linux-kernel - Re: [PART2 PATCH v3 06/11] iommu/amd: Adding GALOG interrupt handler

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] [day] [month] [year] [list]

Message-ID: <578638EE.6030200@amd.com>
Date:	Wed, 13 Jul 2016 19:49:50 +0700
From:	Suravee Suthikulpanit <Suravee.Suthikulpanit@....com>
To:	Radim Krčmář <rkrcmar@...hat.com>
CC:	<joro@...tes.org>, <pbonzini@...hat.com>,
	<alex.williamson@...hat.com>, <kvm@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, <sherry.hurwitz@....com>
Subject: Re: [PART2 PATCH v3 06/11] iommu/amd: Adding GALOG interrupt handler

Hi Radim,

On 07/12/2016 01:47 AM, Radim Krčmář wrote:
> 2016-07-11 05:11-0500, Suravee Suthikulpanit:
>> From: Suravee Suthikulpanit <Suravee.Suthikulpanit@....com>
>>
>> This patch adds AMD IOMMU guest virtual APIC log (GALOG) handler.
>> When IOMMU hardware receives an interrupt targeting a blocking vcpu,
>> it creates an entry in the GALOG, and generates an interrupt to notify
>> the AMD IOMMU driver.
>>
>> At this point, the driver processes the log entry, and notify the SVM
>> driver via the registered iommu_ga_log_notifier function.
>>
>> Signed-off-by: Suravee Suthikulpanit <Suravee.Suthikulpanit@....com>
>> ---
>> diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
>> @@ -741,14 +741,101 @@ static void iommu_poll_ppr_log(struct amd_iommu *iommu)
>> +static void iommu_handle_ga_guest_nr_entry(struct amd_iommu *iommu,
>> +					   u16 devid, u32 ga_tag)
>> +{
>> +	struct amd_ir_data *ir_data;
>> +	unsigned long flags;
>> +	int vec = 0;
>> +
>> +	if (!iommu_ga_log_notifier)
>> +		return;
>> +
>> +	spin_lock_irqsave(&iommu->ga_hash_lock, flags);
>> +	hash_for_each_possible(iommu->ga_hash, ir_data, hnode, ga_tag) {
>> +		struct irte_ga *entry = (struct irte_ga *)ir_data->entry;
>
>    if (ga_tag != entry->fields_vapic.ga_tag)
>    	continue;
>
> Multiple different ga_tags can map to the same bucket, so we need to
> explicitly select the correct entry.
>

In the upcoming V4 changes, we should not need this look up anymore.

>> +
>> +		vec = entry->hi.fields.vector;
>> +		break;
>> +	}
>> +	spin_unlock_irqrestore(&iommu->ga_hash_lock, flags);
>> +
>> +	if (vec) {
>> +		pr_debug("AMD-Vi: %s: devid=%#x, ga_tag=%#x\n",
>> +			 __func__, devid, ga_tag);
>> +
>> +		if (iommu_ga_log_notifier(GATAG_TO_AVICTAG(ga_tag),
>> +					  GATAG_TO_VCPUID(ga_tag), vec) != 0)


The necessary look up (i.e. to get vm and vcpu information should be 
done in the calling notifier here. I'll update this in V4, which getting 
ready to send out soon.

Thanks,
Suravee