From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-pj1-f69.google.com (mail-pj1-f69.google.com [209.85.216.69]) (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 3B7874BB263 for ; Wed, 9 Sep 2026 22:48:26 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.69 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1788994112; cv=none; b=TYeoKqnHK1eyiNC05PJJJ5/+vwC7xALXhgawGZj86dPCjz6VHk6kS1EZg8lPjFju2aGnahuCrqzFce0HnGel88COgvO+Z5yTWBGhI6Vrd3wo792SmFp4c3mxS0PqZXoqi09u2FigSNc9ERsND2pCZwHROzAEU7Sgc/HDkirpTJk= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1788994112; c=relaxed/simple; bh=UaTP3cWHEqPR1qCZPKGgFnJo64FYOR3B0JkSwc/TFdQ=; h=Date:Mime-Version:Message-ID:Subject:From:To:Cc:Content-Type; b=m4L0ehOxhV8KI06h2gUE/jmy0eXpfWfFl/Cu/6URwSyfst09nVCIZSi2cXJYpUOwzJu/nWef/g3818m3FyXrPJ/Hhaf39wd6HU83ZR4+ILe/Q8G7KMPYfNirEJHs2uU+uENxsTtGmxIHXRxG4zszT2/Dh1ZGzruM5V+h01Lzzh0= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--mrungta.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=gLrbILLq; arc=none smtp.client-ip=209.85.216.69 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--mrungta.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="gLrbILLq" Received: by mail-pj1-f69.google.com with SMTP id 98e67ed59e1d1-398dcfabbf8so5412579a91.0 for ; Wed, 09 Sep 2026 15:48:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1788994100; x=1789598900; darn=lists.linux.dev; h=content-type:cc:to:from:subject:message-id:mime-version:date:from :to:cc:subject:date:message-id:reply-to:content-type; bh=UdIQWh93komvxnsnnPp25GJniHOVDvlbRHRdTzF4xOo=; b=gLrbILLqX23EX41gNprc0BQC3iz00blDIrFMnaJoF4jnlo0DLIGuweR6hpvBNWKyw5 GrKFC4Q5rzzwuYXNKHDD6AtQNu0DaAxNDBxz9K31aKHf7WTKX7kB62JOB3DCastBewLL YTW7M9JI/lhczqJvJBbTqwoN6TVAGJkkkBtIVsJlYMndg+7wDq1YO7pbE72FdemYB0PZ cvnUA0E01NQldYF4N3vLFqnXEMznSkog3XFfz0p3CwXGUOJpKJwHos1DtH4nGyEuJDQJ PUZcO7J/9DZhEH9PI5PG61gEl10tn2cLNhFI6US1RcN/Ku0yvC1vYXIxrDiDN7vn66t5 UlKA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1788994100; x=1789598900; h=content-type:cc:to:from:subject:message-id:mime-version:date :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to :content-type; bh=UdIQWh93komvxnsnnPp25GJniHOVDvlbRHRdTzF4xOo=; b=gyC1JQ7HSs11xKGAPIu5j0cOLTxFE0bQ1sYDK6YdSbc78AaLhw5+8K0ktyXbLBIAZR YIeiT6y1nINTx/47a072vWZTt3mLG1++NdUB5vO/qiG/RsCMMKDC4GhMXzaePLVfJyGu bZoxdNCc/zt3//JQnYPXfMLOULoQjTfMy1AZQCOFssU3eswgYq0trQy98KXfpE6tDQ/k bAim8w2TZfrje3+GAryI9lvEgayk3eaoja2lEKibrlPnJmDYHOm12ToNqllvf7+ZVC6r FEz+VlGzHE+9NqhrrIsTSZhszM5zhU5AyH6HFoSuciqYE+R7upHCiirA6zGg60mSMRSR 41SA== X-Forwarded-Encrypted: i=1; AKwUvBy0FTas1RDzDIpcEbdpTheeAZOTj4Ts+we6j+WDGoA0bMXBVVgsbjnn9dRrUlagZV1UOI3IQnAwxjd+zw==@lists.linux.dev X-Gm-Message-State: AFuF++kPMcnp5Z/yCQJnuG/LY+RmzVU73ZiETPXgVbwhV98z/djcrzrr NLM18JTTSqk9BdIJb4GX4G3CGC3WLTvcEiup5qv60Cz8LGGNc40crC9cWoHXCIoYIZsOAxMd3vC CcUApQy9Rfg== X-Received: from dlbbu10.prod.google.com ([2002:a05:7022:220a:b0:142:d53e:16e1]) (user=mrungta job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:5344:b0:38e:4f41:83df with SMTP id 98e67ed59e1d1-39b261e0f14mr56399683a91.15.1788994100079; Wed, 09 Sep 2026 15:48:20 -0700 (PDT) Date: Wed, 9 Sep 2026 15:48:18 -0700 Precedence: bulk X-Mailing-List: driver-core@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 X-Mailer: git-send-email 2.55.0.1003.g10538fe699-goog Message-ID: <20260909224818.2177311-1-mrungta@google.com> Subject: [PATCH v2] PM: sleep: Add DPM watchdog to prepare/late/early/noirq/complete phases From: Mayank Rungta To: "Rafael J . Wysocki" , Greg Kroah-Hartman , Danilo Krummrich Cc: Len Brown , Pavel Machek , Douglas Anderson , Tzung-Bi Shih , linux-pm@vger.kernel.org, driver-core@lists.linux.dev, linux-kernel@vger.kernel.org, Mayank Rungta Content-Type: text/plain; charset="UTF-8" Extend the DPM watchdog to wrap device_prepare, device_suspend_late, device_suspend_noirq, device_resume_noirq, device_resume_early, and device_complete callbacks. If a driver hangs during these transitions, the watchdog will fire and dump a stack trace to help identify the offending driver. To prevent false-positive timeouts, the watchdog is set only after waiting for subordinate (during suspend) and superior (during resume) devices. With this change, DPM watchdog coverage is extended across all phases of system sleep transitions. Signed-off-by: Mayank Rungta --- v2: - Extend DPM watchdog to device_prepare and device_complete callbacks. - Update commit subject and description to reflect coverage across all system sleep phases. v1: - Initial version adding DPM watchdog to late/early/noirq phases. Testing: - Tested on an ARM64 SoC (Pixel platform running a 6.18-based kernel with DPM watchdog enabled; Android 6.18 kernel has latest DPM changes backported). - Verified hang simulation in device_prepare and device_complete triggers DPM watchdog panic and stack trace as expected. drivers/base/power/main.c | 31 ++++++++++++++++++++++++++++--- 1 file changed, 28 insertions(+), 3 deletions(-) diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 184dc4b3b938..5e3e40312f71 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -802,6 +802,7 @@ static void device_resume_noirq(struct device *dev, pm_message_t state, bool asy const char *info = NULL; bool skip_resume; int error = 0; + DECLARE_DPM_WATCHDOG_ON_STACK(wd); TRACE_DEVICE(dev); TRACE_RESUME(0); @@ -827,6 +828,7 @@ static void device_resume_noirq(struct device *dev, pm_message_t state, bool asy if (!dpm_wait_for_superior(dev, async)) goto Out; + dpm_watchdog_set(&wd, dev); skip_resume = dev_pm_skip_resume(dev); /* * If the driver callback is skipped below or by the middle layer @@ -871,6 +873,7 @@ static void device_resume_noirq(struct device *dev, pm_message_t state, bool asy error = dpm_run_callback(callback, dev, state, info); Skip: + dpm_watchdog_clear(&wd); dev->power.is_noirq_suspended = false; Out: @@ -971,6 +974,7 @@ static void device_resume_early(struct device *dev, pm_message_t state, bool asy pm_callback_t callback = NULL; const char *info = NULL; int error = 0; + DECLARE_DPM_WATCHDOG_ON_STACK(wd); TRACE_DEVICE(dev); TRACE_RESUME(0); @@ -987,6 +991,7 @@ static void device_resume_early(struct device *dev, pm_message_t state, bool asy if (!dpm_wait_for_superior(dev, async)) goto Out; + dpm_watchdog_set(&wd, dev); if (dev->pm_domain) { info = "early power domain "; callback = pm_late_early_op(&dev->pm_domain->ops, state); @@ -1004,7 +1009,7 @@ static void device_resume_early(struct device *dev, pm_message_t state, bool asy goto Run; if (dev_pm_skip_resume(dev)) - goto Skip; + goto End; if (dev->driver && dev->driver->pm) { info = "early driver "; @@ -1014,6 +1019,9 @@ static void device_resume_early(struct device *dev, pm_message_t state, bool asy Run: error = dpm_run_callback(callback, dev, state, info); +End: + dpm_watchdog_clear(&wd); + Skip: dev->power.is_late_suspended = false; pm_runtime_enable(dev); @@ -1281,10 +1289,12 @@ static void device_complete(struct device *dev, pm_message_t state) { void (*callback)(struct device *) = NULL; const char *info = NULL; + DECLARE_DPM_WATCHDOG_ON_STACK(wd); if (dev->power.syscore) goto out; + dpm_watchdog_set(&wd, dev); device_lock(dev); if (dev->pm_domain) { @@ -1312,6 +1322,7 @@ static void device_complete(struct device *dev, pm_message_t state) } device_unlock(dev); + dpm_watchdog_clear(&wd); out: /* If enabling runtime PM for the device is blocked, unblock it. */ @@ -1508,6 +1519,7 @@ static void device_suspend_noirq(struct device *dev, pm_message_t state, bool as pm_callback_t callback = NULL; const char *info = NULL; int error = 0; + DECLARE_DPM_WATCHDOG_ON_STACK(wd); TRACE_DEVICE(dev); TRACE_SUSPEND(0); @@ -1520,6 +1532,7 @@ static void device_suspend_noirq(struct device *dev, pm_message_t state, bool as if (dev->power.syscore || dev->power.direct_complete) goto Complete; + dpm_watchdog_set(&wd, dev); if (dev->pm_domain) { info = "noirq power domain "; callback = pm_noirq_op(&dev->pm_domain->ops, state); @@ -1550,7 +1563,7 @@ static void device_suspend_noirq(struct device *dev, pm_message_t state, bool as WRITE_ONCE(async_error, error); dpm_save_failed_dev(dev_name(dev)); pm_dev_err(dev, state, async ? " async noirq" : " noirq", error); - goto Complete; + goto End; } Skip: @@ -1569,6 +1582,9 @@ static void device_suspend_noirq(struct device *dev, pm_message_t state, bool as if (dev->power.must_resume) dpm_superior_set_must_resume(dev); +End: + dpm_watchdog_clear(&wd); + Complete: complete_all(&dev->power.completion); TRACE_SUSPEND(error); @@ -1703,6 +1719,7 @@ static void device_suspend_late(struct device *dev, pm_message_t state, bool asy pm_callback_t callback = NULL; const char *info = NULL; int error = 0; + DECLARE_DPM_WATCHDOG_ON_STACK(wd); TRACE_DEVICE(dev); TRACE_SUSPEND(0); @@ -1720,6 +1737,8 @@ static void device_suspend_late(struct device *dev, pm_message_t state, bool asy if (dev->power.direct_complete) goto Complete; + dpm_watchdog_set(&wd, dev); + /* * After this point, any runtime PM operations targeting the device * will fail until the corresponding pm_runtime_enable() call in @@ -1761,13 +1780,16 @@ static void device_suspend_late(struct device *dev, pm_message_t state, bool asy dpm_save_failed_dev(dev_name(dev)); pm_dev_err(dev, state, async ? " async late" : " late", error); pm_runtime_enable(dev); - goto Complete; + goto End; } dpm_propagate_wakeup_to_parent(dev); Skip: dev->power.is_late_suspended = true; +End: + dpm_watchdog_clear(&wd); + Complete: TRACE_SUSPEND(error); complete_all(&dev->power.completion); @@ -2201,6 +2223,7 @@ static int device_prepare(struct device *dev, pm_message_t state) int (*callback)(struct device *) = NULL; bool smart_suspend; int ret = 0; + DECLARE_DPM_WATCHDOG_ON_STACK(wd); /* * If a device's parent goes into runtime suspend at the wrong time, @@ -2220,6 +2243,7 @@ static int device_prepare(struct device *dev, pm_message_t state) if (dev->power.syscore) return 0; + dpm_watchdog_set(&wd, dev); device_lock(dev); dev->power.wakeup_path = false; @@ -2245,6 +2269,7 @@ static int device_prepare(struct device *dev, pm_message_t state) unlock: device_unlock(dev); + dpm_watchdog_clear(&wd); if (ret < 0) { suspend_report_result(dev, callback, ret); -- 2.55.0.1003.g10538fe699-goog