock(&sqd->lock); cond_resched(); mutex_lock(&sqd->lock); + sqd->sq_cpu = raw_smp_processor_id(); } continue; } @@ -295,6 +302,7 @@ static int io_sq_thread(void *data) mutex_unlock(&sqd->lock); schedule(); mutex_lock(&sqd->lock); + sqd->sq_cpu = raw_smp_processor_id(); } list_for_each_entry(ctx, &sqd->ctx_list, sqd_list) atomic_andnot(IORING_SQ_NEED_WAKEUP,