qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Alejandro Jimenez <alejandro.j.jimenez@oracle.com>
To: "Igor Mammedov" <imammedo@redhat.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>
Cc: qemu-devel@nongnu.org,
	"Clément Mathieu--Drif" <clement.mathieu--drif@eviden.com>,
	"Zhao Liu" <zhao1.liu@intel.com>,
	"Hanna Reitz" <hreitz@redhat.com>,
	"Marcel Apfelbaum" <marcel.apfelbaum@gmail.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Eduardo Habkost" <eduardo@habkost.net>,
	qemu-block@nongnu.org, "Jason Wang" <jasowang@redhat.com>,
	"Richard Henderson" <richard.henderson@linaro.org>,
	"Yanan Wang" <wangyanan55@huawei.com>,
	"Ani Sinha" <anisinha@redhat.com>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	"Thomas Huth" <thuth@redhat.com>,
	"Gerd Hoffmann" <kraxel@redhat.com>,
	"Kevin Wolf" <kwolf@redhat.com>, "Yi Liu" <yi.l.liu@intel.com>,
	peterx@redhat.com
Subject: Re: [PATCH 18/18] hw/i386/x86-iommu: Remove X86IOMMUState::pt_supported field
Date: Thu, 12 Jun 2025 17:39:19 -0400	[thread overview]
Message-ID: <06a32bb9-087e-4508-881c-df26176ab71c@oracle.com> (raw)
In-Reply-To: <20250606154219.37957b34@imammedo.users.ipa.redhat.com>



On 6/6/25 9:42 AM, Igor Mammedov wrote:
> On Thu,  1 May 2025 23:04:56 +0200
> Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
> 
>> @@ -1486,15 +1485,8 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
>>                                               AMDVI_INT_ADDR_FIRST,
>>                                               &amdvi_dev_as->iommu_ir, 1);
>>   
>> -        if (!x86_iommu->pt_supported) {
>> -            memory_region_set_enabled(&amdvi_dev_as->iommu_nodma, false);
>> -            memory_region_set_enabled(MEMORY_REGION(&amdvi_dev_as->iommu),
>> -                                      true);
>> -        } else {
>> -            memory_region_set_enabled(MEMORY_REGION(&amdvi_dev_as->iommu),
>> -                                      false);
>> -            memory_region_set_enabled(&amdvi_dev_as->iommu_nodma, true);
>> -        }
>> +        memory_region_set_enabled(&amdvi_dev_as->iommu_nodma, false);
>> +        memory_region_set_enabled(MEMORY_REGION(&amdvi_dev_as->iommu), true);
> 
> given default is 'true', this hunk seems to be backwards,
> shouldn't it be 'else' branch code
> 

I believe this is trying to preserve the change recently introduced in:
31753d5a336f ("hw/i386/amd_iommu: Fix device setup failure when PT is on.")

However, that doesn't explain the logic of enabling those specific 
memory regions, which I think is also prone to cause confusion. I tried 
to explain the "trick" and argued against it here[0].

I am ultimately rewriting that code in a series to add DMA remap 
support[1], which hopefully makes it easier to follow which memory 
regions are active under specific configurations.

Thank you,
Alejandro

[0] 
https://lore.kernel.org/qemu-devel/914314b3-611d-4da3-9050-3c8c1b881e40@oracle.com/

[1] 
https://lore.kernel.org/qemu-devel/20250502021605.1795985-1-alejandro.j.jimenez@oracle.com/

> 
>>       }
>>       return &iommu_as[devfn]->as;
>>   }
>> diff --git a/hw/i386/intel_iommu.c b/hw/i386/intel_iommu.c
>> index c980cecb4ee..cc08dc41441 100644
>> --- a/hw/i386/intel_iommu.c
>> +++ b/hw/i386/intel_iommu.c
>> @@ -1066,6 +1066,7 @@ static inline bool vtd_ce_type_check(X86IOMMUState *x86_iommu,
>>   {
>>       switch (vtd_ce_get_type(ce)) {
>>       case VTD_CONTEXT_TT_MULTI_LEVEL:
>> +    case VTD_CONTEXT_TT_PASS_THROUGH:
>>           /* Always supported */
>>           break;
>>       case VTD_CONTEXT_TT_DEV_IOTLB:
>> @@ -1074,12 +1075,6 @@ static inline bool vtd_ce_type_check(X86IOMMUState *x86_iommu,
>>               return false;
>>           }
>>           break;
>> -    case VTD_CONTEXT_TT_PASS_THROUGH:
>> -        if (!x86_iommu->pt_supported) {
>> -            error_report_once("%s: PT specified but not supported", __func__);
>> -            return false;
>> -        }
>> -        break;
>>       default:
>>           /* Unknown type */
>>           error_report_once("%s: unknown ce type: %"PRIu32, __func__,
>> @@ -4520,7 +4515,7 @@ static void vtd_cap_init(IntelIOMMUState *s)
>>   {
>>       X86IOMMUState *x86_iommu = X86_IOMMU_DEVICE(s);
>>   
>> -    s->cap = VTD_CAP_FRO | VTD_CAP_NFR | VTD_CAP_ND |
>> +    s->cap = VTD_CAP_FRO | VTD_CAP_NFR | VTD_CAP_ND | VTD_ECAP_PT |
>>                VTD_CAP_MAMV | VTD_CAP_PSI | VTD_CAP_SLLPS |
>>                VTD_CAP_MGAW(s->aw_bits);
>>       if (s->dma_drain) {
>> @@ -4548,10 +4543,6 @@ static void vtd_cap_init(IntelIOMMUState *s)
>>           s->ecap |= VTD_ECAP_DT;
>>       }
>>   
>> -    if (x86_iommu->pt_supported) {
>> -        s->ecap |= VTD_ECAP_PT;
>> -    }
>> -
>>       if (s->caching_mode) {
>>           s->cap |= VTD_CAP_CM;
>>       }
>> diff --git a/hw/i386/x86-iommu.c b/hw/i386/x86-iommu.c
>> index d34a6849f4a..ca7cd953e98 100644
>> --- a/hw/i386/x86-iommu.c
>> +++ b/hw/i386/x86-iommu.c
>> @@ -129,7 +129,6 @@ static const Property x86_iommu_properties[] = {
>>       DEFINE_PROP_ON_OFF_AUTO("intremap", X86IOMMUState,
>>                               intr_supported, ON_OFF_AUTO_AUTO),
>>       DEFINE_PROP_BOOL("device-iotlb", X86IOMMUState, dt_supported, false),
>> -    DEFINE_PROP_BOOL("pt", X86IOMMUState, pt_supported, true),
>>   };
>>   
>>   static void x86_iommu_class_init(ObjectClass *klass, const void *data)
> 
> 



      reply	other threads:[~2025-06-12 21:40 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-05-01 21:04 [PATCH 00/18] hw/i386/pc: Remove deprecated 2.8 and 2.9 PC machines Philippe Mathieu-Daudé
2025-05-01 21:04 ` [PATCH 01/18] hw/i386/pc: Remove deprecated pc-q35-2.8 and pc-i440fx-2.8 machines Philippe Mathieu-Daudé
2025-06-03 15:21   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 02/18] hw/i386/pc: Remove pc_compat_2_8[] array Philippe Mathieu-Daudé
2025-06-03 15:21   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 03/18] hw/southbridge/ich9: Remove ICH9_LPC_SMI_F_BROADCAST_BIT definition Philippe Mathieu-Daudé
2025-06-03 15:25   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 04/18] hw/i386/kvm: Remove KVMClockState::mach_use_reliable_get_clock field Philippe Mathieu-Daudé
2025-06-03 15:27   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 05/18] hw/core/machine: Remove hw_compat_2_8[] array Philippe Mathieu-Daudé
2025-06-03 15:28   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 06/18] hw/block/pflash: Remove PFlashCFI01::old_multiple_chip_handling field Philippe Mathieu-Daudé
2025-06-05 15:06   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 07/18] hw/pci/pcie: Remove QEMU_PCIE_EXTCAP_INIT definition Philippe Mathieu-Daudé
2025-06-05 15:14   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 08/18] hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_INIT_DEVERR definition Philippe Mathieu-Daudé
2025-06-06 13:07   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 09/18] hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_INIT_LNKCTL definition Philippe Mathieu-Daudé
2025-06-06 13:10   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 10/18] hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_INIT_PM definition Philippe Mathieu-Daudé
2025-06-06 13:11   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 11/18] hw/nvram/fw_cfg: Remove FWCfgState::file_slots field Philippe Mathieu-Daudé
2025-06-06 13:24   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 12/18] hw/nvram/fw_cfg: Inline fw_cfg_file_slots() Philippe Mathieu-Daudé
2025-06-06 13:29   ` Igor Mammedov
2025-06-16  5:05     ` Philippe Mathieu-Daudé
2025-05-01 21:04 ` [PATCH 13/18] hw/i386/pc: Remove deprecated pc-q35-2.9 and pc-i440fx-2.9 machines Philippe Mathieu-Daudé
2025-06-06 13:30   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 14/18] hw/i386/pc: Remove pc_compat_2_9[] array Philippe Mathieu-Daudé
2025-06-06 13:31   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 15/18] hw/core/machine: Remove hw_compat_2_9[] array Philippe Mathieu-Daudé
2025-06-06 13:31   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 16/18] hw/net/virtio-net: Remove VirtIONet::mtu_bypass_backend field Philippe Mathieu-Daudé
2025-06-06 13:33   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 17/18] hw/pci-bridge/gen_pcie_rp: Remove GenPCIERootPort::migrate_msix field Philippe Mathieu-Daudé
2025-06-06 13:34   ` Igor Mammedov
2025-05-01 21:04 ` [PATCH 18/18] hw/i386/x86-iommu: Remove X86IOMMUState::pt_supported field Philippe Mathieu-Daudé
2025-06-06 13:42   ` Igor Mammedov
2025-06-12 21:39     ` Alejandro Jimenez [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=06a32bb9-087e-4508-881c-df26176ab71c@oracle.com \
    --to=alejandro.j.jimenez@oracle.com \
    --cc=anisinha@redhat.com \
    --cc=clement.mathieu--drif@eviden.com \
    --cc=eduardo@habkost.net \
    --cc=hreitz@redhat.com \
    --cc=imammedo@redhat.com \
    --cc=jasowang@redhat.com \
    --cc=kraxel@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=marcel.apfelbaum@gmail.com \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peterx@redhat.com \
    --cc=philmd@linaro.org \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.org \
    --cc=thuth@redhat.com \
    --cc=wangyanan55@huawei.com \
    --cc=yi.l.liu@intel.com \
    --cc=zhao1.liu@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).