Received: from mail-wr1-f43.google.com (mail-wr1-f43.google.com [209.85.221.43]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1BB8A23C503 for ; Tue, 27 Jan 2026 19:07:04 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.221.43 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1769540826; cv=none; b=kCdoGS781mi1BEK/Pz7cGIWKlabgKo+479XwaF+DQ7BFwyfYedeorWPQq4exr3pYkTdN9Ha8ZDaeK6O7atpZ2Y90l/0Iin8YC3ye50rxifaunLda+SV+DCpU0OlCpu7HNz3YClZf10RH1qTBIG4iagaDXLv/mTPS+wdYiyacUPg= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1769540826; c=relaxed/simple; bh=FqVkiqS9rd727usiavtaBldcChihlNxSJdQkx5eLni4=; h=Message-ID:Date:MIME-Version:Subject:To:Cc:References:From: In-Reply-To:Content-Type; b=O4w3scO7JIdrBG9w3MOnQ+W2K9IJ0PW+mbWf5HvulcHrQxI52YiNbK4FKbBsIiw/58BcnrbK9jwlAHMfOrNpiT1hBK8U7Dk5bFtiRHALSofuvXpzICHKiW2/AMhhkg9xujaKy6hT0qXWcyQ8KF1N0dl29cqdQ6ZBSv+nL6VPxgM= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com; spf=pass smtp.mailfrom=gmail.com; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b=P2rgn1Cr; arc=none smtp.client-ip=209.85.221.43 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="P2rgn1Cr" Received: by mail-wr1-f43.google.com with SMTP id ffacd0b85a97d-43246af170aso102647f8f.0 for ; Tue, 27 Jan 2026 11:07:04 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1769540823; x=1770145623; darn=vger.kernel.org; h=content-transfer-encoding:in-reply-to:from:content-language :references:cc:to:subject:user-agent:mime-version:date:message-id :from:to:cc:subject:date:message-id:reply-to; bh=hUGAgqQ38zIXAYKYCeaxIKZeDwSnTqOgRfXtiSfl/WE=; b=P2rgn1CrJ1Jvf8TFmEwEzeoWY9JroTdgp99iP3XmJ/FVY49rZyOH7T4nfufsVCbZJa IT9WenAKfbnOeB9yMXjl9R40L8N38kpAvsd2fxYywOWMhzH1G6EyQBsgDWnzBfr8XFkp TPVOMrTLijnlP1ojl2MQWH9MqQbIdjdV5y4QgAu1g4U0XbFCd902wkDqS8PHjtBr9+wk E9GRx41gmtgmgzWk/dc3+y5BL1wLS38C2RfLjUwvJ2bAHdJofMwocnR3lZv3Sb0+cC3m 1bfbR4OAbJdYZPH1+tjrULUUu88QnYgwq9mIb3eKcWPv7Dgwneo1p85PuEY9HNfMLbbP g+yg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1769540823; x=1770145623; h=content-transfer-encoding:in-reply-to:from:content-language :references:cc:to:subject:user-agent:mime-version:date:message-id :x-gm-gg:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=hUGAgqQ38zIXAYKYCeaxIKZeDwSnTqOgRfXtiSfl/WE=; b=JbdsgpS0sd76HFTyehoUpl92CIS4zah5mKV6DMFJb73DEZG3MB0Uy/+JpBreLo9Qfe UKmOkXI+9YYEIbrazMc6+Z1gfzomuGpWEVje1Ook8MoIDtn1673EnQzdh39KZH9fczgS v4k+KI+AOW+Yg0MWjJAlaQMk+JkgjEebtmvQmghD8hqrlpUSERiJJP4ziFLkxkAQFzOe XFI56OhOxcEFa9jPGhISNWK9vXR/3KG72ua28I/CG+ylPH9QRAFKesVcf70aFtSBvU6p hrpdQ3Sn1W4sCy670pDweqt2perfy5eGkv5I7srOZUvO5wk6lpRYmXhxBn+KRd9Hf7/F U2Cg== X-Gm-Message-State: AOJu0YxqRcEkJc2EPq4EnsJRCu7brRerp63yCKwmKAPZFWhj41Zf7ttR SRz7TBImOTA0V0OUhDdNJPvM4r8VQcwemQzaooVn2G94Vo7P4ZMOgfN1 X-Gm-Gg: AZuq6aJ8cb0a+iTBEP0OxIH9H6lF8nQPM4kIy9xX0iWlywhFHLfpdjVF8mxm0Hog1n4 UgsQ9TYmbMcCEFVOWOaHDyjlcN9TLB0R6uNs/boCvWRnvCzOzRKUHGZDsYcElBxvcaz3kn8t2oJ Y9ukrrRtnRhSyjR2juaQboe1xbubmmkvkAYjg7Xid0cuXX7zhngrUBvhwq1vF32IQcFDi9o4yN1 u8bLl1zUKN8NaC0SKWOs3LH0bPJJrpGCowsqk8QCuan2URxIJ0gpWuJHEqlZ6J6YEss/qzH+Kpf p1hJEfRb/rkgyAJ/onQ23voSkta+BNEH0bPhdZnM0HnUmuHCDmFIYa91WnWcN+sTiiuOICNd/iW sXZ7vp9MACzLbPncVYILz2aioBdoWve2HrUIvzatfbI6Sf6w6oXxiF97GJa1blQyP5d4Nrd4up7 sC8jPLTlPG1VAI6r+KSJju8MnDGepG22GzSKP6HgW69A5iTeM= X-Received: by 2002:a05:6000:2483:b0:429:cb8b:b58e with SMTP id ffacd0b85a97d-435dd22fc21mr4840759f8f.28.1769540823271; Tue, 27 Jan 2026 11:07:03 -0800 (PST) Received: from [172.16.1.11] (i68975BB5.versanet.de. [104.151.91.181]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-435e10e476dsm906954f8f.4.2026.01.27.11.07.01 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Tue, 27 Jan 2026 11:07:02 -0800 (PST) Message-ID: <7689224e-6cf4-4be9-b8b1-686dc689fd5a@gmail.com> Date: Tue, 27 Jan 2026 19:07:01 +0000 Precedence: bulk X-Mailing-List: bpf@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH bpf-next v7 1/8] bpf: Enable bpf timer and workqueue use in NMI To: Kumar Kartikeya Dwivedi Cc: bpf@vger.kernel.org, ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, kafai@meta.com, kernel-team@meta.com, eddyz87@gmail.com, Mykyta Yatsenko References: <20260122-timer_nolock-v7-0-04a45c55c2e2@meta.com> <20260122-timer_nolock-v7-1-04a45c55c2e2@meta.com> Content-Language: en-US From: Mykyta Yatsenko In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit On 1/27/26 17:45, Kumar Kartikeya Dwivedi wrote: > Resending because I screwed up formatting. > > On Thu, 22 Jan 2026 at 19:03, Mykyta Yatsenko > wrote: >> From: Mykyta Yatsenko >> >> Refactor bpf timer and workqueue helpers to allow calling them from NMI >> context by making all operations lock-free and deferring NMI-unsafe >> work to irq_work. >> >> Previously, bpf_timer_start(), and bpf_wq_start() >> could not be called from NMI context because they acquired >> bpf_spin_lock and called hrtimer/schedule_work APIs directly. This >> patch removes these limitations. >> >> Key changes: >> * Remove bpf_spin_lock from struct bpf_async_kern. >> * Initialize/Destroy via setting/unsetting bpf_async_cb pointer >> atomically. >> * Add per-bpf_async_cb irq_work to defer NMI-unsafe >> operations (hrtimer_start, hrtimer_try_to_cancel, schedule_work) from >> NMI to softirq context. >> * Use the lock-free seqcount_latch_t to pass operation >> commands (start/cancel/free) and parameters >> from NMI-safe callers to the irq_work handler. >> * Add reference counting to bpf_async_cb to ensure the object stays >> alive until all scheduled irq_work completes. >> * Move bpf_prog_put() to RCU callback to handle races between >> set_callback() and cancel_and_free(). >> * Modify cancel_and_free() path: >> * Detach bpf_async_cb. >> * Signal destruction to irq_work side via setting last_seq to >> BPF_ASYNC_DESTROY. >> * On receiving BPF_ASYNC_DESTROY, cancel timer/wq. >> * Free bpf_async_cb on refcnt reaching 0, wait for both rcu and rcu >> task trace grace periods before freeing the bpf_async_cb. Removed >> unnecessary rcu locks, as kfunc/helper allways assumes rcu or rcu >> task trace lock. >> >> This enables BPF programs attached to NMI-context hooks (perf >> events) to use timers and workqueues for deferred processing. >> >> Signed-off-by: Mykyta Yatsenko >> --- > Acked-by: Kumar Kartikeya Dwivedi > >> [...] >> + >> +static int bpf_async_read_op(struct bpf_async_cb *cb, enum bpf_async_op *op, >> + u64 *nsec, u32 *flags) >> +{ >> + u32 seq, idx; >> + s64 last_seq; >> + >> + while (true) { >> + last_seq = atomic64_read_acquire(&cb->last_seq); >> + if (unlikely(last_seq > U32_MAX)) { /* Check if terminal seq num has been set */ >> + s64 expected = BPF_ASYNC_DESTROY; >> + >> + if (last_seq != BPF_ASYNC_DESTROY) >> + return -EBUSY; >> + >> + if (!atomic64_try_cmpxchg_release(&cb->last_seq, >> + &expected, >> + BPF_ASYNC_DESTROYED)) >> + return -EBUSY; /* Someone else set it to DESTROYED, bail */ >> + >> + *op = BPF_ASYNC_CANCEL; >> + return 0; >> + } >> + >> + seq = raw_read_seqcount_latch(&cb->latch); >> + >> + /* Return -EBUSY if current seq is consumed by another reader */ >> + if (seq == last_seq) >> + return -EBUSY; >> + >> + idx = seq & 1; >> + *nsec = cb->cmd[idx].nsec; >> + *flags = cb->cmd[idx].mode; >> + *op = cb->cmd[idx].op; >> + >> + if (raw_read_seqcount_latch_retry(&cb->latch, seq)) >> + continue; >> + >> + /* Commit read sequence number, own snapshot exclusively */ >> + if (atomic64_try_cmpxchg_release(&cb->last_seq, &last_seq, seq)) >> + break; > I guess you could fold in the change we discussed offline here: > > return atomic64_try_cmpxchg_release(&cb->last_seq, &last_seq, seq) ? 0 : -EBUSY; yes, thanks for help with the stress testing! > >> [...] >>