linux-kernel - Re: [PATCH v5 1/9] iommu: Introduce domain attachment handle

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Sun, 19 May 2024 18:07:42 +0800
From: Baolu Lu <baolu.lu@...ux.intel.com>
To: "Tian, Kevin" <kevin.tian@...el.com>, Jason Gunthorpe <jgg@...pe.ca>,
 Joerg Roedel <joro@...tes.org>, Will Deacon <will@...nel.org>,
 Robin Murphy <robin.murphy@....com>,
 Jean-Philippe Brucker <jean-philippe@...aro.org>,
 Nicolin Chen <nicolinc@...dia.com>, "Liu, Yi L" <yi.l.liu@...el.com>,
 Jacob Pan <jacob.jun.pan@...ux.intel.com>,
 Joel Granados <j.granados@...sung.com>
Cc: baolu.lu@...ux.intel.com, "iommu@...ts.linux.dev"
 <iommu@...ts.linux.dev>,
 "virtualization@...ts.linux-foundation.org"
 <virtualization@...ts.linux-foundation.org>,
 "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
 Jason Gunthorpe <jgg@...dia.com>
Subject: Re: [PATCH v5 1/9] iommu: Introduce domain attachment handle

On 5/15/24 3:17 PM, Tian, Kevin wrote:
>> From: Lu Baolu <baolu.lu@...ux.intel.com>
>> Sent: Tuesday, April 30, 2024 10:57 PM
>>
>> +/* Add an attach handle to the group's pasid array. */
>> +static struct iommu_attach_handle *
>> +iommu_attach_handle_set(struct iommu_domain *domain,
>> +			struct iommu_group *group, ioasid_t pasid)
>> +{
>> +	struct iommu_attach_handle *handle;
>> +	void *curr;
>> +
>> +	handle = kzalloc(sizeof(*handle), GFP_KERNEL);
>> +	if (!handle)
>> +		return ERR_PTR(-ENOMEM);
>> +
>> +	handle->domain = domain;
>> +	curr = xa_cmpxchg(&group->pasid_array, pasid, NULL, handle,
>> GFP_KERNEL);
> 
> this could be just xa_insert() which returns -EBUSY if the entry isn't NULL.

Yes, exactly.

> 
>> +	if (curr) {
>> +		kfree(handle);
>> +		return xa_err(curr) ? curr : ERR_PTR(-EBUSY);
> 
> 'curr' is not an error pointer. You need ERR_PTR(xa_err(curr)).

Fixed.

> 
>>   int iommu_attach_group(struct iommu_domain *domain, struct
>> iommu_group *group)
>>   {
>> +	struct iommu_attach_handle *handle;
>>   	int ret;
>>
>>   	mutex_lock(&group->mutex);
>> +	handle = iommu_attach_handle_set(domain, group,
>> IOMMU_NO_PASID);
>> +	if (IS_ERR(handle)) {
>> +		ret = PTR_ERR(handle);
>> +		goto out_unlock;
>> +	}
>>   	ret = __iommu_attach_group(domain, group);
>> +	if (ret)
>> +		goto out_remove_handle;
>>   	mutex_unlock(&group->mutex);
> 
> It's slightly better to set the handler after __iommu_attach_group().
> 
> doing so is aligned to the sequence in iommu_group_replace_domain().

I did it in this way because erasing an attach handle is easier than
rolling back the group to the previous domain from the perspective of
coding. I didn't realize attaching the group and storing the attach
handle had to be done in a specific order.

> 
>> @@ -2211,13 +2307,33 @@ EXPORT_SYMBOL_GPL(iommu_attach_group);
>>   int iommu_group_replace_domain(struct iommu_group *group,
>>   			       struct iommu_domain *new_domain)
>>   {
>> +	struct iommu_domain *old_domain = group->domain;
>> +	struct iommu_attach_handle *handle;
>>   	int ret;
>>
>>   	if (!new_domain)
>>   		return -EINVAL;
>>
>> +	if (new_domain == old_domain)
>> +		return 0;
>> +
> 
> __iommu_group_set_domain() already does the check.

Removed.

Best regards,
baolu