[PATCH v2 7/9] iommu/arm-smmu: remove arm_smmu_devices
leizhen
thunder.leizhen at huawei.com
Wed Jul 8 19:43:29 PDT 2015
On 2015/7/8 21:13, Robin Murphy wrote:
> On 07/07/15 04:30, Zhen Lei wrote:
>> It can be replaced by of_iommu_list(in of_iommu.c).
>>
>> Signed-off-by: Zhen Lei <thunder.leizhen at huawei.com>
>> ---
>> drivers/iommu/arm-smmu-v3.c | 22 ++--------------------
>> 1 file changed, 2 insertions(+), 20 deletions(-)
>>
>> diff --git a/drivers/iommu/arm-smmu-v3.c b/drivers/iommu/arm-smmu-v3.c
>> index c569539..39c55f6 100644
>> --- a/drivers/iommu/arm-smmu-v3.c
>> +++ b/drivers/iommu/arm-smmu-v3.c
>> @@ -604,10 +604,6 @@ struct arm_smmu_domain {
>> struct iommu_domain domain;
>> };
>>
>> -/* Our list of SMMU instances */
>> -static DEFINE_SPINLOCK(arm_smmu_devices_lock);
>> -static LIST_HEAD(arm_smmu_devices);
>> -
>> struct arm_smmu_option_prop {
>> u32 opt;
>> const char *prop;
>> @@ -2675,11 +2671,6 @@ static int arm_smmu_device_dt_probe(struct platform_device *pdev)
>> if (ret)
>> goto out_free_structures;
>>
>> - /* Record our private device structure */
>> - INIT_LIST_HEAD(&smmu->list);
>> - spin_lock(&arm_smmu_devices_lock);
>> - list_add(&smmu->list, &arm_smmu_devices);
>> - spin_unlock(&arm_smmu_devices_lock);
>> platform_set_drvdata(pdev, smmu);
>> of_iommu_set_ops(smmu->dev->of_node, &arm_smmu_ops);
>>
>> @@ -2692,19 +2683,10 @@ out_free_structures:
>>
>> static int arm_smmu_device_remove(struct platform_device *pdev)
>> {
>> - struct arm_smmu_device *curr, *smmu = NULL;
>> + struct arm_smmu_device *smmu;
>> struct device *dev = &pdev->dev;
>>
>> - spin_lock(&arm_smmu_devices_lock);
>> - list_for_each_entry(curr, &arm_smmu_devices, list) {
>> - if (curr->dev == dev) {
>> - smmu = curr;
>> - list_del(&smmu->list);
>> - break;
>> - }
>> - }
>> - spin_unlock(&arm_smmu_devices_lock);
>> -
>> + smmu = find_smmu_by_node(dev->of_node);
>
> Isn't that just a really long round-trip to platform_get_drvdata(pdev)?
But we should first check that pdev is a real smmu device, not others. So we can not omit:
if (!of_iommu_get_ops(np))
I think arm_smmu_device_remove will rarely be called, so it will not impact performance.
>
>> if (!smmu)
>> return -ENODEV;
>>
>> --
>> 1.8.0
>>
>>
>> _______________________________________________
>> iommu mailing list
>> iommu at lists.linux-foundation.org
>> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>>
>
>
> .
>
More information about the linux-arm-kernel
mailing list