From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mx0a-0031df01.pphosted.com (mx0a-0031df01.pphosted.com [205.220.168.131]) (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 DB04620E023; Fri, 10 Jan 2025 13:30:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=205.220.168.131 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1736515804; cv=none; b=NFdF0NEz4Y3jz8J9zoi5Ik8RyYIZxfD0fAviX+QqYTpd6TYk+B2gYD6UykgKR2DI88O/bzBw7yjB4A0h4AOw/lwjsIuyJtPejBaPf00qhwEuCn35Sd55lFitOHVuxP5qIC0/IE6jpIx0f4mD2NOMpzYtFQshYlFLhJIKddyFkEs= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1736515804; c=relaxed/simple; bh=5C4ZtgRDxqr9qSuURkK6QNcOIRV536fVjqYDChheTFw=; h=Message-ID:Date:MIME-Version:Subject:To:CC:References:From: In-Reply-To:Content-Type; b=PVv2Z0caFcjEOxr5Dyxtaq37Z+OkHCKjMRcaXhufpR0Q/lpdxx6yI6ZULapuY8JYN1Mjfybw2nUEPAopGPbNtwljOz1IP4hMTwf8HmTnQWucGFt2yTTtfM1No6gQE+zsuUQDGkPu1oYgGBwThfuCSyZhxO6Fqkv5kms5yfOoIGQ= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=quicinc.com; spf=pass smtp.mailfrom=quicinc.com; dkim=pass (2048-bit key) header.d=quicinc.com header.i=@quicinc.com header.b=HOIt6ESE; arc=none smtp.client-ip=205.220.168.131 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=quicinc.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=quicinc.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=quicinc.com header.i=@quicinc.com header.b="HOIt6ESE" Received: from pps.filterd (m0279862.ppops.net [127.0.0.1]) by mx0a-0031df01.pphosted.com (8.18.1.2/8.18.1.2) with ESMTP id 50A9sAsu027967; Fri, 10 Jan 2025 13:29:54 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=quicinc.com; h= cc:content-transfer-encoding:content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to; s=qcppdkim1; bh= 2Z9LJ3RXC8Zn5fQ8mt7kySq1gfbl4M8FWYlANQPAOQ8=; b=HOIt6ESEgK+n3fXg MmUIL//u6JIqZtcuuem9fTfuAgItj4TOAnRPvfp09YdusM9rqppfiJWyosYgv71q GcOsuraZex6OmkMMNYorHnI5EFuxclHaZfhQAnuVXjaJsBsa7QK4SLmuKiylxVNt pBwyCIsXo3+CFoWmFoK2JDgXr1TSOd2hWolkGF6xAR/8twV57TIqdycDtiF1S9Ft 4QZ60QxzNpvVdTzGo228CMV2eA633OKTYVOsaZSoWtV9lSDgT6aP1+NiCNWTE4D5 ccgrMH6QrKN5hEkcuMWFrkRRWl8L9+DEOEmnSMToYE4Onndd3MdJpH3U8xQjk7Bg GoICCw== Received: from nalasppmta01.qualcomm.com (Global_NAT1.qualcomm.com [129.46.96.20]) by mx0a-0031df01.pphosted.com (PPS) with ESMTPS id 4431as0j5s-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Fri, 10 Jan 2025 13:29:54 +0000 (GMT) Received: from nalasex01a.na.qualcomm.com (nalasex01a.na.qualcomm.com [10.47.209.196]) by NALASPPMTA01.qualcomm.com (8.18.1.2/8.18.1.2) with ESMTPS id 50ADTrd6026891 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Fri, 10 Jan 2025 13:29:53 GMT Received: from [10.216.20.159] (10.80.80.8) by nalasex01a.na.qualcomm.com (10.47.209.196) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.1544.9; Fri, 10 Jan 2025 05:29:51 -0800 Message-ID: Date: Fri, 10 Jan 2025 18:59:48 +0530 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH RFC] iommu: fix wrong DMA ops for iommu device To: Robin Murphy , Will Deacon CC: , , , References: <20241213150415.653821-1-quic_charante@quicinc.com> <91d15d11-ec53-4ca2-a385-9fa69d861f2c@quicinc.com> <20250103153434.GC3816@willie-the-truck> <20250107113126.GA6932@willie-the-truck> Content-Language: en-US From: Charan Teja Kalla In-Reply-To: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8bit X-ClientProxiedBy: nasanex01a.na.qualcomm.com (10.52.223.231) To nalasex01a.na.qualcomm.com (10.47.209.196) X-QCInternal: smtphost X-Proofpoint-Virus-Version: vendor=nai engine=6200 definitions=5800 signatures=585085 X-Proofpoint-GUID: qDCwT6KVc9LX_hEYn9riQ69pEmhYAjry X-Proofpoint-ORIG-GUID: qDCwT6KVc9LX_hEYn9riQ69pEmhYAjry X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.293,Aquarius:18.0.1039,Hydra:6.0.680,FMLib:17.12.60.29 definitions=2024-09-06_09,2024-09-06_01,2024-09-02_01 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 lowpriorityscore=0 adultscore=0 spamscore=0 mlxscore=0 priorityscore=1501 phishscore=0 mlxlogscore=999 suspectscore=0 malwarescore=0 bulkscore=0 impostorscore=0 clxscore=1015 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.19.0-2411120000 definitions=main-2501100106 Thanks Will and Robin for the response!! On 1/8/2025 6:07 PM, Robin Murphy wrote: > On 07/01/2025 11:31 am, Will Deacon wrote: >> On Mon, Jan 06, 2025 at 06:46:08PM +0530, Charan Teja Kalla wrote: >>> On 1/3/2025 9:04 PM, Will Deacon wrote: >>>> On Thu, Jan 02, 2025 at 04:09:46PM +0530, Charan Teja Kalla wrote: >>>>> On 12/13/2024 8:34 PM, Charan Teja Kalla wrote: >>>>>> Race between smmu device probe and iommu device probe is resulting >>>>>> into >>>>>> wrong DMA ops used for the device. >>>>>> >>>>>> bus_iommu_probe(P1)                  of_iommu_configure(P2) >>>>>> (from iommu_device_register)        (from insmod of a driver) >>>>>> --------------------------         -------------------------- >>>>>> 1) probe_iommu_group() >>>>>> (Fills just dev->iommu_group >>>>>> under iommu_probe_device_lock) >>>>>>                 2) iommu_probe_device(): >>>>>>                   (acquires the iommu_probe_device_lock, >>>>>>                    but since dev->iommu_group is already >>>>>>                    set, it just returns without >>>>>>                    allocating the domain.) >>>>>> >>>>>>                 3) of_dma_configure_id()-> >>>>>>                    arch_setup_dma_ops() gets called for >>>>>>                    this device, but returns with the >>>>>>                    error message: >>>>>>                    "Failed to set up IOMMU for device; >>>>>>                     retaining platform DMA ops" >>>>>> >>>>>>                 4) device probe is succeeded where it >>>>>>                    can now setup iommu mappings using >>>>>>                    wrong ->dma_ops >>>>>> >>>>>> 5) domain will be allocated later >>>>>> and fills iommu_group->domain. >>>>>> >>>>>> Step 3) and 4) denies iommu device from using 'iommu_dma_ops'. >>>>>> >>>>>> This issue does exists on 6.6 because of commit f188056352bc ("iommu: >>>>>> Avoid locking/unlocking for iommu_probe_device()") without which 2) >>>>>> returns only after filling the domain under group->mutex, thus no >>>>>> issue. >>>>>> >>>>>> With commit b5c58b2fdc42("dma-mapping: direct calls for dma-iommu"), >>>>>> ->iommu_dma_ops is removed altogether and iommu api are directly >>>>>> being >>>>>> called under "use_dma_iommu(): return dev->dma_iommu". Again, >>>>>> ->dma_iommu flag is set only after domain allocation. So, If >>>>>> there is a sufficient delay between steps 4) and 5), issue is still >>>>>> exists but that seems very narrow to me. >>>>>> >>>>>> I am thinking, setup of the domain before returning from step2) is >>>>>> the >>>>>> way to fix this issue. Can you please help if you have any >>>>>> suggestions? >>>>>> >>>>>> Signed-off-by: Charan Teja Kalla >>>>>> --- >>>>>>   drivers/iommu/iommu.c | 26 +++++++++++++++++++++++++- >>>>>>   1 file changed, 25 insertions(+), 1 deletion(-) >>>>> >>>>> Just a quick ask if you have any comments for the below. BTW, in >>>>> internal tests, when we revert the commit f188056352bc ("iommu: Avoid >>>>> locking/unlocking for iommu_probe_device()", it is helping. >>>>> >>>>> I can share more details If the below is not sufficient. >>>>> >>>>> Thanks in advance... >>>> >>>> Can you reproduce the issue on 6.13-rc5? We bodged some of the probe >>>> code recently: >>>> >>>> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/ >>>> commit/?id=229e6ee43d2a160a1592b83aad620d6027084aad >>>> >>>> and a better set of fixes are queued in -next. >>> >>> Yes, we did try a better patch from Robin[1] and the issue is still >>> reproducible. >> >> Did you try the -EPROBE_DEFER hack that got merged? >> >> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/ >> commit/?id=229e6ee43d2a160a1592b83aad620d6027084aad >> @Will, We just tested and the issue is still reported with just the mentioned hack that got merged. May be I will try to convince again with more details about why this is a race and why the -EPROBE_DEFER hack you mentioned will not help. S: SMMU device probe in bus_iommu_probe(). C: SMMU client device probe in of_iommu_configure(). 1) S: In probe_iommu_group(): Fills the client device ->iommu_group under iommu_probe_device_lock. But iommu_group->domain is not yet allocated because it is a 2 step process from probe_iommu_group. 2) C: iommu_probe_device() --> __iommu_probe_device(): Acquires the iommu_probe_device_lock, and since dev->iommu_group is already set, it just returns without allocating the domain. static int __iommu_probe_device(struct device *dev, ...) { /* Device is probed already if in a group */ if (dev->iommu_group) return 0; --> __driver call returns from here.__ ret = iommu_init_device(dev, ops); if (ret) return ret; } The change you mentioned will come into play iff it enters iommu_init_device(), where it returns -EPROBE_DEFER till smmu driver bound happens. __Here we didn't even enter into this function__. 3) C: Continues probing and calls of_dma_configure_id()->arch_setup_dma_ops()->iommu_setup_dma_ops(). Since ->domain is not yet allocated, it just returns without setting 'iommu_dma_ops' and with a warn message of 'Failed to set up IOMMU for device' for this client device. And also the client probe is successfully completed. void iommu_setup_dma_ops(struct device *dev, u64 dma_base, u64 dma_limit) { if (!domain) goto out_err; if (iommu_is_dma_domain(domain)) { ..... dev->dma_ops = &iommu_dma_ops;<-- This step will not be executed. } out_err: pr_warn("Failed to set up IOMMU for device %s; retaining platform DMA ops\n", dev_name(dev)); } 4) S: domain will be allocated later in the second step and fills iommu_group->domain. Since, the SMMU client device probe is already successfully completed with invalid dma_ops and when client uses the dma api can lead to issues. >> Otherwise, you can try the stack queued for 6.14 in iommu/linux.git: >> >> 46b3df8eb9bd iommu: Manage driver probe deferral better >> fcbd62156742 iommu/arm-smmu-v3: Clean up more on probe failure >> 97cb1fa02726 iommu/arm-smmu: Retire probe deferral workaround >> 7d835134d4e1 iommu/arm-smmu: Make instance lookup robust >> >>> IIUC, these patches are trying to get valid smmu device, during the >>> probe of an smmu client device, by traversing proper 'klist_devices' >>> happening from iommu_init_device(). Please CMIW. >> >> Well, they're fixing a race between the IOMMU probing and its client >> devices attaching. >> >>> But this bug makes iommu_init_device() to completely gets skipped. >>> >>> static int __iommu_probe_device(struct device *dev, ...) >>> { >>>     /* Device is probed already if in a group */ >>>          if (dev->iommu_group) >>>                  return 0;  --> driver call returns from here. >>> >>>          ret = iommu_init_device(dev, ops); >>>          if (ret) >>>                  return ret; >>> >>> >>> } >>> >>> Regarding the testing, we work on Android that is on LTS kernels, so, it >>> may not be possible for us to try the thing on 6.13 and this seems a bug >>> on LTS 6.6 kernel. >> >> We need to fix 6.13 before we fix 6.6 unless you can show that 6.13 is >> unaffected (in which case, we can backport the fix(es)). > > Certainly the reasoning can't apply to mainline as given, since > arch_setup_dma_ops() stopped touching IOMMU stuff at all back in 6.10 > (and indeed iommu_dma_ops itself no longer exists since 6.12). > @Robin Agree that we don't have iommu_dma_ops but I can say that same race still exists. Although dma_ops is not used, but the decision to call into the iommu api's is determined by the 'dev->dma_iommu' flag, which again, is set after domain is allocated for a device. In the same race mentioned above, 1) S: Domain is not allocated but the dev->iommu_group. 2) C: Just returns as dev->iommu_group is filled. 3) C: Continues probing and succeeds. 4) C: Calls dma_alloc/map/.... But, it won't enter into iommu_ api's because the 'dev->dma_iommu' is still 'false'. 5) S: Domain is allocated and sets the 'dev->dma_iommu' to 'true'. 4) above is the problematic step. Although issue exists but seems to me that very narrow to get triggered. Please CMIW. > Robin. Thanks.