From nobody Mon Jun 8 19:55:00 2026 Received: from spam.asrmicro.com (asrmicro.com [210.13.118.86]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 6723A2BEC23; Wed, 27 May 2026 04:33:00 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=210.13.118.86 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1779856384; cv=none; b=dhoo6a4/9TL7+uFNIIXoFdXNbO5Ux+n5UGKew/AXUQovcVg1cAalLNx6iN9zEujILs/V3fym9L8agaERhVees15e+J6dKal1TspeIG/9bE8ewYuPKAQYgWF8BsE1wVXhqPTW5pBIAc/PUNhotRTnvN+eESYtRXcLAbapIbXRKsc= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1779856384; c=relaxed/simple; bh=FRfwH3MIKqmKPNdllLLIaxCN3KeQ27UpXSQdkzfsalA=; h=From:To:CC:Subject:Date:Message-ID:MIME-Version:Content-Type; b=ZjMWYVr8g655IHrRXhLfNBPA/KXu6QU2lUaOhf47xOzwPatH/prQbJd4w9CmRgxywUz8ke8/wt5wcSEsPoaSsl1QXevbxNnaKdmuL740QgsVseTTMZ+JuSA7aCyJal5eWnxzETKe5tds5/61wP4eNjnf0wZt80Vv95hpRLp7W2U= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=asrmicro.com; spf=pass smtp.mailfrom=asrmicro.com; arc=none smtp.client-ip=210.13.118.86 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=asrmicro.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=asrmicro.com Received: from exch02.asrmicro.com (exch02.asrmicro.com [10.1.24.122]) by spam.asrmicro.com with ESMTPS id 64R4Vh2p099197 (version=TLSv1 cipher=ECDHE-RSA-AES256-SHA bits=256 verify=FAIL); Wed, 27 May 2026 12:31:43 +0800 (GMT-8) (envelope-from hongjiefang@asrmicro.com) Received: from localhost (10.1.170.248) by exch02.asrmicro.com (10.1.24.122) with Microsoft SMTP Server (TLS) id 15.0.847.32; Wed, 27 May 2026 12:31:45 +0800 From: Hongjie Fang To: , , , , CC: , Subject: [PATCH v3] scsi: core: pair EH runtime PM get/put Date: Wed, 27 May 2026 12:31:44 +0800 Message-ID: <20260527043144.255619-1-hongjiefang@asrmicro.com> X-Mailer: git-send-email 2.25.1 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: exch03.asrmicro.com (10.1.24.118) To exch02.asrmicro.com (10.1.24.122) X-DNSRBL: X-SPAM-SOURCE-CHECK: pass X-MAIL: spam.asrmicro.com 64R4Vh2p099197 Content-Type: text/plain; charset="utf-8" shost->eh_noresume is currently consulted twice in one error handling iteration: once before scsi_autopm_get_host() and once again before scsi_autopm_put_host(). That is racy when a PM-triggered error path flips shost->eh_noresume while the SCSI EH thread is still running. The problem flow looks like this: PM path ufshcd_set_dev_pwr_mode() shost->eh_noresume =3D 1 ufshcd_execute_start_stop <-- trigger EH ... shost->eh_noresume =3D 0 EH path scsi_error_handler() if (!shost->eh_noresume) scsi_autopm_get_host() <-- skipped ... if (!shost->eh_noresume) scsi_autopm_put_host() <-- executed later In that case one EH iteration can skip autoresume on entry and still drop a runtime PM reference on exit. That leaves an unmatched runtime PM put and can trigger a runtime PM usage count underflow. Fix this by calling scsi_autopm_put_host() only if the same iteration successfully acquired a runtime PM reference through the scsi_autopm_get_host(). Fixes: ae0751ffc77e ("[SCSI] add flag to skip the runtime PM calls on the h= ost") Signed-off-by: Hongjie Fang --- drivers/scsi/scsi_error.c | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index 147127fb4db9..c78ea64641a9 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -2342,6 +2342,7 @@ static void scsi_unjam_host(struct Scsi_Host *shost) int scsi_error_handler(void *data) { struct Scsi_Host *shost =3D data; + bool autopm_put; =20 /* * We use TASK_INTERRUPTIBLE so that the thread is not @@ -2383,12 +2384,16 @@ int scsi_error_handler(void *data) * what we need to do to get it up and online again (if we can). * If we fail, we end up taking the thing offline. */ - if (!shost->eh_noresume && scsi_autopm_get_host(shost) !=3D 0) { - SCSI_LOG_ERROR_RECOVERY(1, - shost_printk(KERN_ERR, shost, - "scsi_eh_%d: unable to autoresume\n", - shost->host_no)); - continue; + autopm_put =3D false; + if (!shost->eh_noresume) { + if (scsi_autopm_get_host(shost) !=3D 0) { + SCSI_LOG_ERROR_RECOVERY(1, + shost_printk(KERN_ERR, shost, + "scsi_eh_%d: unable to autoresume\n", + shost->host_no)); + continue; + } + autopm_put =3D true; } =20 if (shost->transportt->eh_strategy_handler) @@ -2407,7 +2412,7 @@ int scsi_error_handler(void *data) * which are still online. */ scsi_restart_operations(shost); - if (!shost->eh_noresume) + if (autopm_put) scsi_autopm_put_host(shost); } __set_current_state(TASK_RUNNING); --=20 2.25.1